Research Article Distributive Implications of Fertility ...
12
Research Article Distributive Implications of Fertility Changes in Latin America Nicolás Badaracco, Leonardo Gasparini, and Mariana Marchionni Centro de Estudios Distributivos, Laborales y Sociales (CEDLAS), Facultad de Ciencias Econ´ omicas, Universidad Nacional de La Plata and CONICET, 1900 La Plata, Argentina Correspondence should be addressed to Nicol´ as Badaracco; [email protected] Received 13 February 2016; Accepted 22 May 2016 Academic Editor: Jonathan Haughton Copyright © 2016 Nicol´ as Badaracco et al. is is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Fertility rates significantly fell over the last decades in Latin America. In order to assess the extent to which these changes contributed to the observed reduction in income poverty and inequality, we apply microeconometric decomposition to microdata from national household surveys from seven Latin American countries. We find that changes in fertility rates were associated with a nonnegligible reduction in inequality and poverty in the region. e main channel was straightforward: lower fertility implied smaller families and hence larger per capita incomes. Lower fertility also fostered labor force participation, especially among women, which contributed to the reduction of poverty and inequality in most countries, although the size of this effect was smaller. 1. Introduction Fertility rates have been significantly falling during the last decades in Latin America. e average number of children decreased in families from all population groups, but the decline was sharper among poor households: the fertility gap between the most and the least vulnerable groups shrunk in the region. e distributive impact of these demographic changes could be sizeable. Ceteris paribus, a fall in the number of children in poor households and in those marginally above the poverty line, reduces income poverty, whereas heterogeneous changes in family size across income groups could reduce income inequality. In this paper, we assess the extent to which changes in fertility in seven Latin American countries (Argentina, Brazil, Chile, El Salvador, Mexico, Peru, and Uruguay) contributed to the observed reduction in income poverty and inequality during the 1990s and 2000s. To that aim, we simulate the counterfactual household per capita income distribution if fertility outcomes in a given year would have been deter- mined as in another different year. e term fertility is used as a shortcut for the number of children in the household, which in most cases changes as a consequence of fertility decisions. Of course, actual changes in poverty and inequality are driven by various factors, including some of those affecting fertility. Estimating a general equilibrium model or complex structural equations that take into account all the interactions is not feasible, given lack of data, among other limitations. Instead, in this paper, we follow a more modest but potentially useful methodology: we compute the first-round partial- equilibrium impact on the income distribution of changes in fertility. at impact could be taken as an approximation of the magnitude that fertility changes may have on poverty and inequality. In particular, we examine three effects: first, more children in the household imply a reduction in current per capita income, as a similar budget should be divided among more people; second, higher fertility may affect the intensive and extensive margins of labor supply among adults in the household; and, third, it may also affect the income transfers received through social programs targeted at families with children. If we observe that family size declines for the poor more than for the rich, it is rather obvious that poverty and inequality measured over the distribution of household current income per capita will also decline. is paper makes two contributions to this intuition. First, it provides estimates of the magnitude of the direct distributive impact of the changes in fertility, that is, the extent to which the actual decline in poverty and inequality can be accounted for by only the change in the reproductive behavior. Second, the methodology allows tracing and measuring some not-so- obvious effects. e fall in the number of children in the Hindawi Publishing Corporation International Journal of Population Research Volume 2016, Article ID 8717265, 11 pages http://dx.doi.org/10.1155/2016/8717265
Transcript of Research Article Distributive Implications of Fertility ...
Research ArticleDistributive Implications of Fertility Changes in Latin America
Nicolaacutes Badaracco Leonardo Gasparini and Mariana Marchionni
Centro de Estudios Distributivos Laborales y Sociales (CEDLAS) Facultad de Ciencias EconomicasUniversidad Nacional de La Plata and CONICET 1900 La Plata Argentina
Correspondence should be addressed to Nicolas Badaracco badaracconicolasgmailcom
Received 13 February 2016 Accepted 22 May 2016
Academic Editor Jonathan Haughton
Copyright copy 2016 Nicolas Badaracco et al This is an open access article distributed under the Creative Commons AttributionLicense which permits unrestricted use distribution and reproduction in any medium provided the original work is properlycited
Fertility rates significantly fell over the last decades in LatinAmerica In order to assess the extent towhich these changes contributedto the observed reduction in income poverty and inequality we applymicroeconometric decomposition tomicrodata fromnationalhousehold surveys from seven Latin American countriesWe find that changes in fertility rates were associated with a nonnegligiblereduction in inequality and poverty in the regionThemain channel was straightforward lower fertility implied smaller families andhence larger per capita incomes Lower fertility also fostered labor force participation especially among women which contributedto the reduction of poverty and inequality in most countries although the size of this effect was smaller
1 Introduction
Fertility rates have been significantly falling during the lastdecades in Latin America The average number of childrendecreased in families from all population groups but thedecline was sharper among poor households the fertility gapbetween the most and the least vulnerable groups shrunkin the region The distributive impact of these demographicchanges could be sizeableCeteris paribus a fall in the numberof children in poor households and in those marginallyabove the poverty line reduces income poverty whereasheterogeneous changes in family size across income groupscould reduce income inequality
In this paper we assess the extent to which changes infertility in seven LatinAmerican countries (Argentina BrazilChile El Salvador Mexico Peru and Uruguay) contributedto the observed reduction in income poverty and inequalityduring the 1990s and 2000s To that aim we simulate thecounterfactual household per capita income distribution iffertility outcomes in a given year would have been deter-mined as in another different yearThe term fertility is used asa shortcut for the number of children in the household whichin most cases changes as a consequence of fertility decisionsOf course actual changes in poverty and inequality aredriven by various factors including some of those affectingfertility Estimating a general equilibrium model or complex
structural equations that take into account all the interactionsis not feasible given lack of data among other limitationsInstead in this paper we follow amoremodest but potentiallyuseful methodology we compute the first-round partial-equilibrium impact on the income distribution of changes infertility That impact could be taken as an approximation ofthe magnitude that fertility changes may have on poverty andinequality In particular we examine three effects first morechildren in the household imply a reduction in current percapita income as a similar budget should be divided amongmore people second higher fertility may affect the intensiveand extensive margins of labor supply among adults in thehousehold and third it may also affect the income transfersreceived through social programs targeted at families withchildren
If we observe that family size declines for the poormore than for the rich it is rather obvious that povertyand inequality measured over the distribution of householdcurrent income per capita will also declineThis paper makestwo contributions to this intuition First it provides estimatesof the magnitude of the direct distributive impact of thechanges in fertility that is the extent to which the actualdecline in poverty and inequality can be accounted for byonly the change in the reproductive behavior Second themethodology allows tracing and measuring some not-so-obvious effects The fall in the number of children in the
Hindawi Publishing CorporationInternational Journal of Population ResearchVolume 2016 Article ID 8717265 11 pageshttpdxdoiorg10115520168717265
2 International Journal of Population Research
bottom strata of the distribution may induce some low-income women to enter the labor market or to work morehours In that case the decline in poverty and inequalitymight be larger than expected if one only considers the directimpact of the decline in family size In addition a reductionin fertility may reduce the income transfers received by poorhouseholds in terms of social assistance directed to familieswith children
In this paper we find that the changes in fertility thattook place over the last decades in Latin America contributedto the reduction in income inequality as well as in povertyThe fall in fertility among the income-deprived helped toreduce the incidence of income poverty Besides since it waslarger than the fall among the nonpoor it contributed to thereduction in income inequality as wellThemain channel wasstraightforward lower fertility rates implied smaller familiesand hence larger per capita incomes Lower fertility alsofostered labor force participation especially among womenwhich contributed to the reduction of poverty and inequalityin most countries although the size of this effect was small
The rest of the paper is organized in a straightforwardwayIn the next section we document changes in fertility acrossincome groups and present trends in income inequality andpoverty in Latin America In Section 3 we lay out themethodology discuss its limitations and introduce the dataused in the estimations Section 4 presents the main resultsof the paper whereas Section 5 closes with some concludingremarks
2 Fertility Changes and Distributive Trends
In the early 1960s the total fertility rate (TFR) was around6 children per woman in Latin America higher than theworldrsquos average (total fertility rate is the average numberof children a hypothetical cohort of women would haveat the end of their reproductive period if during theirwhole lives they were subject to the fertility rates of agiven period and if they were not subject to mortalityit is expressed as children per woman (United NationsWorld Population Prospects The 2012 Revision definitionavailable at httpesaunorgwppExcel-Datafertilityhtm))It was not until the mid-1960s that TFR began its downwardtrend which eventually led to clear convergence towards thelevels of the most advanced regions of the world (Figure 1) Inparticular theTFR continued falling over the last twodecadesin Latin America reaching an average of 25 children perwoman in 2005ndash2010 (this decreasing trend has been presentin developed countries since the beginning of the past century[1 2] For an analysis of fertility dynamics in Latin Americasee Chackiel [3] and ECLAC [4 5]
Following a similar trend the number of children perhousehold (the proxy for fertility that can be implementedwith household survey microdata) has been falling over thelast two decades In 1992 there were 23 children under theage of 16 per household (for households with a woman aged25ndash45) in a typical Latin American country the figure was 18in 2012 [6]
The average number of children decreased in householdsfrom all population groups but the gap between the most
0010203040506070
1950
ndash195
5
1955
ndash196
0
1960
ndash196
5
1965
ndash197
0
1970
ndash197
5
1975
ndash198
0
1980
ndash198
5
1985
ndash199
0
1990
ndash199
5
1995
ndash200
0
2000
ndash200
5
2005
ndash201
0
World
Latin AmericaThe Caribbean
North America
Europe and Central AsiaEast Asia and the Pacific
South Asia
Middle East and North Africa
Sub-Saharan Africa
Figure 1 Total fertility rate (children per woman) Regions of theworld 1950ndash2010 Source Gasparini and Marchionni [6] based onWorld Population Prospects The 2012 Revision DVD Edition
10121416182022242628
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
Richest 20Poorest 20
Figure 2 Number of children under 16 per household Bottom andtop quintiles of parental income distribution 1992ndash2012 Sourceown calculations based on microdata from national household sur-veys Note average of seven Latin American countries ArgentinaBrazil Chile El Salvador Mexico Peru and Uruguay Householdswith head aged 25ndash45 Unweighted means
and the least vulnerable groups shrank owing to a sharperdecline in the number of children living in poor householdsFigure 2 takes a sample of seven Latin American countries toshow the average number of children under 16 per householdin the poorest 20 and richest 20 of the prime-age parents(the sample includes countries in which we implement themicrosimulations that follow in the next section) In bothgroups fertility went down over the two decades the fall issomewhat more pronounced among the poorest couplesThegap in the number of children per household shrank from067 in 1992 to 047 in 2012
With regard to the income distribution Latin Americaexperienced an increase in inequality during the 1990s whilepoverty slightly decreased [7ndash9] In contrast in the 2000sthere were strong reductions of both income inequality and
International Journal of Population Research 3
15
20
25
30
35
40
4519
9219
9319
9419
9519
9619
9719
9819
9920
0020
0120
0220
0320
0420
0520
0620
0720
0820
0920
1020
1120
12
Headcount ratio (US$ 4 line)
45
47
49
51
53
55
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
Gini coefficient
Figure 3 Poverty and income inequality Latin America Source own calculations based on microdata from national household surveysNote average of seven Latin American countries Argentina Brazil Chile El Salvador Mexico Peru and Uruguay Unweighted meansPoverty and inequality computed over the distribution of household per capita income
poverty Figure 3 shows the average income inequality andpoverty trends of seven Latin American countries Povertyis measured by the headcount ratio with a daily US$ 4 linewhile income inequality is measured by the Gini coefficientBoth indices are calculated over the household per capitaincome distribution Income poverty slightly decreased dur-ing the 1990s and fell sharply during the 2000s ArgentinaUruguay and Mexico suffered an increase in poverty duringthe 1990s and then a continuous fall over the following decade(Figure 4) Meanwhile Brazil Chile El Salvador and Peruexperienced a decrease in poverty over the entire periodthough steeper during the 2000s
On average for our sample of seven countries incomeinequality increased during the first decade under analysisand then significantly decreased during the 2000s The trendfor the 1990s is driven by the sharp increase in incomedispersion inArgentina andUruguay (Figure 5) and the smallchanges in the rest of the countries In contrast during the2000s the fall in inequality was strong and generalized to alleconomies
3 Data and Methodology
We focus on seven Latin American countries ArgentinaBrazil Chile El Salvador Mexico Peru and Uruguay duringthe 1990s and 2000sThe analysis is based onmicrodata fromnational household surveys collected by the correspondingNational Statistical Offices Table 1 presents a brief descrip-tion of these databases Since surveys are not homogeneouswe have made all possible efforts to make variables compara-ble across countries and over time by using similar definitionsof variables in each countryyear and by applying consistentmethods of processing the data
The ideal variable for capturing fertility would be thenumber of children either living in the household or notUnfortunately this variable is usually not available in LatinAmerican household surveys with the exception of BrazilAs a proxy we use the number of children living in thehousehold Consequently two main problems emerge First
Table 1 Household surveys
Country Survey Acronym Years
Argentina Encuesta Permanente deHogares EPH 1992ndash2012
Brazil Pesquisa Nacional porAmostra de Domicilios PNAD 1990ndash2012
ChileEncuesta de
CaracterizacionSocioeconomica Nacional
CASEN 1990ndash2011
El Salvador Encuesta de Hogares dePropositos Multiples EHPM 1991ndash2010
MexicoEncuesta Nacional deIngresos y Gasto de los
HogaresENIGH 1992ndash2012
Peru Encuesta Nacional deHogares ENAHO 1997ndash2012
Uruguay Encuesta Continua deHogares ECH 1995ndash2012
the measurement error of this proxy depends on the ageof parents and children that is on average the older theparents and the older the children are the more likely theyleave the parental home Therefore the observed differencesin the number of children across households are not onlydue to reproductive decisions but also due to the life cycleTo deal with this shortcoming we restrict our sample tohouseholds where the head of household is between 25 and45 years old Many people below 25 years old may not haveyet started childbearing The average age at childbearing forthe countries under analysis was 275 years in 1993 and 276years in 2011 The dispersion across countries is also smallranging from 265 in Mexico to 283 in Peru The average ageat marriage in the early 1990s was 229 and 256 years forwomen and men respectively and about two years more inthe early 2010s (250 and 275)
Another problem arises since Latin American house-hold surveys usually register family relationships among
4 International Journal of Population Research
0102030405060
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
0102030405060
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
0102030405060
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
0102030405060
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
0102030405060
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
0102030405060
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
0102030405060
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
Argentina Brazil Chile
El Salvador Mexico Peru
Uruguay
Figure 4 PovertyHeadcount ratio (US$ 4 line) Source own calculations based onmicrodata fromnational household surveys Note povertycomputed over the distribution of household per capita income
household members only in terms of the head of householdthat is it is possible to match children with their parentsonly when they are the head of household or the spouseTherefore we further restrict the analysis to only considerfertility decisions of the head of household and the spouse
Regarding the empirical specification and estimationmethodology we follow Marchionni and Gasparini [10]The main inputs to carry out the microsimulations arethe estimates of the parameters that govern fertility deci-sionsoutcomes (for simplicity we refer to fertility decisionsalthough fertility outcomes could be the result of free con-scious choices but also the consequence of various other cir-cumstances) and the response of labor market participationto changes in family size We assume that the number ofchildren in a household follows a Poisson process and thatits parameters can be consistently estimated using a Poissonregression model Hourly wages and hours of work areassumed to be simultaneously determined in an equilibriummodel of the labor market
After estimating the parameters we carry out the sim-ulations That is we simulate the counterfactual incomedistribution that arises in a given base year by assumingthat the population in that year takes fertility decisions
according to the parameters estimated for a different yearTheresulting poverty and inequality measures over the simulateddistribution are compared to those actually observed in thebase year The difference between the simulated value ofan indicator of poverty or inequality and its actual value isinterpreted as a measure of the direct impact of the change infertility behavior that is the impact that would take place incase all other covariates remained fixed
The methodology requires keeping all other things con-stant when simulating the impact of changes in fertilitydecisions Naturally as it was stressed in previous sectionssome of these factors may be correlated with fertility Forinstance the structure ofwagesmay respond to changes in thelabor supply triggered by a change in fertility Additionallychanges in reproductive behavior may have been inducedby income changes in which case the microsimulation onlycaptures a single round of effects from fertility to incomes ofa more complicated process
In a given period income poverty and inequality areaffected by a host of factors including economic shocks andsocial policies The microsimulation methodology is usefulas it allows approximating the size of the impact that agiven factor would have in case all other factors remained
International Journal of Population Research 5
40
45
50
55
60
6519
9219
9419
9619
9820
0020
0220
0420
0620
0820
1020
12
40
45
50
55
60
65
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
40
45
50
55
60
65
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
40
45
50
55
60
65
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
40
45
50
55
60
65
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
40
45
50
55
60
65
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
40
45
50
55
60
65
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
Argentina Brazil Chile
El Salvador Mexico Peru
Uruguay
Figure 5 Income inequality Gini coefficient Source own calculations based onmicrodata fromnational household surveys Note inequalitycomputed over the distribution of household per capita income
fixed Although this is clearly an unrealistic scenario theestimations are still useful as a first approximation of theimpact especially considering the fact that the more ambi-tious alternative of estimating a general equilibriummodel isusually unfeasible
31 Microsimulations We carry out the simulation over thedistribution of household per capita income defined as
119910119894119905=
119884119871
ℎ119905+ 119884119873119871
ℎ119905
119873ℎ119905
forall119894 isin ℎ at time 119905 (1)
where 119894 indexes individuals ℎ households and 119905 time periods(years) 119884119871
ℎ119905denotes total labor income of household ℎ at time
119905 119884119873119871
ℎ119905labels nonlabor income and 119873
ℎ119905is the family size
which is the sum of the number of children up to 16 yearsold119867
ℎ119905and the rest of the household members 119877
ℎ119905
119873ℎ119905
= 119867ℎ119905+ 119877ℎ119905 (2)
Weneed to specify how119867ℎ119905is determined in order to simulate
the counterfactual distributions The literature argues that
fertility outcomes are the result of a process affected bycharacteristics of each spouse and of the household amongother factors This process can be formalized as
119867ℎ119905
= 119867 (119885ℎ119905 119890ℎ119905 120578119905) (3)
where 119885ℎ119905
is a vector of observable characteristics and119890ℎ119905
includes all unobservable characteristics that influencethe familyrsquos reproductive behavior while 120578
119905is the set of
parameters that govern fertility decisionsWe assume that labor income for individual 119894 is given by
119884119871
119894119905= 119908119894119905119871119894119905 (4)
where 119908119894119905
is the hourly wage rate and 119871119894119905
is the num-ber of hours worked by individual 119894 Wages and hoursof work are defined as 119908
119894119905= 119908(119883
1119894119905 120576119882
119894119905 120573119905) and 119871
119894119905=
119871(1198832119894119905 119867119894119905 120576119871
119894119905 120574119905 120582119905) respectively Where 119883
1119894119905and 119883
2119894119905are
observable characteristics 120576119882119894119905
and 120576119871
119894119905are unobservable char-
acteristics while 120573119905 120574119905 and 120582
119905are parameters In particular
120582119905is the parameter that relates the hours of work to the
number of children
6 International Journal of Population Research
Nonlabor income is usually assumed to be exogenous inthemicrosimulation literature However as stated in previoussections the expansion of social programs to families withchildren suggests a potential link between fertility decisionsand nonlabor income Accordingly we assume that nonlaborincome is given by
119884119873119871
ℎ119905= 119898ℎ119905+ 119892 (119867
ℎ119905) (5)
where119898ℎ119905is exogenous and119892(119867
ℎ119905) represents the component
of nonlabor income that depends on the number of childrenWe label the parameters of the fertility decisions at time 119905
as 120578119905and those at time 1199051015840 as 120578
1199051015840 A key step in themethodology
is to estimate the counterfactual number of children in agiven year 119905 if fertility outcomes were determined as in analternative year 119905
1015840 Once this term labeled as 119867ℎ119905(1205781199051015840) is
estimated three microsimulation exercises are carried out byreplacing this estimate in the household per capita incomeequation
The first exercise aims to capture the contribution of thechange in fertility parameters 120578 to the actual change in theincome distribution through the denominator of (1) Thesimulated per capita income is given by
119910119863
119894119905=
119884119871
ℎ119905+ 119884119873119871
ℎ119905
119873ℎ119905(1205781199051015840)
(6)
where 119873ℎ119905(1205781199051015840) = 119877
ℎ119905+ 119867ℎ119905(1205781199051015840) Superscript 119863 in the
simulated per capita income stands for the direct-size effectthat is the change in the income distribution due to changesin the number of household members among whom totalhousehold income should be distributed
The second exercise involves simulating the labor incomethat a household would have in year 119905 if parameters thatgovern fertility decisionswere those of year 1199051015840 In other wordsthe exercise consists in replacing the simulated number ofchildren 119867
ℎ119905(1205781199051015840) in the equation of individual labor income
(4) to obtain
119910119867
119894119905=
119884119871
ℎ119905(1205781199051015840) + 119884
119873119871
ℎ119905
119873ℎ119905
(7)
Superscript 119867 in the simulated per capita income stands forthe hours-size effect that is the contribution of the changein fertility parameters 120578 to the actual change in the incomedistribution through the indirect channel of affecting thehours-of-work decisions
We carry out a third exercise by simulating the coun-terfactual distribution arising from a change in nonlaborincome driven by changes in fertility decisions We refer tothe distributional impact of changes through this channelas the nonlabor-income effect and thus label the per capitaincome variable in (8) with the NL superscript
119910119873119871
119894119905=
119884119871
ℎ119905+ 119884119873119871
ℎ119905(1205781199051015840)
119873ℎ119905
(8)
Finally (9) considers the three channels simultaneously toobtain the total effect (superscript119879) of changes in the fertilityparameters
119910119879
119894119905=
119884119871
ℎ119905(1205781199051015840) + 119884
119873119871
ℎ119905(1205781199051015840)
119873ℎ119905(1205781199051015840)
(9)
We denote the per capita income distribution among indi-viduals in year 119905 as 119863
119905(119910119905) = (119910
1119905 1199102119905 119910
119899119905) where 119899
denotes the total number of individuals in the populationLet 119868(sdot) be a distributive indicator as the Gini coefficientor the poverty headcount ratio Then the contribution ofthe change in fertility parameters 120578 to the actual change inthe income distribution through the direct channel is givenby (10) The contributions through the other channels arecomputed analogously
119868 (119863119905(119910119905)) minus 119868 (119863
119905(119910119863
119905)) (10)
A shortcoming of the microsimulation methodology is thatdecomposition is path-dependent This means that takingyear 119905 or year 1199051015840 as the base year does not necessarily implythe same result Accordingly we perform the simulationsalternating the base year and report the average effect for eachexercise
In the following subsections we provide details concern-ing the specification and estimation of fertility labor incomeand nonlabor-income equations
32 Fertility Decisions We assume that the number of chil-dren in (3) follows a Poisson process with parameter 120583
ℎ119905
Formally
119867ℎ119905
sim Poisson (120583ℎ119905) with 120583
ℎ119905= 119864 (119867
ℎ119905| 119885ℎ119905) = exp (119885
1015840
ℎ119905120578119905) (11)
Then
Prob (119867ℎ119905
= 1198670) =
exp (minus120583ℎ119905) (120583ℎ119905)1198670
1198670
with 1198670= 0 1 2 (12)
Equation (12) is the Poisson regressionmodel fromwhich it ispossible to consistently estimate parameters 120578
119905by maximum
likelihood (it can be shown that consistency holds for themaximum likelihood estimators of 120578
119905as long as the real
distribution is any of the linear exponential family to whichthe Poisson distribution belongs provided that the condi-tional mean in (11) is correctly specified see eg Wooldridge[11]) We estimate separate models for two-parent householdand single-parent households For simplicity we denote theestimates by 120578
Once parameters 120578 are estimated we proceed to simulatechanges in the number of children Since the objective isto simulate these changes as a consequence of changes onlyin the parameters 120578 it is necessary to keep unobservablefactors fixed Therefore we characterize each household bythe quantile it occupies in the distribution of children ofyear 119905 Let 119865
120578119905|119885ℎ119905
(sdot) be the cumulative probability function ofa random variable that follows a Poisson distribution withparameter equal to exp(1198851015840
ℎ119905120578119905) and let 119902
ℎ119905be the quantile
for household ℎ at time 119905 that is 119865120578119905|119885ℎ119905
(119867ℎ119905) = 119902
ℎ119905 The
International Journal of Population Research 7
simulated number of children in household ℎ will be the onethat places it in the 119902
ℎ119905quantile of the distribution of children
with the relevant parameters of time 1199051015840(1205781199051015840) conditional to
the observable characteristics 119885ℎ119905 Formally the simulated
number of children in household ℎ at year 119905 using theestimated fertility parameters for year 1199051015840 is given by
119867ℎ119905(1205781199051015840) = 119865
minus1
1205781199051015840 |119885ℎ119905
∘ 119865120578119905|119885ℎ119905
(119867ℎ119905) (13)
33 Labor Incomes To estimate individual labor income in(4) we follow Gasparini et al [12] in assuming that bothwages and hours worked are determined in a reduced-formmodel of the labor market equilibrium
ln119908lowast
119894119905= 1198831015840
1119894119905120573119905+ 120576119882
119894119905 (14)
119871lowast
119894119905= 1198831015840
2119894119905120574119905+ 119867119894119905120582119905+ 120576119871
119894119905 (15)
with
if 119871lowast119894119905gt 0
119908119894119905= 119908lowast
119894119905
119871119894119905= 119871lowast
119894119905
if 119871lowast119894119905le 0
119908119894119905= 0
119871119894119905= 0
(120576119882
119894119905 120576119871
119894119905) sim 119873(0 0 120590
2
119882119905 1205902
119871119905 120588119905)
(16)
where119908lowast119894119905and119871lowast
119894119905are unobserved latent variables Vectors119883
1119894119905
and 1198832119894119905
include observable factors affecting hourly wagesand hours of work respectively while 119867
119894119905stands for the
number of children 120573119905 120574119905(vectors) and 120582
119905are parameters
to be estimated along with 1205902
119882119905 1205902119871119905 and 120588
119905
We estimate (14) by Heckmanrsquos maximum likelihoodmethod where a censored version of (15) is used as a selectionequation replacing hours of work by a binary variable thatindicates whether the individual works For estimation of(15) we use aTobitmodel (although this estimation strategy isnot fully efficient efficiency loss is not necessarily significantfor a given sample size and this alternative has certaincomputational advantages over a full information procedurefor more details see Gasparini et al [12]) We estimateseparate models for head of households and spouses
To simulate changes in individual labor incomes wereplace the observed number of children of individual 119894 atyear 119905 in (15) by the simulated number of children 119867
119905ℎ(1205781199051015840)
while assigning himher a wage based on (14) and an errorterm drawn from the bivariate distribution implicit in themodel
34 Nonlabor Incomes The measure of nonlabor incomeincludes pensions capital income and transfers Conditionalcash transfers are included as government transfers Wesimulate nonlabor incomes in an ad hoc manner taking
into account the characteristics of each countryrsquos socialprogram In those cases where it is not possible to identifythe recipients in the household survey we identify potentialbeneficiaries and estimate the amount of the transfers basedon the design of each program The general framework is asfollows nonlabor income of household ℎ at year 119905 is given by(5) where 119892(119867
ℎ119905) represents the part of nonlabor income that
depends on the number of children Typically cash transfersin poverty-alleviation programs vary with the number ofchildren in the family For example if the program has a fixedamount 120579
119905per child the nonlabor income becomes
119884119873119871
ℎ119905= 119898ℎ119905+ 120579119905119867ℎ119905 (17)
The simulated nonlabor income is calculated using thecounterfactual number of children in a given year 119905 if fertilityoutcomes were determined as in an alternative year 1199051015840
119884119873119871
ℎ119905(1205781199051015840) = 119898
ℎ119905+ 120579ℎ119905119867ℎ119905(1205781199051015840) (18)
Even though all programs have a similar setting to the onedescribed they differ in some aspects maximum number ofchildren age and in some cases how to identify the programrsquosrecipients The details are given in the Appendix We sim-ulate nonlabor incomes for Argentina Brazil Mexico andUruguay Even though the programs in Chile El Salvadorand Peru have similar characteristics it is not possible toestimate how the amount of transfers varies with the numberof children given the information in the household surveysFor these countries we assume that nonlabor income iscompletely exogenous
4 The Results
The results of the microsimulations regarding poverty arepresented inTable 2 Column (i) displays the observed changein the headcount ratio while the rest of the columns presentthe impact of the changes in fertility on poverty Column(v) shows the sum of all effects explored In order to betterunderstand the information in the table take the case ofBrazil as an example Column (i) shows that between 1990and 2012 the poverty headcount ratio in that country (usingthe line of US$ 4 a day) fell by around 25 points Thevalue in the last column has the following interpretationif fertility had been the only factor that changed duringthat period then the poverty headcount ratio in Brazilwould have fallen by 392 points This change is statisticallysignificant although not very big compared to the actual fallin poverty in that period The overall effect is the result ofa significant direct effect (minus374) and two smaller effects thatgo in different directions On the one hand the differentialreduction in fertility implied a greater increase in hours ofwork among more disadvantaged households which in turncontributed though very slightly to a further reduction inincome poverty On the other hand the falling patterns infertility among the poor ameliorated the poverty-decreasingimpact of the conditional cash transfer programs that aretargeted at families with children (mainly the Bolsa Familiain the 2000s) However this effect was quantitatively almostinsignificant
8 International Journal of Population Research
Table 2 Impact of fertility changes on poverty Changes in the poverty headcount ratio (US$ 4 line)
Country Period Observed change (i) EffectsDirect (ii) Hours (iii) NLI (iv) Total (v)
Argentina 1992ndash2012 minus1122 minus139 004 039 minus080(008) (002) (001) (001) (002)
Brazil 1990ndash2012 minus2516 minus374 minus017 014 minus392(003) (001) (000) (000) (001)
Chile 1990ndash2011 minus3324 minus156 018 mdash minus143(005) (001) (000) mdash (001)
El Salvador 1991ndash2010 minus1412 minus498 minus034 mdash minus531(006) (004) (001) mdash (005)
Mexico 1992ndash2012 minus350 minus435 minus030 025 minus443(010) (004) (002) (001) (004)
Peru 1997ndash2012 minus2320 minus383 minus019 mdash minus394(009) (003) (001) mdash (003)
Uruguay 1995ndash2012 minus156 minus008 000 002 minus005(003) (002) (000) (000) (002)
Source own calculations based on microdata from national household surveysNote All effects are significant at the 1 level The standard errors were calculated using bootstrap with 200 replications The values of each effect are averagesthat result from taking alternatively each year in the comparison as the base year The sample includes only households in which the head is between 25 and45 years old NLI nonlabor income
Table 3 Impact of fertility changes on inequality Changes in the Gini coefficient
Country Period Observed change (i) EffectsDirect (ii) Hours (iii) NLI (iv) Total (v)
Argentina 1992ndash2012 minus544 minus104 minus001 013 minus095(010) (001) (000) (000) (002)
Brazil 1990ndash2012 minus718 minus130 minus006 006 minus132(004) (000) (000) (000) (000)
Chile 1990ndash2011 minus178 minus075 004 mdash minus073(008) (001) (000) mdash (001)
El Salvador 1991ndash2010 minus768 minus130 minus007 mdash minus140(006) (002) (000) mdash (002)
Mexico 1992ndash2012 minus086 minus168 minus011 012 minus169(019) (002) (000) (000) (002)
Peru 1997ndash2012 minus732 minus192 minus006 mdash minus199(013) (001) (000) mdash (001)
Uruguay 1995ndash2012 minus203 minus012 minus001 002 minus013(004) (001) (000) (000) (001)
Source own calculations based on microdata from national household surveysNote All effects are significant at the 1 level The standard errors were calculated using bootstrap with 200 replications The values of each effect are averagesthat result from taking alternatively each year in the comparison as the base year The sample includes only households in which the head is between 25 and45 years old NLI nonlabor income
Changes in fertility patterns over the last two decades inLatin America have implied a reduction in income povertyThe impact is statistically significant and in some caseseconomically large For instance the estimated poverty-reduction effect was more than 4 points in Mexico andEl Salvador Most of the effect comes through the directeffect a reduction in fertility rates among the most dis-advantaged groups reduced family size and increased percapita income The hours-of-work effect is in most casespoverty reducing and the nonlabor-income effect is poverty
increasing but in both cases the estimated sizes aresmall
Table 3 shows the results of the simulations on incomeinequality measured by the Gini coefficient Inequality inBrazil as measured by the Gini coefficient fell by 718points between 1990 and 2012 If fertility had been the onlyfactor that changed in that period then the Gini coefficientwould have fallen by 132 points (column v) This change isstatistically significant representing around 18 of the actualreduction in inequality during that period The differential
International Journal of Population Research 9
reduction in fertility across socioeconomic groups in Brazilcontributed to the observed decline in inequality in the lasttwo decadesThe overall effect ofminus132 points in column (v) isthe result of a significant direct effect (minus130) and two smallereffects going in different directions
Some interesting general results emerge from Table 3First changes in fertility have implied a decline in incomeinequality The differential pattern in fertility across groupsexperienced in most Latin American countries over thelast decades translated into an equalizing impact on theincome distribution Second this effect is small althoughnot negligible To be sure demographic changes are not thecentral reason behind changes in income inequality but theyare statistically significant and economically relevant Onaverage fertility changes account for a fall of around onepoint in the Gini coefficient
Third most of the impact comes from the direct effectThe differential fall in fertility rates among socioeconomicgroups had a larger impact on the family size of poorerfamilies implying a proportionally larger increasing effect ontheir per capita incomes
Fourth the effect of fertility changes on hours of workand in turn on incomes is smaller and has different signsacross countries It is important to notice that a more intensereduction in fertility among the poor may be consistentwith a positive sign (inequality-increasing) for this effectThis could happen if the elasticity of hours of work withrespect to the number of children is higher among thenonpoor and if the extra hours worked triggered by the fallin fertility imply a household income increase proportionallylarger for the nonpoor Anyway with one exception thehours-of-work effect has a negative sign implying that theasymmetric reduction in fertility contributed to a reductionin income inequality by fostering higher participation in thelabor market among poor parents (mostly mothers)
Finally the effect through nonlabor incomes is alwaysinequality-increasing but small The fall in fertility amongthe poor implied a lower inequality-reducing impact of theconditional cash transfer programs that were introduced andexpanded in the region mainly during the 2000s The effecthowever is in most cases very small
5 Concluding Remarks
Poverty and inequality fell in Latin America over the lastdecades driven by a large number of factors In thispaper we highlight one factor that is usually overlooked inthe economic literature the impact of changes in fertilityThe average number of children fell in households fromall groups but especially among the more vulnerable afact that could have distributive implications We quantifythe consequences of this demographic pattern by applyingmicrosimulation techniquesThe results suggest that changesin fertility that took place over the last decades in LatinAmerica contributed to the reduction in income inequalityas well as in poverty The impacts found were in generalstatistically significant and economically relevant althoughnot very large Although surely there are more importantdeterminants behind the observed patterns in the income
distributions of the Latin American countries demographicfactors should not be overlooked
The paper has focused on three simple yet central chan-nels the change in the number of children (i) enlarges thefamily size and reduces per capita income (ii) affects thelabor supply decisions and (iii) modifies the income supportfrom social programs Certainly there are other potentiallyrelevant channels that are ignored in the paper For instancea reduction in the number of children per family couldimply higher inheritances and higher family spending oneducation and health per child and hence better incomeperspectives Also smaller families may be better positionedto take advantage of labor market opportunities for instancethrough migration Moreover lower fertility rates may beassociated with a reduction in the relative supply of unskilledlabor and hence an increase in its relative wage
Although in the short and medium run lower fertility inmore disadvantaged households may contribute to reducingpoverty and inequality in the long run the implicationsare more nuanced When children become adults they maycontribute to their parentsrsquo incomes a fall in fertility mayreduce incomes for the elderly especially in countries wherethe pension system is weak and ultimately contribute tohigher poverty through that intertemporal channel
Appendix
In this appendix we describe the issues regarding the simu-lations of the nonlabor incomes for each country
A Argentina Asignacioacuten Universal porHijo (AUH)
Argentina launched the conditional cash transfer programAUH in 2009 The beneficiaries are children under the ageof 18 in households whose members are unemployed orinformal workers The program provides a monthly cashsubsidy per child up to 5 children Since Argentinarsquos house-hold survey (EPH) does not include a question to identifyrecipients of this program we follow Garganta and Gasparini[13] assuming full take-up among those who qualify Thesimulation is made according to the following rule
If119867ℎ119905(1205781199051015840) + 119876
ℎ119905le 5 and119867
ℎ119905+ 119876ℎ119905
le 5
119884119873119871
ℎ119905(1205781199051015840) = 119884
119873119871
ℎ119905+ 120579119905(119867ℎ119905(1205781199051015840) minus 119867
ℎ119905) (A1)
If119867ℎ119905(1205781199051015840) + 119876
ℎ119905gt 5 and119867
ℎ119905+ 119876ℎ119905
le 5
119884119873119871
ℎ119905(1205781199051015840) = 119884
119873119871
ℎ119905+ 120579119905(5 minus 119867
ℎ119905minus 119876ℎ119905) (A2)
If119867ℎ119905(1205781199051015840) + 119876
ℎ119905le 5 and119867
ℎ119905+ 119876ℎ119905
gt 5
119884119873119871
ℎ119905(1205781199051015840) = 119884
119873119871
ℎ119905+ 120579119905(119867ℎ119905(1205781199051015840) + 119876
ℎ119905minus 5) (A3)
If119867ℎ119905(1205781199051015840) + 119876
ℎ119905gt 5 and119867
ℎ119905+ 119876ℎ119905
gt 5
119884119873119871
ℎ119905(1205781199051015840) = 119884
119873119871
ℎ119905 (A4)
10 International Journal of Population Research
where 119876ℎ119905
is the number of children aged 17 or 18(recall that we carry out the simulations for childrenaged 16 or younger) and 120579
119905is equal to AR$270 for
the third trimester of 2012 and AR$340 for the fourthtrimester
B Brazil Bolsa Familia Program (BFP)
The Bolsa Familia implemented in 2003 is the flagshipconditional cash transfer program in Brazil aimed at poorhouseholds The estimated nonlabor incomes in this case arecalculated as follows
If119867119905ℎ
gt 0 and beneficiary
119884119873119871
ℎ119905(1205781199051015840) = 119884
119873119871
119905ℎ+ 119872119905ℎ(119867119905ℎ(1205781199051015840)
119867119905ℎ
minus 1) (B1)
If119867119905ℎ
= 0119867119905ℎ(1205781199051015840) = 119896 and beneficiary
119884119873119871
ℎ119905(1205781199051015840) = 119884
119873119871
119905ℎ+ 119872119896
119905ℎminus 119897119905ℎ (B2)
with 119896 = 1 2 3 119872119905ℎ
being the amount of thetransfer119872119896
119905ℎbeing the average benefit of households
with 119896 children and 119897119905ℎbeing the standard benefit of
recipients without children
C Mexico Oportunidades
Oportunidades is the main antipoverty program of theMexican Government The size of the transfer depends onthe number of children among other characteristics of thehouseholdMexicorsquos ENIGH identifies the beneficiaries of theprogram and the amount of the subsidy that each householdreceives The simulation is made as follows
119884119873119871
ℎ119905(1205781199051015840) = 119884
119873119871
119905ℎ+ 119872119896
119905ℎminus 119872119905ℎ
if 119867119905ℎ(1205781199051015840) = 119896 119867
119905ℎ= 119896 and beneficiary
(C1)
where 119872119896119905ℎis the average transfer of a beneficiary household
with 119896 children and119872119905ℎis the original household transfer
D Uruguay Asignaciones Familiares
The Asignaciones Familiares program in Uruguay is targetedat the children of workers The amount of the transfer varieswith the householdrsquos income Since the size of the transferdepends on characteristics of the child such as the age thesimulation is carried out as follows
If119867119905ℎ
gt 0 and beneficiary
119884119873119871
ℎ119905(1205781199051015840) = 119884
119873119871
119905ℎ+ 119872119905ℎ(119867119905ℎ(1205781199051015840)
119867119905ℎ
minus 1) (D1)
If119867119905ℎ
= 0119867119905ℎ(1205781199051015840) = 119896 and beneficiary
119884119873119871
ℎ119905(1205781199051015840) = 119884
119873119871
119905ℎ+ 119872119896
119905ℎ (D2)
with 119896 = 1 2 3 119872119905ℎ
being the total amount ofthe transfer and 119872
119896
119905ℎbeing the average transfer of a
household with 119896 children
Disclosure
The information presented herein is solely the responsibilityof the authors and does not necessarily reflect the opinions ofthe aforementioned entities
Competing Interests
The authors declare that they have no competing interests
Acknowledgments
This paper was written in the framework of the projectEnhancing Womenrsquos Economic Empowerment throughBetter Policies in Latin America a joint initiative betweenCEDLAS-UNLP and CIEDUR with the support ofIDRC Canada The paper includes material from NicolasBadaraccorsquos Masters dissertation at Universidad Nacionalde La Plata The authors are grateful to Carolina RobinoGuillermo Cruces Jorge Paz Facundo Albornoz andseminar participants at AAEP (Posadas 2014) Network ofInequality and Poverty (UNGS 2014) and Conference onSocial Protection (IDRC and IPC Brasilia 2014) for valuablecomments and suggestions
References
[1] T W Guinnane ldquoThe historical fertility transition a guide foreconomistsrdquo Journal of Economic Literature vol 49 no 3 pp589ndash614 2011
[2] L E Jones and M Tertilt ldquoAn economic history of fertility inthe US 1826ndash1960rdquo NBERWorking Paper 12796 2006
[3] J Chackiel La Dinamica Demografica en America LatinaSerie Poblacion y Desarrollo No 52 Centro Latinoamericanoy Caribeno de Demografıa-Division de Poblacion de CEPAL2004
[4] ECLAC ldquoLa fecundidad en America Latina un descensoacelerado y heterogeneo con profundas transformacionesdemograficas y socialesrdquo in Fertility Latin American and theCaribbean Demographic Observatory no 5 Centro Latinoa-mericano y Caribeno de DemografıamdashDivision de Poblacionde CEPAL 2008
[5] ECLAC ldquoChapter II current situation and outlook for fertilityin Latin Americardquo in Social Panorama of Latin America 2011ECLAC 2011
[6] L Gasparini and M Marchionni Bridging Gender Gaps TheRise and Deceleration of Female Labor Force ParticipationCEDLAS-UNLP 2015
[7] L Gasparini and N Lustig ldquoThe rise and fall of incomeinequality in Latin Americardquo inThe Oxford Handbook of LatinAmerican Economics J Ocampo and J Ros Eds chapter 27 pp691ndash714 Oxford University Press Oxford UK 2011
[8] G Cornia Ed Falling Inequality in Latin America PolicyChanges and Lessons Oxford University Press 2014
International Journal of Population Research 11
[9] F Alvaredo and L Gasparini ldquoRecent trends in inequalityand poverty in developing countriesrdquo in Handbook of IncomeDistribution Atkinson and Bourguignon Eds vol 2 chapter 9pp 697ndash805 Elsevier New York NY USA 2015
[10] M Marchionni and L Gasparini ldquoTracing out the effects ofdemographic changes on the income distribution The caseof Greater Buenos Aires 1980ndash1998rdquo Journal of EconomicInequality vol 5 no 1 pp 97ndash114 2007
[11] J M Wooldridge Econometric Analysis of Cross Section andPanel Data MIT Press Boston Mass USA 2002
[12] L Gasparini M Marchionni and W Sosa Escudero ldquoChar-acterization of inequality changes through microeconometricdecompositions The case of Greater Buenos Airesrdquo in TheMicroeconomics of Income Distribution Dynamics in East Asiaand Latina America F Bourguignon F Ferreira and N LustigEds Oxford University Press New York NY USA 2004
[13] S Garganta and L Gasparini ldquoThe impact of a social programon labor informality the case of AUH in Argentinardquo Journal ofDevelopment Economics vol 115 pp 99ndash110 2015
Submit your manuscripts athttpwwwhindawicom
Child Development Research
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Education Research International
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Biomedical EducationJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Psychiatry Journal
ArchaeologyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
AnthropologyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Research and TreatmentSchizophrenia
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Urban Studies Research
Population ResearchInternational Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CriminologyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Aging ResearchJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
NursingResearch and Practice
Current Gerontologyamp Geriatrics Research
Hindawi Publishing Corporationhttpwwwhindawicom
Volume 2014
Sleep DisordersHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
AddictionJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Depression Research and TreatmentHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Geography Journal
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Research and TreatmentAutism
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Economics Research International
2 International Journal of Population Research
bottom strata of the distribution may induce some low-income women to enter the labor market or to work morehours In that case the decline in poverty and inequalitymight be larger than expected if one only considers the directimpact of the decline in family size In addition a reductionin fertility may reduce the income transfers received by poorhouseholds in terms of social assistance directed to familieswith children
In this paper we find that the changes in fertility thattook place over the last decades in Latin America contributedto the reduction in income inequality as well as in povertyThe fall in fertility among the income-deprived helped toreduce the incidence of income poverty Besides since it waslarger than the fall among the nonpoor it contributed to thereduction in income inequality as wellThemain channel wasstraightforward lower fertility rates implied smaller familiesand hence larger per capita incomes Lower fertility alsofostered labor force participation especially among womenwhich contributed to the reduction of poverty and inequalityin most countries although the size of this effect was small
The rest of the paper is organized in a straightforwardwayIn the next section we document changes in fertility acrossincome groups and present trends in income inequality andpoverty in Latin America In Section 3 we lay out themethodology discuss its limitations and introduce the dataused in the estimations Section 4 presents the main resultsof the paper whereas Section 5 closes with some concludingremarks
2 Fertility Changes and Distributive Trends
In the early 1960s the total fertility rate (TFR) was around6 children per woman in Latin America higher than theworldrsquos average (total fertility rate is the average numberof children a hypothetical cohort of women would haveat the end of their reproductive period if during theirwhole lives they were subject to the fertility rates of agiven period and if they were not subject to mortalityit is expressed as children per woman (United NationsWorld Population Prospects The 2012 Revision definitionavailable at httpesaunorgwppExcel-Datafertilityhtm))It was not until the mid-1960s that TFR began its downwardtrend which eventually led to clear convergence towards thelevels of the most advanced regions of the world (Figure 1) Inparticular theTFR continued falling over the last twodecadesin Latin America reaching an average of 25 children perwoman in 2005ndash2010 (this decreasing trend has been presentin developed countries since the beginning of the past century[1 2] For an analysis of fertility dynamics in Latin Americasee Chackiel [3] and ECLAC [4 5]
Following a similar trend the number of children perhousehold (the proxy for fertility that can be implementedwith household survey microdata) has been falling over thelast two decades In 1992 there were 23 children under theage of 16 per household (for households with a woman aged25ndash45) in a typical Latin American country the figure was 18in 2012 [6]
The average number of children decreased in householdsfrom all population groups but the gap between the most
0010203040506070
1950
ndash195
5
1955
ndash196
0
1960
ndash196
5
1965
ndash197
0
1970
ndash197
5
1975
ndash198
0
1980
ndash198
5
1985
ndash199
0
1990
ndash199
5
1995
ndash200
0
2000
ndash200
5
2005
ndash201
0
World
Latin AmericaThe Caribbean
North America
Europe and Central AsiaEast Asia and the Pacific
South Asia
Middle East and North Africa
Sub-Saharan Africa
Figure 1 Total fertility rate (children per woman) Regions of theworld 1950ndash2010 Source Gasparini and Marchionni [6] based onWorld Population Prospects The 2012 Revision DVD Edition
10121416182022242628
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
Richest 20Poorest 20
Figure 2 Number of children under 16 per household Bottom andtop quintiles of parental income distribution 1992ndash2012 Sourceown calculations based on microdata from national household sur-veys Note average of seven Latin American countries ArgentinaBrazil Chile El Salvador Mexico Peru and Uruguay Householdswith head aged 25ndash45 Unweighted means
and the least vulnerable groups shrank owing to a sharperdecline in the number of children living in poor householdsFigure 2 takes a sample of seven Latin American countries toshow the average number of children under 16 per householdin the poorest 20 and richest 20 of the prime-age parents(the sample includes countries in which we implement themicrosimulations that follow in the next section) In bothgroups fertility went down over the two decades the fall issomewhat more pronounced among the poorest couplesThegap in the number of children per household shrank from067 in 1992 to 047 in 2012
With regard to the income distribution Latin Americaexperienced an increase in inequality during the 1990s whilepoverty slightly decreased [7ndash9] In contrast in the 2000sthere were strong reductions of both income inequality and
International Journal of Population Research 3
15
20
25
30
35
40
4519
9219
9319
9419
9519
9619
9719
9819
9920
0020
0120
0220
0320
0420
0520
0620
0720
0820
0920
1020
1120
12
Headcount ratio (US$ 4 line)
45
47
49
51
53
55
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
Gini coefficient
Figure 3 Poverty and income inequality Latin America Source own calculations based on microdata from national household surveysNote average of seven Latin American countries Argentina Brazil Chile El Salvador Mexico Peru and Uruguay Unweighted meansPoverty and inequality computed over the distribution of household per capita income
poverty Figure 3 shows the average income inequality andpoverty trends of seven Latin American countries Povertyis measured by the headcount ratio with a daily US$ 4 linewhile income inequality is measured by the Gini coefficientBoth indices are calculated over the household per capitaincome distribution Income poverty slightly decreased dur-ing the 1990s and fell sharply during the 2000s ArgentinaUruguay and Mexico suffered an increase in poverty duringthe 1990s and then a continuous fall over the following decade(Figure 4) Meanwhile Brazil Chile El Salvador and Peruexperienced a decrease in poverty over the entire periodthough steeper during the 2000s
On average for our sample of seven countries incomeinequality increased during the first decade under analysisand then significantly decreased during the 2000s The trendfor the 1990s is driven by the sharp increase in incomedispersion inArgentina andUruguay (Figure 5) and the smallchanges in the rest of the countries In contrast during the2000s the fall in inequality was strong and generalized to alleconomies
3 Data and Methodology
We focus on seven Latin American countries ArgentinaBrazil Chile El Salvador Mexico Peru and Uruguay duringthe 1990s and 2000sThe analysis is based onmicrodata fromnational household surveys collected by the correspondingNational Statistical Offices Table 1 presents a brief descrip-tion of these databases Since surveys are not homogeneouswe have made all possible efforts to make variables compara-ble across countries and over time by using similar definitionsof variables in each countryyear and by applying consistentmethods of processing the data
The ideal variable for capturing fertility would be thenumber of children either living in the household or notUnfortunately this variable is usually not available in LatinAmerican household surveys with the exception of BrazilAs a proxy we use the number of children living in thehousehold Consequently two main problems emerge First
Table 1 Household surveys
Country Survey Acronym Years
Argentina Encuesta Permanente deHogares EPH 1992ndash2012
Brazil Pesquisa Nacional porAmostra de Domicilios PNAD 1990ndash2012
ChileEncuesta de
CaracterizacionSocioeconomica Nacional
CASEN 1990ndash2011
El Salvador Encuesta de Hogares dePropositos Multiples EHPM 1991ndash2010
MexicoEncuesta Nacional deIngresos y Gasto de los
HogaresENIGH 1992ndash2012
Peru Encuesta Nacional deHogares ENAHO 1997ndash2012
Uruguay Encuesta Continua deHogares ECH 1995ndash2012
the measurement error of this proxy depends on the ageof parents and children that is on average the older theparents and the older the children are the more likely theyleave the parental home Therefore the observed differencesin the number of children across households are not onlydue to reproductive decisions but also due to the life cycleTo deal with this shortcoming we restrict our sample tohouseholds where the head of household is between 25 and45 years old Many people below 25 years old may not haveyet started childbearing The average age at childbearing forthe countries under analysis was 275 years in 1993 and 276years in 2011 The dispersion across countries is also smallranging from 265 in Mexico to 283 in Peru The average ageat marriage in the early 1990s was 229 and 256 years forwomen and men respectively and about two years more inthe early 2010s (250 and 275)
Another problem arises since Latin American house-hold surveys usually register family relationships among
4 International Journal of Population Research
0102030405060
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
0102030405060
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
0102030405060
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
0102030405060
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
0102030405060
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
0102030405060
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
0102030405060
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
Argentina Brazil Chile
El Salvador Mexico Peru
Uruguay
Figure 4 PovertyHeadcount ratio (US$ 4 line) Source own calculations based onmicrodata fromnational household surveys Note povertycomputed over the distribution of household per capita income
household members only in terms of the head of householdthat is it is possible to match children with their parentsonly when they are the head of household or the spouseTherefore we further restrict the analysis to only considerfertility decisions of the head of household and the spouse
Regarding the empirical specification and estimationmethodology we follow Marchionni and Gasparini [10]The main inputs to carry out the microsimulations arethe estimates of the parameters that govern fertility deci-sionsoutcomes (for simplicity we refer to fertility decisionsalthough fertility outcomes could be the result of free con-scious choices but also the consequence of various other cir-cumstances) and the response of labor market participationto changes in family size We assume that the number ofchildren in a household follows a Poisson process and thatits parameters can be consistently estimated using a Poissonregression model Hourly wages and hours of work areassumed to be simultaneously determined in an equilibriummodel of the labor market
After estimating the parameters we carry out the sim-ulations That is we simulate the counterfactual incomedistribution that arises in a given base year by assumingthat the population in that year takes fertility decisions
according to the parameters estimated for a different yearTheresulting poverty and inequality measures over the simulateddistribution are compared to those actually observed in thebase year The difference between the simulated value ofan indicator of poverty or inequality and its actual value isinterpreted as a measure of the direct impact of the change infertility behavior that is the impact that would take place incase all other covariates remained fixed
The methodology requires keeping all other things con-stant when simulating the impact of changes in fertilitydecisions Naturally as it was stressed in previous sectionssome of these factors may be correlated with fertility Forinstance the structure ofwagesmay respond to changes in thelabor supply triggered by a change in fertility Additionallychanges in reproductive behavior may have been inducedby income changes in which case the microsimulation onlycaptures a single round of effects from fertility to incomes ofa more complicated process
In a given period income poverty and inequality areaffected by a host of factors including economic shocks andsocial policies The microsimulation methodology is usefulas it allows approximating the size of the impact that agiven factor would have in case all other factors remained
International Journal of Population Research 5
40
45
50
55
60
6519
9219
9419
9619
9820
0020
0220
0420
0620
0820
1020
12
40
45
50
55
60
65
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
40
45
50
55
60
65
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
40
45
50
55
60
65
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
40
45
50
55
60
65
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
40
45
50
55
60
65
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
40
45
50
55
60
65
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
Argentina Brazil Chile
El Salvador Mexico Peru
Uruguay
Figure 5 Income inequality Gini coefficient Source own calculations based onmicrodata fromnational household surveys Note inequalitycomputed over the distribution of household per capita income
fixed Although this is clearly an unrealistic scenario theestimations are still useful as a first approximation of theimpact especially considering the fact that the more ambi-tious alternative of estimating a general equilibriummodel isusually unfeasible
31 Microsimulations We carry out the simulation over thedistribution of household per capita income defined as
119910119894119905=
119884119871
ℎ119905+ 119884119873119871
ℎ119905
119873ℎ119905
forall119894 isin ℎ at time 119905 (1)
where 119894 indexes individuals ℎ households and 119905 time periods(years) 119884119871
ℎ119905denotes total labor income of household ℎ at time
119905 119884119873119871
ℎ119905labels nonlabor income and 119873
ℎ119905is the family size
which is the sum of the number of children up to 16 yearsold119867
ℎ119905and the rest of the household members 119877
ℎ119905
119873ℎ119905
= 119867ℎ119905+ 119877ℎ119905 (2)
Weneed to specify how119867ℎ119905is determined in order to simulate
the counterfactual distributions The literature argues that
fertility outcomes are the result of a process affected bycharacteristics of each spouse and of the household amongother factors This process can be formalized as
119867ℎ119905
= 119867 (119885ℎ119905 119890ℎ119905 120578119905) (3)
where 119885ℎ119905
is a vector of observable characteristics and119890ℎ119905
includes all unobservable characteristics that influencethe familyrsquos reproductive behavior while 120578
119905is the set of
parameters that govern fertility decisionsWe assume that labor income for individual 119894 is given by
119884119871
119894119905= 119908119894119905119871119894119905 (4)
where 119908119894119905
is the hourly wage rate and 119871119894119905
is the num-ber of hours worked by individual 119894 Wages and hoursof work are defined as 119908
119894119905= 119908(119883
1119894119905 120576119882
119894119905 120573119905) and 119871
119894119905=
119871(1198832119894119905 119867119894119905 120576119871
119894119905 120574119905 120582119905) respectively Where 119883
1119894119905and 119883
2119894119905are
observable characteristics 120576119882119894119905
and 120576119871
119894119905are unobservable char-
acteristics while 120573119905 120574119905 and 120582
119905are parameters In particular
120582119905is the parameter that relates the hours of work to the
number of children
6 International Journal of Population Research
Nonlabor income is usually assumed to be exogenous inthemicrosimulation literature However as stated in previoussections the expansion of social programs to families withchildren suggests a potential link between fertility decisionsand nonlabor income Accordingly we assume that nonlaborincome is given by
119884119873119871
ℎ119905= 119898ℎ119905+ 119892 (119867
ℎ119905) (5)
where119898ℎ119905is exogenous and119892(119867
ℎ119905) represents the component
of nonlabor income that depends on the number of childrenWe label the parameters of the fertility decisions at time 119905
as 120578119905and those at time 1199051015840 as 120578
1199051015840 A key step in themethodology
is to estimate the counterfactual number of children in agiven year 119905 if fertility outcomes were determined as in analternative year 119905
1015840 Once this term labeled as 119867ℎ119905(1205781199051015840) is
estimated three microsimulation exercises are carried out byreplacing this estimate in the household per capita incomeequation
The first exercise aims to capture the contribution of thechange in fertility parameters 120578 to the actual change in theincome distribution through the denominator of (1) Thesimulated per capita income is given by
119910119863
119894119905=
119884119871
ℎ119905+ 119884119873119871
ℎ119905
119873ℎ119905(1205781199051015840)
(6)
where 119873ℎ119905(1205781199051015840) = 119877
ℎ119905+ 119867ℎ119905(1205781199051015840) Superscript 119863 in the
simulated per capita income stands for the direct-size effectthat is the change in the income distribution due to changesin the number of household members among whom totalhousehold income should be distributed
The second exercise involves simulating the labor incomethat a household would have in year 119905 if parameters thatgovern fertility decisionswere those of year 1199051015840 In other wordsthe exercise consists in replacing the simulated number ofchildren 119867
ℎ119905(1205781199051015840) in the equation of individual labor income
(4) to obtain
119910119867
119894119905=
119884119871
ℎ119905(1205781199051015840) + 119884
119873119871
ℎ119905
119873ℎ119905
(7)
Superscript 119867 in the simulated per capita income stands forthe hours-size effect that is the contribution of the changein fertility parameters 120578 to the actual change in the incomedistribution through the indirect channel of affecting thehours-of-work decisions
We carry out a third exercise by simulating the coun-terfactual distribution arising from a change in nonlaborincome driven by changes in fertility decisions We refer tothe distributional impact of changes through this channelas the nonlabor-income effect and thus label the per capitaincome variable in (8) with the NL superscript
119910119873119871
119894119905=
119884119871
ℎ119905+ 119884119873119871
ℎ119905(1205781199051015840)
119873ℎ119905
(8)
Finally (9) considers the three channels simultaneously toobtain the total effect (superscript119879) of changes in the fertilityparameters
119910119879
119894119905=
119884119871
ℎ119905(1205781199051015840) + 119884
119873119871
ℎ119905(1205781199051015840)
119873ℎ119905(1205781199051015840)
(9)
We denote the per capita income distribution among indi-viduals in year 119905 as 119863
119905(119910119905) = (119910
1119905 1199102119905 119910
119899119905) where 119899
denotes the total number of individuals in the populationLet 119868(sdot) be a distributive indicator as the Gini coefficientor the poverty headcount ratio Then the contribution ofthe change in fertility parameters 120578 to the actual change inthe income distribution through the direct channel is givenby (10) The contributions through the other channels arecomputed analogously
119868 (119863119905(119910119905)) minus 119868 (119863
119905(119910119863
119905)) (10)
A shortcoming of the microsimulation methodology is thatdecomposition is path-dependent This means that takingyear 119905 or year 1199051015840 as the base year does not necessarily implythe same result Accordingly we perform the simulationsalternating the base year and report the average effect for eachexercise
In the following subsections we provide details concern-ing the specification and estimation of fertility labor incomeand nonlabor-income equations
32 Fertility Decisions We assume that the number of chil-dren in (3) follows a Poisson process with parameter 120583
ℎ119905
Formally
119867ℎ119905
sim Poisson (120583ℎ119905) with 120583
ℎ119905= 119864 (119867
ℎ119905| 119885ℎ119905) = exp (119885
1015840
ℎ119905120578119905) (11)
Then
Prob (119867ℎ119905
= 1198670) =
exp (minus120583ℎ119905) (120583ℎ119905)1198670
1198670
with 1198670= 0 1 2 (12)
Equation (12) is the Poisson regressionmodel fromwhich it ispossible to consistently estimate parameters 120578
119905by maximum
likelihood (it can be shown that consistency holds for themaximum likelihood estimators of 120578
119905as long as the real
distribution is any of the linear exponential family to whichthe Poisson distribution belongs provided that the condi-tional mean in (11) is correctly specified see eg Wooldridge[11]) We estimate separate models for two-parent householdand single-parent households For simplicity we denote theestimates by 120578
Once parameters 120578 are estimated we proceed to simulatechanges in the number of children Since the objective isto simulate these changes as a consequence of changes onlyin the parameters 120578 it is necessary to keep unobservablefactors fixed Therefore we characterize each household bythe quantile it occupies in the distribution of children ofyear 119905 Let 119865
120578119905|119885ℎ119905
(sdot) be the cumulative probability function ofa random variable that follows a Poisson distribution withparameter equal to exp(1198851015840
ℎ119905120578119905) and let 119902
ℎ119905be the quantile
for household ℎ at time 119905 that is 119865120578119905|119885ℎ119905
(119867ℎ119905) = 119902
ℎ119905 The
International Journal of Population Research 7
simulated number of children in household ℎ will be the onethat places it in the 119902
ℎ119905quantile of the distribution of children
with the relevant parameters of time 1199051015840(1205781199051015840) conditional to
the observable characteristics 119885ℎ119905 Formally the simulated
number of children in household ℎ at year 119905 using theestimated fertility parameters for year 1199051015840 is given by
119867ℎ119905(1205781199051015840) = 119865
minus1
1205781199051015840 |119885ℎ119905
∘ 119865120578119905|119885ℎ119905
(119867ℎ119905) (13)
33 Labor Incomes To estimate individual labor income in(4) we follow Gasparini et al [12] in assuming that bothwages and hours worked are determined in a reduced-formmodel of the labor market equilibrium
ln119908lowast
119894119905= 1198831015840
1119894119905120573119905+ 120576119882
119894119905 (14)
119871lowast
119894119905= 1198831015840
2119894119905120574119905+ 119867119894119905120582119905+ 120576119871
119894119905 (15)
with
if 119871lowast119894119905gt 0
119908119894119905= 119908lowast
119894119905
119871119894119905= 119871lowast
119894119905
if 119871lowast119894119905le 0
119908119894119905= 0
119871119894119905= 0
(120576119882
119894119905 120576119871
119894119905) sim 119873(0 0 120590
2
119882119905 1205902
119871119905 120588119905)
(16)
where119908lowast119894119905and119871lowast
119894119905are unobserved latent variables Vectors119883
1119894119905
and 1198832119894119905
include observable factors affecting hourly wagesand hours of work respectively while 119867
119894119905stands for the
number of children 120573119905 120574119905(vectors) and 120582
119905are parameters
to be estimated along with 1205902
119882119905 1205902119871119905 and 120588
119905
We estimate (14) by Heckmanrsquos maximum likelihoodmethod where a censored version of (15) is used as a selectionequation replacing hours of work by a binary variable thatindicates whether the individual works For estimation of(15) we use aTobitmodel (although this estimation strategy isnot fully efficient efficiency loss is not necessarily significantfor a given sample size and this alternative has certaincomputational advantages over a full information procedurefor more details see Gasparini et al [12]) We estimateseparate models for head of households and spouses
To simulate changes in individual labor incomes wereplace the observed number of children of individual 119894 atyear 119905 in (15) by the simulated number of children 119867
119905ℎ(1205781199051015840)
while assigning himher a wage based on (14) and an errorterm drawn from the bivariate distribution implicit in themodel
34 Nonlabor Incomes The measure of nonlabor incomeincludes pensions capital income and transfers Conditionalcash transfers are included as government transfers Wesimulate nonlabor incomes in an ad hoc manner taking
into account the characteristics of each countryrsquos socialprogram In those cases where it is not possible to identifythe recipients in the household survey we identify potentialbeneficiaries and estimate the amount of the transfers basedon the design of each program The general framework is asfollows nonlabor income of household ℎ at year 119905 is given by(5) where 119892(119867
ℎ119905) represents the part of nonlabor income that
depends on the number of children Typically cash transfersin poverty-alleviation programs vary with the number ofchildren in the family For example if the program has a fixedamount 120579
119905per child the nonlabor income becomes
119884119873119871
ℎ119905= 119898ℎ119905+ 120579119905119867ℎ119905 (17)
The simulated nonlabor income is calculated using thecounterfactual number of children in a given year 119905 if fertilityoutcomes were determined as in an alternative year 1199051015840
119884119873119871
ℎ119905(1205781199051015840) = 119898
ℎ119905+ 120579ℎ119905119867ℎ119905(1205781199051015840) (18)
Even though all programs have a similar setting to the onedescribed they differ in some aspects maximum number ofchildren age and in some cases how to identify the programrsquosrecipients The details are given in the Appendix We sim-ulate nonlabor incomes for Argentina Brazil Mexico andUruguay Even though the programs in Chile El Salvadorand Peru have similar characteristics it is not possible toestimate how the amount of transfers varies with the numberof children given the information in the household surveysFor these countries we assume that nonlabor income iscompletely exogenous
4 The Results
The results of the microsimulations regarding poverty arepresented inTable 2 Column (i) displays the observed changein the headcount ratio while the rest of the columns presentthe impact of the changes in fertility on poverty Column(v) shows the sum of all effects explored In order to betterunderstand the information in the table take the case ofBrazil as an example Column (i) shows that between 1990and 2012 the poverty headcount ratio in that country (usingthe line of US$ 4 a day) fell by around 25 points Thevalue in the last column has the following interpretationif fertility had been the only factor that changed duringthat period then the poverty headcount ratio in Brazilwould have fallen by 392 points This change is statisticallysignificant although not very big compared to the actual fallin poverty in that period The overall effect is the result ofa significant direct effect (minus374) and two smaller effects thatgo in different directions On the one hand the differentialreduction in fertility implied a greater increase in hours ofwork among more disadvantaged households which in turncontributed though very slightly to a further reduction inincome poverty On the other hand the falling patterns infertility among the poor ameliorated the poverty-decreasingimpact of the conditional cash transfer programs that aretargeted at families with children (mainly the Bolsa Familiain the 2000s) However this effect was quantitatively almostinsignificant
8 International Journal of Population Research
Table 2 Impact of fertility changes on poverty Changes in the poverty headcount ratio (US$ 4 line)
Country Period Observed change (i) EffectsDirect (ii) Hours (iii) NLI (iv) Total (v)
Argentina 1992ndash2012 minus1122 minus139 004 039 minus080(008) (002) (001) (001) (002)
Brazil 1990ndash2012 minus2516 minus374 minus017 014 minus392(003) (001) (000) (000) (001)
Chile 1990ndash2011 minus3324 minus156 018 mdash minus143(005) (001) (000) mdash (001)
El Salvador 1991ndash2010 minus1412 minus498 minus034 mdash minus531(006) (004) (001) mdash (005)
Mexico 1992ndash2012 minus350 minus435 minus030 025 minus443(010) (004) (002) (001) (004)
Peru 1997ndash2012 minus2320 minus383 minus019 mdash minus394(009) (003) (001) mdash (003)
Uruguay 1995ndash2012 minus156 minus008 000 002 minus005(003) (002) (000) (000) (002)
Source own calculations based on microdata from national household surveysNote All effects are significant at the 1 level The standard errors were calculated using bootstrap with 200 replications The values of each effect are averagesthat result from taking alternatively each year in the comparison as the base year The sample includes only households in which the head is between 25 and45 years old NLI nonlabor income
Table 3 Impact of fertility changes on inequality Changes in the Gini coefficient
Country Period Observed change (i) EffectsDirect (ii) Hours (iii) NLI (iv) Total (v)
Argentina 1992ndash2012 minus544 minus104 minus001 013 minus095(010) (001) (000) (000) (002)
Brazil 1990ndash2012 minus718 minus130 minus006 006 minus132(004) (000) (000) (000) (000)
Chile 1990ndash2011 minus178 minus075 004 mdash minus073(008) (001) (000) mdash (001)
El Salvador 1991ndash2010 minus768 minus130 minus007 mdash minus140(006) (002) (000) mdash (002)
Mexico 1992ndash2012 minus086 minus168 minus011 012 minus169(019) (002) (000) (000) (002)
Peru 1997ndash2012 minus732 minus192 minus006 mdash minus199(013) (001) (000) mdash (001)
Uruguay 1995ndash2012 minus203 minus012 minus001 002 minus013(004) (001) (000) (000) (001)
Source own calculations based on microdata from national household surveysNote All effects are significant at the 1 level The standard errors were calculated using bootstrap with 200 replications The values of each effect are averagesthat result from taking alternatively each year in the comparison as the base year The sample includes only households in which the head is between 25 and45 years old NLI nonlabor income
Changes in fertility patterns over the last two decades inLatin America have implied a reduction in income povertyThe impact is statistically significant and in some caseseconomically large For instance the estimated poverty-reduction effect was more than 4 points in Mexico andEl Salvador Most of the effect comes through the directeffect a reduction in fertility rates among the most dis-advantaged groups reduced family size and increased percapita income The hours-of-work effect is in most casespoverty reducing and the nonlabor-income effect is poverty
increasing but in both cases the estimated sizes aresmall
Table 3 shows the results of the simulations on incomeinequality measured by the Gini coefficient Inequality inBrazil as measured by the Gini coefficient fell by 718points between 1990 and 2012 If fertility had been the onlyfactor that changed in that period then the Gini coefficientwould have fallen by 132 points (column v) This change isstatistically significant representing around 18 of the actualreduction in inequality during that period The differential
International Journal of Population Research 9
reduction in fertility across socioeconomic groups in Brazilcontributed to the observed decline in inequality in the lasttwo decadesThe overall effect ofminus132 points in column (v) isthe result of a significant direct effect (minus130) and two smallereffects going in different directions
Some interesting general results emerge from Table 3First changes in fertility have implied a decline in incomeinequality The differential pattern in fertility across groupsexperienced in most Latin American countries over thelast decades translated into an equalizing impact on theincome distribution Second this effect is small althoughnot negligible To be sure demographic changes are not thecentral reason behind changes in income inequality but theyare statistically significant and economically relevant Onaverage fertility changes account for a fall of around onepoint in the Gini coefficient
Third most of the impact comes from the direct effectThe differential fall in fertility rates among socioeconomicgroups had a larger impact on the family size of poorerfamilies implying a proportionally larger increasing effect ontheir per capita incomes
Fourth the effect of fertility changes on hours of workand in turn on incomes is smaller and has different signsacross countries It is important to notice that a more intensereduction in fertility among the poor may be consistentwith a positive sign (inequality-increasing) for this effectThis could happen if the elasticity of hours of work withrespect to the number of children is higher among thenonpoor and if the extra hours worked triggered by the fallin fertility imply a household income increase proportionallylarger for the nonpoor Anyway with one exception thehours-of-work effect has a negative sign implying that theasymmetric reduction in fertility contributed to a reductionin income inequality by fostering higher participation in thelabor market among poor parents (mostly mothers)
Finally the effect through nonlabor incomes is alwaysinequality-increasing but small The fall in fertility amongthe poor implied a lower inequality-reducing impact of theconditional cash transfer programs that were introduced andexpanded in the region mainly during the 2000s The effecthowever is in most cases very small
5 Concluding Remarks
Poverty and inequality fell in Latin America over the lastdecades driven by a large number of factors In thispaper we highlight one factor that is usually overlooked inthe economic literature the impact of changes in fertilityThe average number of children fell in households fromall groups but especially among the more vulnerable afact that could have distributive implications We quantifythe consequences of this demographic pattern by applyingmicrosimulation techniquesThe results suggest that changesin fertility that took place over the last decades in LatinAmerica contributed to the reduction in income inequalityas well as in poverty The impacts found were in generalstatistically significant and economically relevant althoughnot very large Although surely there are more importantdeterminants behind the observed patterns in the income
distributions of the Latin American countries demographicfactors should not be overlooked
The paper has focused on three simple yet central chan-nels the change in the number of children (i) enlarges thefamily size and reduces per capita income (ii) affects thelabor supply decisions and (iii) modifies the income supportfrom social programs Certainly there are other potentiallyrelevant channels that are ignored in the paper For instancea reduction in the number of children per family couldimply higher inheritances and higher family spending oneducation and health per child and hence better incomeperspectives Also smaller families may be better positionedto take advantage of labor market opportunities for instancethrough migration Moreover lower fertility rates may beassociated with a reduction in the relative supply of unskilledlabor and hence an increase in its relative wage
Although in the short and medium run lower fertility inmore disadvantaged households may contribute to reducingpoverty and inequality in the long run the implicationsare more nuanced When children become adults they maycontribute to their parentsrsquo incomes a fall in fertility mayreduce incomes for the elderly especially in countries wherethe pension system is weak and ultimately contribute tohigher poverty through that intertemporal channel
Appendix
In this appendix we describe the issues regarding the simu-lations of the nonlabor incomes for each country
A Argentina Asignacioacuten Universal porHijo (AUH)
Argentina launched the conditional cash transfer programAUH in 2009 The beneficiaries are children under the ageof 18 in households whose members are unemployed orinformal workers The program provides a monthly cashsubsidy per child up to 5 children Since Argentinarsquos house-hold survey (EPH) does not include a question to identifyrecipients of this program we follow Garganta and Gasparini[13] assuming full take-up among those who qualify Thesimulation is made according to the following rule
If119867ℎ119905(1205781199051015840) + 119876
ℎ119905le 5 and119867
ℎ119905+ 119876ℎ119905
le 5
119884119873119871
ℎ119905(1205781199051015840) = 119884
119873119871
ℎ119905+ 120579119905(119867ℎ119905(1205781199051015840) minus 119867
ℎ119905) (A1)
If119867ℎ119905(1205781199051015840) + 119876
ℎ119905gt 5 and119867
ℎ119905+ 119876ℎ119905
le 5
119884119873119871
ℎ119905(1205781199051015840) = 119884
119873119871
ℎ119905+ 120579119905(5 minus 119867
ℎ119905minus 119876ℎ119905) (A2)
If119867ℎ119905(1205781199051015840) + 119876
ℎ119905le 5 and119867
ℎ119905+ 119876ℎ119905
gt 5
119884119873119871
ℎ119905(1205781199051015840) = 119884
119873119871
ℎ119905+ 120579119905(119867ℎ119905(1205781199051015840) + 119876
ℎ119905minus 5) (A3)
If119867ℎ119905(1205781199051015840) + 119876
ℎ119905gt 5 and119867
ℎ119905+ 119876ℎ119905
gt 5
119884119873119871
ℎ119905(1205781199051015840) = 119884
119873119871
ℎ119905 (A4)
10 International Journal of Population Research
where 119876ℎ119905
is the number of children aged 17 or 18(recall that we carry out the simulations for childrenaged 16 or younger) and 120579
119905is equal to AR$270 for
the third trimester of 2012 and AR$340 for the fourthtrimester
B Brazil Bolsa Familia Program (BFP)
The Bolsa Familia implemented in 2003 is the flagshipconditional cash transfer program in Brazil aimed at poorhouseholds The estimated nonlabor incomes in this case arecalculated as follows
If119867119905ℎ
gt 0 and beneficiary
119884119873119871
ℎ119905(1205781199051015840) = 119884
119873119871
119905ℎ+ 119872119905ℎ(119867119905ℎ(1205781199051015840)
119867119905ℎ
minus 1) (B1)
If119867119905ℎ
= 0119867119905ℎ(1205781199051015840) = 119896 and beneficiary
119884119873119871
ℎ119905(1205781199051015840) = 119884
119873119871
119905ℎ+ 119872119896
119905ℎminus 119897119905ℎ (B2)
with 119896 = 1 2 3 119872119905ℎ
being the amount of thetransfer119872119896
119905ℎbeing the average benefit of households
with 119896 children and 119897119905ℎbeing the standard benefit of
recipients without children
C Mexico Oportunidades
Oportunidades is the main antipoverty program of theMexican Government The size of the transfer depends onthe number of children among other characteristics of thehouseholdMexicorsquos ENIGH identifies the beneficiaries of theprogram and the amount of the subsidy that each householdreceives The simulation is made as follows
119884119873119871
ℎ119905(1205781199051015840) = 119884
119873119871
119905ℎ+ 119872119896
119905ℎminus 119872119905ℎ
if 119867119905ℎ(1205781199051015840) = 119896 119867
119905ℎ= 119896 and beneficiary
(C1)
where 119872119896119905ℎis the average transfer of a beneficiary household
with 119896 children and119872119905ℎis the original household transfer
D Uruguay Asignaciones Familiares
The Asignaciones Familiares program in Uruguay is targetedat the children of workers The amount of the transfer varieswith the householdrsquos income Since the size of the transferdepends on characteristics of the child such as the age thesimulation is carried out as follows
If119867119905ℎ
gt 0 and beneficiary
119884119873119871
ℎ119905(1205781199051015840) = 119884
119873119871
119905ℎ+ 119872119905ℎ(119867119905ℎ(1205781199051015840)
119867119905ℎ
minus 1) (D1)
If119867119905ℎ
= 0119867119905ℎ(1205781199051015840) = 119896 and beneficiary
119884119873119871
ℎ119905(1205781199051015840) = 119884
119873119871
119905ℎ+ 119872119896
119905ℎ (D2)
with 119896 = 1 2 3 119872119905ℎ
being the total amount ofthe transfer and 119872
119896
119905ℎbeing the average transfer of a
household with 119896 children
Disclosure
The information presented herein is solely the responsibilityof the authors and does not necessarily reflect the opinions ofthe aforementioned entities
Competing Interests
The authors declare that they have no competing interests
Acknowledgments
This paper was written in the framework of the projectEnhancing Womenrsquos Economic Empowerment throughBetter Policies in Latin America a joint initiative betweenCEDLAS-UNLP and CIEDUR with the support ofIDRC Canada The paper includes material from NicolasBadaraccorsquos Masters dissertation at Universidad Nacionalde La Plata The authors are grateful to Carolina RobinoGuillermo Cruces Jorge Paz Facundo Albornoz andseminar participants at AAEP (Posadas 2014) Network ofInequality and Poverty (UNGS 2014) and Conference onSocial Protection (IDRC and IPC Brasilia 2014) for valuablecomments and suggestions
References
[1] T W Guinnane ldquoThe historical fertility transition a guide foreconomistsrdquo Journal of Economic Literature vol 49 no 3 pp589ndash614 2011
[2] L E Jones and M Tertilt ldquoAn economic history of fertility inthe US 1826ndash1960rdquo NBERWorking Paper 12796 2006
[3] J Chackiel La Dinamica Demografica en America LatinaSerie Poblacion y Desarrollo No 52 Centro Latinoamericanoy Caribeno de Demografıa-Division de Poblacion de CEPAL2004
[4] ECLAC ldquoLa fecundidad en America Latina un descensoacelerado y heterogeneo con profundas transformacionesdemograficas y socialesrdquo in Fertility Latin American and theCaribbean Demographic Observatory no 5 Centro Latinoa-mericano y Caribeno de DemografıamdashDivision de Poblacionde CEPAL 2008
[5] ECLAC ldquoChapter II current situation and outlook for fertilityin Latin Americardquo in Social Panorama of Latin America 2011ECLAC 2011
[6] L Gasparini and M Marchionni Bridging Gender Gaps TheRise and Deceleration of Female Labor Force ParticipationCEDLAS-UNLP 2015
[7] L Gasparini and N Lustig ldquoThe rise and fall of incomeinequality in Latin Americardquo inThe Oxford Handbook of LatinAmerican Economics J Ocampo and J Ros Eds chapter 27 pp691ndash714 Oxford University Press Oxford UK 2011
[8] G Cornia Ed Falling Inequality in Latin America PolicyChanges and Lessons Oxford University Press 2014
International Journal of Population Research 11
[9] F Alvaredo and L Gasparini ldquoRecent trends in inequalityand poverty in developing countriesrdquo in Handbook of IncomeDistribution Atkinson and Bourguignon Eds vol 2 chapter 9pp 697ndash805 Elsevier New York NY USA 2015
[10] M Marchionni and L Gasparini ldquoTracing out the effects ofdemographic changes on the income distribution The caseof Greater Buenos Aires 1980ndash1998rdquo Journal of EconomicInequality vol 5 no 1 pp 97ndash114 2007
[11] J M Wooldridge Econometric Analysis of Cross Section andPanel Data MIT Press Boston Mass USA 2002
[12] L Gasparini M Marchionni and W Sosa Escudero ldquoChar-acterization of inequality changes through microeconometricdecompositions The case of Greater Buenos Airesrdquo in TheMicroeconomics of Income Distribution Dynamics in East Asiaand Latina America F Bourguignon F Ferreira and N LustigEds Oxford University Press New York NY USA 2004
[13] S Garganta and L Gasparini ldquoThe impact of a social programon labor informality the case of AUH in Argentinardquo Journal ofDevelopment Economics vol 115 pp 99ndash110 2015
Submit your manuscripts athttpwwwhindawicom
Child Development Research
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Education Research International
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Biomedical EducationJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Psychiatry Journal
ArchaeologyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
AnthropologyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Research and TreatmentSchizophrenia
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Urban Studies Research
Population ResearchInternational Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CriminologyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Aging ResearchJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
NursingResearch and Practice
Current Gerontologyamp Geriatrics Research
Hindawi Publishing Corporationhttpwwwhindawicom
Volume 2014
Sleep DisordersHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
AddictionJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Depression Research and TreatmentHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Geography Journal
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Research and TreatmentAutism
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Economics Research International
International Journal of Population Research 3
15
20
25
30
35
40
4519
9219
9319
9419
9519
9619
9719
9819
9920
0020
0120
0220
0320
0420
0520
0620
0720
0820
0920
1020
1120
12
Headcount ratio (US$ 4 line)
45
47
49
51
53
55
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
Gini coefficient
Figure 3 Poverty and income inequality Latin America Source own calculations based on microdata from national household surveysNote average of seven Latin American countries Argentina Brazil Chile El Salvador Mexico Peru and Uruguay Unweighted meansPoverty and inequality computed over the distribution of household per capita income
poverty Figure 3 shows the average income inequality andpoverty trends of seven Latin American countries Povertyis measured by the headcount ratio with a daily US$ 4 linewhile income inequality is measured by the Gini coefficientBoth indices are calculated over the household per capitaincome distribution Income poverty slightly decreased dur-ing the 1990s and fell sharply during the 2000s ArgentinaUruguay and Mexico suffered an increase in poverty duringthe 1990s and then a continuous fall over the following decade(Figure 4) Meanwhile Brazil Chile El Salvador and Peruexperienced a decrease in poverty over the entire periodthough steeper during the 2000s
On average for our sample of seven countries incomeinequality increased during the first decade under analysisand then significantly decreased during the 2000s The trendfor the 1990s is driven by the sharp increase in incomedispersion inArgentina andUruguay (Figure 5) and the smallchanges in the rest of the countries In contrast during the2000s the fall in inequality was strong and generalized to alleconomies
3 Data and Methodology
We focus on seven Latin American countries ArgentinaBrazil Chile El Salvador Mexico Peru and Uruguay duringthe 1990s and 2000sThe analysis is based onmicrodata fromnational household surveys collected by the correspondingNational Statistical Offices Table 1 presents a brief descrip-tion of these databases Since surveys are not homogeneouswe have made all possible efforts to make variables compara-ble across countries and over time by using similar definitionsof variables in each countryyear and by applying consistentmethods of processing the data
The ideal variable for capturing fertility would be thenumber of children either living in the household or notUnfortunately this variable is usually not available in LatinAmerican household surveys with the exception of BrazilAs a proxy we use the number of children living in thehousehold Consequently two main problems emerge First
Table 1 Household surveys
Country Survey Acronym Years
Argentina Encuesta Permanente deHogares EPH 1992ndash2012
Brazil Pesquisa Nacional porAmostra de Domicilios PNAD 1990ndash2012
ChileEncuesta de
CaracterizacionSocioeconomica Nacional
CASEN 1990ndash2011
El Salvador Encuesta de Hogares dePropositos Multiples EHPM 1991ndash2010
MexicoEncuesta Nacional deIngresos y Gasto de los
HogaresENIGH 1992ndash2012
Peru Encuesta Nacional deHogares ENAHO 1997ndash2012
Uruguay Encuesta Continua deHogares ECH 1995ndash2012
the measurement error of this proxy depends on the ageof parents and children that is on average the older theparents and the older the children are the more likely theyleave the parental home Therefore the observed differencesin the number of children across households are not onlydue to reproductive decisions but also due to the life cycleTo deal with this shortcoming we restrict our sample tohouseholds where the head of household is between 25 and45 years old Many people below 25 years old may not haveyet started childbearing The average age at childbearing forthe countries under analysis was 275 years in 1993 and 276years in 2011 The dispersion across countries is also smallranging from 265 in Mexico to 283 in Peru The average ageat marriage in the early 1990s was 229 and 256 years forwomen and men respectively and about two years more inthe early 2010s (250 and 275)
Another problem arises since Latin American house-hold surveys usually register family relationships among
4 International Journal of Population Research
0102030405060
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
0102030405060
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
0102030405060
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
0102030405060
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
0102030405060
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
0102030405060
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
0102030405060
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
Argentina Brazil Chile
El Salvador Mexico Peru
Uruguay
Figure 4 PovertyHeadcount ratio (US$ 4 line) Source own calculations based onmicrodata fromnational household surveys Note povertycomputed over the distribution of household per capita income
household members only in terms of the head of householdthat is it is possible to match children with their parentsonly when they are the head of household or the spouseTherefore we further restrict the analysis to only considerfertility decisions of the head of household and the spouse
Regarding the empirical specification and estimationmethodology we follow Marchionni and Gasparini [10]The main inputs to carry out the microsimulations arethe estimates of the parameters that govern fertility deci-sionsoutcomes (for simplicity we refer to fertility decisionsalthough fertility outcomes could be the result of free con-scious choices but also the consequence of various other cir-cumstances) and the response of labor market participationto changes in family size We assume that the number ofchildren in a household follows a Poisson process and thatits parameters can be consistently estimated using a Poissonregression model Hourly wages and hours of work areassumed to be simultaneously determined in an equilibriummodel of the labor market
After estimating the parameters we carry out the sim-ulations That is we simulate the counterfactual incomedistribution that arises in a given base year by assumingthat the population in that year takes fertility decisions
according to the parameters estimated for a different yearTheresulting poverty and inequality measures over the simulateddistribution are compared to those actually observed in thebase year The difference between the simulated value ofan indicator of poverty or inequality and its actual value isinterpreted as a measure of the direct impact of the change infertility behavior that is the impact that would take place incase all other covariates remained fixed
The methodology requires keeping all other things con-stant when simulating the impact of changes in fertilitydecisions Naturally as it was stressed in previous sectionssome of these factors may be correlated with fertility Forinstance the structure ofwagesmay respond to changes in thelabor supply triggered by a change in fertility Additionallychanges in reproductive behavior may have been inducedby income changes in which case the microsimulation onlycaptures a single round of effects from fertility to incomes ofa more complicated process
In a given period income poverty and inequality areaffected by a host of factors including economic shocks andsocial policies The microsimulation methodology is usefulas it allows approximating the size of the impact that agiven factor would have in case all other factors remained
International Journal of Population Research 5
40
45
50
55
60
6519
9219
9419
9619
9820
0020
0220
0420
0620
0820
1020
12
40
45
50
55
60
65
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
40
45
50
55
60
65
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
40
45
50
55
60
65
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
40
45
50
55
60
65
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
40
45
50
55
60
65
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
40
45
50
55
60
65
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
Argentina Brazil Chile
El Salvador Mexico Peru
Uruguay
Figure 5 Income inequality Gini coefficient Source own calculations based onmicrodata fromnational household surveys Note inequalitycomputed over the distribution of household per capita income
fixed Although this is clearly an unrealistic scenario theestimations are still useful as a first approximation of theimpact especially considering the fact that the more ambi-tious alternative of estimating a general equilibriummodel isusually unfeasible
31 Microsimulations We carry out the simulation over thedistribution of household per capita income defined as
119910119894119905=
119884119871
ℎ119905+ 119884119873119871
ℎ119905
119873ℎ119905
forall119894 isin ℎ at time 119905 (1)
where 119894 indexes individuals ℎ households and 119905 time periods(years) 119884119871
ℎ119905denotes total labor income of household ℎ at time
119905 119884119873119871
ℎ119905labels nonlabor income and 119873
ℎ119905is the family size
which is the sum of the number of children up to 16 yearsold119867
ℎ119905and the rest of the household members 119877
ℎ119905
119873ℎ119905
= 119867ℎ119905+ 119877ℎ119905 (2)
Weneed to specify how119867ℎ119905is determined in order to simulate
the counterfactual distributions The literature argues that
fertility outcomes are the result of a process affected bycharacteristics of each spouse and of the household amongother factors This process can be formalized as
119867ℎ119905
= 119867 (119885ℎ119905 119890ℎ119905 120578119905) (3)
where 119885ℎ119905
is a vector of observable characteristics and119890ℎ119905
includes all unobservable characteristics that influencethe familyrsquos reproductive behavior while 120578
119905is the set of
parameters that govern fertility decisionsWe assume that labor income for individual 119894 is given by
119884119871
119894119905= 119908119894119905119871119894119905 (4)
where 119908119894119905
is the hourly wage rate and 119871119894119905
is the num-ber of hours worked by individual 119894 Wages and hoursof work are defined as 119908
119894119905= 119908(119883
1119894119905 120576119882
119894119905 120573119905) and 119871
119894119905=
119871(1198832119894119905 119867119894119905 120576119871
119894119905 120574119905 120582119905) respectively Where 119883
1119894119905and 119883
2119894119905are
observable characteristics 120576119882119894119905
and 120576119871
119894119905are unobservable char-
acteristics while 120573119905 120574119905 and 120582
119905are parameters In particular
120582119905is the parameter that relates the hours of work to the
number of children
6 International Journal of Population Research
Nonlabor income is usually assumed to be exogenous inthemicrosimulation literature However as stated in previoussections the expansion of social programs to families withchildren suggests a potential link between fertility decisionsand nonlabor income Accordingly we assume that nonlaborincome is given by
119884119873119871
ℎ119905= 119898ℎ119905+ 119892 (119867
ℎ119905) (5)
where119898ℎ119905is exogenous and119892(119867
ℎ119905) represents the component
of nonlabor income that depends on the number of childrenWe label the parameters of the fertility decisions at time 119905
as 120578119905and those at time 1199051015840 as 120578
1199051015840 A key step in themethodology
is to estimate the counterfactual number of children in agiven year 119905 if fertility outcomes were determined as in analternative year 119905
1015840 Once this term labeled as 119867ℎ119905(1205781199051015840) is
estimated three microsimulation exercises are carried out byreplacing this estimate in the household per capita incomeequation
The first exercise aims to capture the contribution of thechange in fertility parameters 120578 to the actual change in theincome distribution through the denominator of (1) Thesimulated per capita income is given by
119910119863
119894119905=
119884119871
ℎ119905+ 119884119873119871
ℎ119905
119873ℎ119905(1205781199051015840)
(6)
where 119873ℎ119905(1205781199051015840) = 119877
ℎ119905+ 119867ℎ119905(1205781199051015840) Superscript 119863 in the
simulated per capita income stands for the direct-size effectthat is the change in the income distribution due to changesin the number of household members among whom totalhousehold income should be distributed
The second exercise involves simulating the labor incomethat a household would have in year 119905 if parameters thatgovern fertility decisionswere those of year 1199051015840 In other wordsthe exercise consists in replacing the simulated number ofchildren 119867
ℎ119905(1205781199051015840) in the equation of individual labor income
(4) to obtain
119910119867
119894119905=
119884119871
ℎ119905(1205781199051015840) + 119884
119873119871
ℎ119905
119873ℎ119905
(7)
Superscript 119867 in the simulated per capita income stands forthe hours-size effect that is the contribution of the changein fertility parameters 120578 to the actual change in the incomedistribution through the indirect channel of affecting thehours-of-work decisions
We carry out a third exercise by simulating the coun-terfactual distribution arising from a change in nonlaborincome driven by changes in fertility decisions We refer tothe distributional impact of changes through this channelas the nonlabor-income effect and thus label the per capitaincome variable in (8) with the NL superscript
119910119873119871
119894119905=
119884119871
ℎ119905+ 119884119873119871
ℎ119905(1205781199051015840)
119873ℎ119905
(8)
Finally (9) considers the three channels simultaneously toobtain the total effect (superscript119879) of changes in the fertilityparameters
119910119879
119894119905=
119884119871
ℎ119905(1205781199051015840) + 119884
119873119871
ℎ119905(1205781199051015840)
119873ℎ119905(1205781199051015840)
(9)
We denote the per capita income distribution among indi-viduals in year 119905 as 119863
119905(119910119905) = (119910
1119905 1199102119905 119910
119899119905) where 119899
denotes the total number of individuals in the populationLet 119868(sdot) be a distributive indicator as the Gini coefficientor the poverty headcount ratio Then the contribution ofthe change in fertility parameters 120578 to the actual change inthe income distribution through the direct channel is givenby (10) The contributions through the other channels arecomputed analogously
119868 (119863119905(119910119905)) minus 119868 (119863
119905(119910119863
119905)) (10)
A shortcoming of the microsimulation methodology is thatdecomposition is path-dependent This means that takingyear 119905 or year 1199051015840 as the base year does not necessarily implythe same result Accordingly we perform the simulationsalternating the base year and report the average effect for eachexercise
In the following subsections we provide details concern-ing the specification and estimation of fertility labor incomeand nonlabor-income equations
32 Fertility Decisions We assume that the number of chil-dren in (3) follows a Poisson process with parameter 120583
ℎ119905
Formally
119867ℎ119905
sim Poisson (120583ℎ119905) with 120583
ℎ119905= 119864 (119867
ℎ119905| 119885ℎ119905) = exp (119885
1015840
ℎ119905120578119905) (11)
Then
Prob (119867ℎ119905
= 1198670) =
exp (minus120583ℎ119905) (120583ℎ119905)1198670
1198670
with 1198670= 0 1 2 (12)
Equation (12) is the Poisson regressionmodel fromwhich it ispossible to consistently estimate parameters 120578
119905by maximum
likelihood (it can be shown that consistency holds for themaximum likelihood estimators of 120578
119905as long as the real
distribution is any of the linear exponential family to whichthe Poisson distribution belongs provided that the condi-tional mean in (11) is correctly specified see eg Wooldridge[11]) We estimate separate models for two-parent householdand single-parent households For simplicity we denote theestimates by 120578
Once parameters 120578 are estimated we proceed to simulatechanges in the number of children Since the objective isto simulate these changes as a consequence of changes onlyin the parameters 120578 it is necessary to keep unobservablefactors fixed Therefore we characterize each household bythe quantile it occupies in the distribution of children ofyear 119905 Let 119865
120578119905|119885ℎ119905
(sdot) be the cumulative probability function ofa random variable that follows a Poisson distribution withparameter equal to exp(1198851015840
ℎ119905120578119905) and let 119902
ℎ119905be the quantile
for household ℎ at time 119905 that is 119865120578119905|119885ℎ119905
(119867ℎ119905) = 119902
ℎ119905 The
International Journal of Population Research 7
simulated number of children in household ℎ will be the onethat places it in the 119902
ℎ119905quantile of the distribution of children
with the relevant parameters of time 1199051015840(1205781199051015840) conditional to
the observable characteristics 119885ℎ119905 Formally the simulated
number of children in household ℎ at year 119905 using theestimated fertility parameters for year 1199051015840 is given by
119867ℎ119905(1205781199051015840) = 119865
minus1
1205781199051015840 |119885ℎ119905
∘ 119865120578119905|119885ℎ119905
(119867ℎ119905) (13)
33 Labor Incomes To estimate individual labor income in(4) we follow Gasparini et al [12] in assuming that bothwages and hours worked are determined in a reduced-formmodel of the labor market equilibrium
ln119908lowast
119894119905= 1198831015840
1119894119905120573119905+ 120576119882
119894119905 (14)
119871lowast
119894119905= 1198831015840
2119894119905120574119905+ 119867119894119905120582119905+ 120576119871
119894119905 (15)
with
if 119871lowast119894119905gt 0
119908119894119905= 119908lowast
119894119905
119871119894119905= 119871lowast
119894119905
if 119871lowast119894119905le 0
119908119894119905= 0
119871119894119905= 0
(120576119882
119894119905 120576119871
119894119905) sim 119873(0 0 120590
2
119882119905 1205902
119871119905 120588119905)
(16)
where119908lowast119894119905and119871lowast
119894119905are unobserved latent variables Vectors119883
1119894119905
and 1198832119894119905
include observable factors affecting hourly wagesand hours of work respectively while 119867
119894119905stands for the
number of children 120573119905 120574119905(vectors) and 120582
119905are parameters
to be estimated along with 1205902
119882119905 1205902119871119905 and 120588
119905
We estimate (14) by Heckmanrsquos maximum likelihoodmethod where a censored version of (15) is used as a selectionequation replacing hours of work by a binary variable thatindicates whether the individual works For estimation of(15) we use aTobitmodel (although this estimation strategy isnot fully efficient efficiency loss is not necessarily significantfor a given sample size and this alternative has certaincomputational advantages over a full information procedurefor more details see Gasparini et al [12]) We estimateseparate models for head of households and spouses
To simulate changes in individual labor incomes wereplace the observed number of children of individual 119894 atyear 119905 in (15) by the simulated number of children 119867
119905ℎ(1205781199051015840)
while assigning himher a wage based on (14) and an errorterm drawn from the bivariate distribution implicit in themodel
34 Nonlabor Incomes The measure of nonlabor incomeincludes pensions capital income and transfers Conditionalcash transfers are included as government transfers Wesimulate nonlabor incomes in an ad hoc manner taking
into account the characteristics of each countryrsquos socialprogram In those cases where it is not possible to identifythe recipients in the household survey we identify potentialbeneficiaries and estimate the amount of the transfers basedon the design of each program The general framework is asfollows nonlabor income of household ℎ at year 119905 is given by(5) where 119892(119867
ℎ119905) represents the part of nonlabor income that
depends on the number of children Typically cash transfersin poverty-alleviation programs vary with the number ofchildren in the family For example if the program has a fixedamount 120579
119905per child the nonlabor income becomes
119884119873119871
ℎ119905= 119898ℎ119905+ 120579119905119867ℎ119905 (17)
The simulated nonlabor income is calculated using thecounterfactual number of children in a given year 119905 if fertilityoutcomes were determined as in an alternative year 1199051015840
119884119873119871
ℎ119905(1205781199051015840) = 119898
ℎ119905+ 120579ℎ119905119867ℎ119905(1205781199051015840) (18)
Even though all programs have a similar setting to the onedescribed they differ in some aspects maximum number ofchildren age and in some cases how to identify the programrsquosrecipients The details are given in the Appendix We sim-ulate nonlabor incomes for Argentina Brazil Mexico andUruguay Even though the programs in Chile El Salvadorand Peru have similar characteristics it is not possible toestimate how the amount of transfers varies with the numberof children given the information in the household surveysFor these countries we assume that nonlabor income iscompletely exogenous
4 The Results
The results of the microsimulations regarding poverty arepresented inTable 2 Column (i) displays the observed changein the headcount ratio while the rest of the columns presentthe impact of the changes in fertility on poverty Column(v) shows the sum of all effects explored In order to betterunderstand the information in the table take the case ofBrazil as an example Column (i) shows that between 1990and 2012 the poverty headcount ratio in that country (usingthe line of US$ 4 a day) fell by around 25 points Thevalue in the last column has the following interpretationif fertility had been the only factor that changed duringthat period then the poverty headcount ratio in Brazilwould have fallen by 392 points This change is statisticallysignificant although not very big compared to the actual fallin poverty in that period The overall effect is the result ofa significant direct effect (minus374) and two smaller effects thatgo in different directions On the one hand the differentialreduction in fertility implied a greater increase in hours ofwork among more disadvantaged households which in turncontributed though very slightly to a further reduction inincome poverty On the other hand the falling patterns infertility among the poor ameliorated the poverty-decreasingimpact of the conditional cash transfer programs that aretargeted at families with children (mainly the Bolsa Familiain the 2000s) However this effect was quantitatively almostinsignificant
8 International Journal of Population Research
Table 2 Impact of fertility changes on poverty Changes in the poverty headcount ratio (US$ 4 line)
Country Period Observed change (i) EffectsDirect (ii) Hours (iii) NLI (iv) Total (v)
Argentina 1992ndash2012 minus1122 minus139 004 039 minus080(008) (002) (001) (001) (002)
Brazil 1990ndash2012 minus2516 minus374 minus017 014 minus392(003) (001) (000) (000) (001)
Chile 1990ndash2011 minus3324 minus156 018 mdash minus143(005) (001) (000) mdash (001)
El Salvador 1991ndash2010 minus1412 minus498 minus034 mdash minus531(006) (004) (001) mdash (005)
Mexico 1992ndash2012 minus350 minus435 minus030 025 minus443(010) (004) (002) (001) (004)
Peru 1997ndash2012 minus2320 minus383 minus019 mdash minus394(009) (003) (001) mdash (003)
Uruguay 1995ndash2012 minus156 minus008 000 002 minus005(003) (002) (000) (000) (002)
Source own calculations based on microdata from national household surveysNote All effects are significant at the 1 level The standard errors were calculated using bootstrap with 200 replications The values of each effect are averagesthat result from taking alternatively each year in the comparison as the base year The sample includes only households in which the head is between 25 and45 years old NLI nonlabor income
Table 3 Impact of fertility changes on inequality Changes in the Gini coefficient
Country Period Observed change (i) EffectsDirect (ii) Hours (iii) NLI (iv) Total (v)
Argentina 1992ndash2012 minus544 minus104 minus001 013 minus095(010) (001) (000) (000) (002)
Brazil 1990ndash2012 minus718 minus130 minus006 006 minus132(004) (000) (000) (000) (000)
Chile 1990ndash2011 minus178 minus075 004 mdash minus073(008) (001) (000) mdash (001)
El Salvador 1991ndash2010 minus768 minus130 minus007 mdash minus140(006) (002) (000) mdash (002)
Mexico 1992ndash2012 minus086 minus168 minus011 012 minus169(019) (002) (000) (000) (002)
Peru 1997ndash2012 minus732 minus192 minus006 mdash minus199(013) (001) (000) mdash (001)
Uruguay 1995ndash2012 minus203 minus012 minus001 002 minus013(004) (001) (000) (000) (001)
Source own calculations based on microdata from national household surveysNote All effects are significant at the 1 level The standard errors were calculated using bootstrap with 200 replications The values of each effect are averagesthat result from taking alternatively each year in the comparison as the base year The sample includes only households in which the head is between 25 and45 years old NLI nonlabor income
Changes in fertility patterns over the last two decades inLatin America have implied a reduction in income povertyThe impact is statistically significant and in some caseseconomically large For instance the estimated poverty-reduction effect was more than 4 points in Mexico andEl Salvador Most of the effect comes through the directeffect a reduction in fertility rates among the most dis-advantaged groups reduced family size and increased percapita income The hours-of-work effect is in most casespoverty reducing and the nonlabor-income effect is poverty
increasing but in both cases the estimated sizes aresmall
Table 3 shows the results of the simulations on incomeinequality measured by the Gini coefficient Inequality inBrazil as measured by the Gini coefficient fell by 718points between 1990 and 2012 If fertility had been the onlyfactor that changed in that period then the Gini coefficientwould have fallen by 132 points (column v) This change isstatistically significant representing around 18 of the actualreduction in inequality during that period The differential
International Journal of Population Research 9
reduction in fertility across socioeconomic groups in Brazilcontributed to the observed decline in inequality in the lasttwo decadesThe overall effect ofminus132 points in column (v) isthe result of a significant direct effect (minus130) and two smallereffects going in different directions
Some interesting general results emerge from Table 3First changes in fertility have implied a decline in incomeinequality The differential pattern in fertility across groupsexperienced in most Latin American countries over thelast decades translated into an equalizing impact on theincome distribution Second this effect is small althoughnot negligible To be sure demographic changes are not thecentral reason behind changes in income inequality but theyare statistically significant and economically relevant Onaverage fertility changes account for a fall of around onepoint in the Gini coefficient
Third most of the impact comes from the direct effectThe differential fall in fertility rates among socioeconomicgroups had a larger impact on the family size of poorerfamilies implying a proportionally larger increasing effect ontheir per capita incomes
Fourth the effect of fertility changes on hours of workand in turn on incomes is smaller and has different signsacross countries It is important to notice that a more intensereduction in fertility among the poor may be consistentwith a positive sign (inequality-increasing) for this effectThis could happen if the elasticity of hours of work withrespect to the number of children is higher among thenonpoor and if the extra hours worked triggered by the fallin fertility imply a household income increase proportionallylarger for the nonpoor Anyway with one exception thehours-of-work effect has a negative sign implying that theasymmetric reduction in fertility contributed to a reductionin income inequality by fostering higher participation in thelabor market among poor parents (mostly mothers)
Finally the effect through nonlabor incomes is alwaysinequality-increasing but small The fall in fertility amongthe poor implied a lower inequality-reducing impact of theconditional cash transfer programs that were introduced andexpanded in the region mainly during the 2000s The effecthowever is in most cases very small
5 Concluding Remarks
Poverty and inequality fell in Latin America over the lastdecades driven by a large number of factors In thispaper we highlight one factor that is usually overlooked inthe economic literature the impact of changes in fertilityThe average number of children fell in households fromall groups but especially among the more vulnerable afact that could have distributive implications We quantifythe consequences of this demographic pattern by applyingmicrosimulation techniquesThe results suggest that changesin fertility that took place over the last decades in LatinAmerica contributed to the reduction in income inequalityas well as in poverty The impacts found were in generalstatistically significant and economically relevant althoughnot very large Although surely there are more importantdeterminants behind the observed patterns in the income
distributions of the Latin American countries demographicfactors should not be overlooked
The paper has focused on three simple yet central chan-nels the change in the number of children (i) enlarges thefamily size and reduces per capita income (ii) affects thelabor supply decisions and (iii) modifies the income supportfrom social programs Certainly there are other potentiallyrelevant channels that are ignored in the paper For instancea reduction in the number of children per family couldimply higher inheritances and higher family spending oneducation and health per child and hence better incomeperspectives Also smaller families may be better positionedto take advantage of labor market opportunities for instancethrough migration Moreover lower fertility rates may beassociated with a reduction in the relative supply of unskilledlabor and hence an increase in its relative wage
Although in the short and medium run lower fertility inmore disadvantaged households may contribute to reducingpoverty and inequality in the long run the implicationsare more nuanced When children become adults they maycontribute to their parentsrsquo incomes a fall in fertility mayreduce incomes for the elderly especially in countries wherethe pension system is weak and ultimately contribute tohigher poverty through that intertemporal channel
Appendix
In this appendix we describe the issues regarding the simu-lations of the nonlabor incomes for each country
A Argentina Asignacioacuten Universal porHijo (AUH)
Argentina launched the conditional cash transfer programAUH in 2009 The beneficiaries are children under the ageof 18 in households whose members are unemployed orinformal workers The program provides a monthly cashsubsidy per child up to 5 children Since Argentinarsquos house-hold survey (EPH) does not include a question to identifyrecipients of this program we follow Garganta and Gasparini[13] assuming full take-up among those who qualify Thesimulation is made according to the following rule
If119867ℎ119905(1205781199051015840) + 119876
ℎ119905le 5 and119867
ℎ119905+ 119876ℎ119905
le 5
119884119873119871
ℎ119905(1205781199051015840) = 119884
119873119871
ℎ119905+ 120579119905(119867ℎ119905(1205781199051015840) minus 119867
ℎ119905) (A1)
If119867ℎ119905(1205781199051015840) + 119876
ℎ119905gt 5 and119867
ℎ119905+ 119876ℎ119905
le 5
119884119873119871
ℎ119905(1205781199051015840) = 119884
119873119871
ℎ119905+ 120579119905(5 minus 119867
ℎ119905minus 119876ℎ119905) (A2)
If119867ℎ119905(1205781199051015840) + 119876
ℎ119905le 5 and119867
ℎ119905+ 119876ℎ119905
gt 5
119884119873119871
ℎ119905(1205781199051015840) = 119884
119873119871
ℎ119905+ 120579119905(119867ℎ119905(1205781199051015840) + 119876
ℎ119905minus 5) (A3)
If119867ℎ119905(1205781199051015840) + 119876
ℎ119905gt 5 and119867
ℎ119905+ 119876ℎ119905
gt 5
119884119873119871
ℎ119905(1205781199051015840) = 119884
119873119871
ℎ119905 (A4)
10 International Journal of Population Research
where 119876ℎ119905
is the number of children aged 17 or 18(recall that we carry out the simulations for childrenaged 16 or younger) and 120579
119905is equal to AR$270 for
the third trimester of 2012 and AR$340 for the fourthtrimester
B Brazil Bolsa Familia Program (BFP)
The Bolsa Familia implemented in 2003 is the flagshipconditional cash transfer program in Brazil aimed at poorhouseholds The estimated nonlabor incomes in this case arecalculated as follows
If119867119905ℎ
gt 0 and beneficiary
119884119873119871
ℎ119905(1205781199051015840) = 119884
119873119871
119905ℎ+ 119872119905ℎ(119867119905ℎ(1205781199051015840)
119867119905ℎ
minus 1) (B1)
If119867119905ℎ
= 0119867119905ℎ(1205781199051015840) = 119896 and beneficiary
119884119873119871
ℎ119905(1205781199051015840) = 119884
119873119871
119905ℎ+ 119872119896
119905ℎminus 119897119905ℎ (B2)
with 119896 = 1 2 3 119872119905ℎ
being the amount of thetransfer119872119896
119905ℎbeing the average benefit of households
with 119896 children and 119897119905ℎbeing the standard benefit of
recipients without children
C Mexico Oportunidades
Oportunidades is the main antipoverty program of theMexican Government The size of the transfer depends onthe number of children among other characteristics of thehouseholdMexicorsquos ENIGH identifies the beneficiaries of theprogram and the amount of the subsidy that each householdreceives The simulation is made as follows
119884119873119871
ℎ119905(1205781199051015840) = 119884
119873119871
119905ℎ+ 119872119896
119905ℎminus 119872119905ℎ
if 119867119905ℎ(1205781199051015840) = 119896 119867
119905ℎ= 119896 and beneficiary
(C1)
where 119872119896119905ℎis the average transfer of a beneficiary household
with 119896 children and119872119905ℎis the original household transfer
D Uruguay Asignaciones Familiares
The Asignaciones Familiares program in Uruguay is targetedat the children of workers The amount of the transfer varieswith the householdrsquos income Since the size of the transferdepends on characteristics of the child such as the age thesimulation is carried out as follows
If119867119905ℎ
gt 0 and beneficiary
119884119873119871
ℎ119905(1205781199051015840) = 119884
119873119871
119905ℎ+ 119872119905ℎ(119867119905ℎ(1205781199051015840)
119867119905ℎ
minus 1) (D1)
If119867119905ℎ
= 0119867119905ℎ(1205781199051015840) = 119896 and beneficiary
119884119873119871
ℎ119905(1205781199051015840) = 119884
119873119871
119905ℎ+ 119872119896
119905ℎ (D2)
with 119896 = 1 2 3 119872119905ℎ
being the total amount ofthe transfer and 119872
119896
119905ℎbeing the average transfer of a
household with 119896 children
Disclosure
The information presented herein is solely the responsibilityof the authors and does not necessarily reflect the opinions ofthe aforementioned entities
Competing Interests
The authors declare that they have no competing interests
Acknowledgments
This paper was written in the framework of the projectEnhancing Womenrsquos Economic Empowerment throughBetter Policies in Latin America a joint initiative betweenCEDLAS-UNLP and CIEDUR with the support ofIDRC Canada The paper includes material from NicolasBadaraccorsquos Masters dissertation at Universidad Nacionalde La Plata The authors are grateful to Carolina RobinoGuillermo Cruces Jorge Paz Facundo Albornoz andseminar participants at AAEP (Posadas 2014) Network ofInequality and Poverty (UNGS 2014) and Conference onSocial Protection (IDRC and IPC Brasilia 2014) for valuablecomments and suggestions
References
[1] T W Guinnane ldquoThe historical fertility transition a guide foreconomistsrdquo Journal of Economic Literature vol 49 no 3 pp589ndash614 2011
[2] L E Jones and M Tertilt ldquoAn economic history of fertility inthe US 1826ndash1960rdquo NBERWorking Paper 12796 2006
[3] J Chackiel La Dinamica Demografica en America LatinaSerie Poblacion y Desarrollo No 52 Centro Latinoamericanoy Caribeno de Demografıa-Division de Poblacion de CEPAL2004
[4] ECLAC ldquoLa fecundidad en America Latina un descensoacelerado y heterogeneo con profundas transformacionesdemograficas y socialesrdquo in Fertility Latin American and theCaribbean Demographic Observatory no 5 Centro Latinoa-mericano y Caribeno de DemografıamdashDivision de Poblacionde CEPAL 2008
[5] ECLAC ldquoChapter II current situation and outlook for fertilityin Latin Americardquo in Social Panorama of Latin America 2011ECLAC 2011
[6] L Gasparini and M Marchionni Bridging Gender Gaps TheRise and Deceleration of Female Labor Force ParticipationCEDLAS-UNLP 2015
[7] L Gasparini and N Lustig ldquoThe rise and fall of incomeinequality in Latin Americardquo inThe Oxford Handbook of LatinAmerican Economics J Ocampo and J Ros Eds chapter 27 pp691ndash714 Oxford University Press Oxford UK 2011
[8] G Cornia Ed Falling Inequality in Latin America PolicyChanges and Lessons Oxford University Press 2014
International Journal of Population Research 11
[9] F Alvaredo and L Gasparini ldquoRecent trends in inequalityand poverty in developing countriesrdquo in Handbook of IncomeDistribution Atkinson and Bourguignon Eds vol 2 chapter 9pp 697ndash805 Elsevier New York NY USA 2015
[10] M Marchionni and L Gasparini ldquoTracing out the effects ofdemographic changes on the income distribution The caseof Greater Buenos Aires 1980ndash1998rdquo Journal of EconomicInequality vol 5 no 1 pp 97ndash114 2007
[11] J M Wooldridge Econometric Analysis of Cross Section andPanel Data MIT Press Boston Mass USA 2002
[12] L Gasparini M Marchionni and W Sosa Escudero ldquoChar-acterization of inequality changes through microeconometricdecompositions The case of Greater Buenos Airesrdquo in TheMicroeconomics of Income Distribution Dynamics in East Asiaand Latina America F Bourguignon F Ferreira and N LustigEds Oxford University Press New York NY USA 2004
[13] S Garganta and L Gasparini ldquoThe impact of a social programon labor informality the case of AUH in Argentinardquo Journal ofDevelopment Economics vol 115 pp 99ndash110 2015
Submit your manuscripts athttpwwwhindawicom
Child Development Research
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Education Research International
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Biomedical EducationJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Psychiatry Journal
ArchaeologyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
AnthropologyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Research and TreatmentSchizophrenia
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Urban Studies Research
Population ResearchInternational Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CriminologyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Aging ResearchJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
NursingResearch and Practice
Current Gerontologyamp Geriatrics Research
Hindawi Publishing Corporationhttpwwwhindawicom
Volume 2014
Sleep DisordersHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
AddictionJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Depression Research and TreatmentHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Geography Journal
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Research and TreatmentAutism
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Economics Research International
4 International Journal of Population Research
0102030405060
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
0102030405060
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
0102030405060
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
0102030405060
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
0102030405060
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
0102030405060
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
0102030405060
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
Argentina Brazil Chile
El Salvador Mexico Peru
Uruguay
Figure 4 PovertyHeadcount ratio (US$ 4 line) Source own calculations based onmicrodata fromnational household surveys Note povertycomputed over the distribution of household per capita income
household members only in terms of the head of householdthat is it is possible to match children with their parentsonly when they are the head of household or the spouseTherefore we further restrict the analysis to only considerfertility decisions of the head of household and the spouse
Regarding the empirical specification and estimationmethodology we follow Marchionni and Gasparini [10]The main inputs to carry out the microsimulations arethe estimates of the parameters that govern fertility deci-sionsoutcomes (for simplicity we refer to fertility decisionsalthough fertility outcomes could be the result of free con-scious choices but also the consequence of various other cir-cumstances) and the response of labor market participationto changes in family size We assume that the number ofchildren in a household follows a Poisson process and thatits parameters can be consistently estimated using a Poissonregression model Hourly wages and hours of work areassumed to be simultaneously determined in an equilibriummodel of the labor market
After estimating the parameters we carry out the sim-ulations That is we simulate the counterfactual incomedistribution that arises in a given base year by assumingthat the population in that year takes fertility decisions
according to the parameters estimated for a different yearTheresulting poverty and inequality measures over the simulateddistribution are compared to those actually observed in thebase year The difference between the simulated value ofan indicator of poverty or inequality and its actual value isinterpreted as a measure of the direct impact of the change infertility behavior that is the impact that would take place incase all other covariates remained fixed
The methodology requires keeping all other things con-stant when simulating the impact of changes in fertilitydecisions Naturally as it was stressed in previous sectionssome of these factors may be correlated with fertility Forinstance the structure ofwagesmay respond to changes in thelabor supply triggered by a change in fertility Additionallychanges in reproductive behavior may have been inducedby income changes in which case the microsimulation onlycaptures a single round of effects from fertility to incomes ofa more complicated process
In a given period income poverty and inequality areaffected by a host of factors including economic shocks andsocial policies The microsimulation methodology is usefulas it allows approximating the size of the impact that agiven factor would have in case all other factors remained
International Journal of Population Research 5
40
45
50
55
60
6519
9219
9419
9619
9820
0020
0220
0420
0620
0820
1020
12
40
45
50
55
60
65
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
40
45
50
55
60
65
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
40
45
50
55
60
65
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
40
45
50
55
60
65
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
40
45
50
55
60
65
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
40
45
50
55
60
65
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
Argentina Brazil Chile
El Salvador Mexico Peru
Uruguay
Figure 5 Income inequality Gini coefficient Source own calculations based onmicrodata fromnational household surveys Note inequalitycomputed over the distribution of household per capita income
fixed Although this is clearly an unrealistic scenario theestimations are still useful as a first approximation of theimpact especially considering the fact that the more ambi-tious alternative of estimating a general equilibriummodel isusually unfeasible
31 Microsimulations We carry out the simulation over thedistribution of household per capita income defined as
119910119894119905=
119884119871
ℎ119905+ 119884119873119871
ℎ119905
119873ℎ119905
forall119894 isin ℎ at time 119905 (1)
where 119894 indexes individuals ℎ households and 119905 time periods(years) 119884119871
ℎ119905denotes total labor income of household ℎ at time
119905 119884119873119871
ℎ119905labels nonlabor income and 119873
ℎ119905is the family size
which is the sum of the number of children up to 16 yearsold119867
ℎ119905and the rest of the household members 119877
ℎ119905
119873ℎ119905
= 119867ℎ119905+ 119877ℎ119905 (2)
Weneed to specify how119867ℎ119905is determined in order to simulate
the counterfactual distributions The literature argues that
fertility outcomes are the result of a process affected bycharacteristics of each spouse and of the household amongother factors This process can be formalized as
119867ℎ119905
= 119867 (119885ℎ119905 119890ℎ119905 120578119905) (3)
where 119885ℎ119905
is a vector of observable characteristics and119890ℎ119905
includes all unobservable characteristics that influencethe familyrsquos reproductive behavior while 120578
119905is the set of
parameters that govern fertility decisionsWe assume that labor income for individual 119894 is given by
119884119871
119894119905= 119908119894119905119871119894119905 (4)
where 119908119894119905
is the hourly wage rate and 119871119894119905
is the num-ber of hours worked by individual 119894 Wages and hoursof work are defined as 119908
119894119905= 119908(119883
1119894119905 120576119882
119894119905 120573119905) and 119871
119894119905=
119871(1198832119894119905 119867119894119905 120576119871
119894119905 120574119905 120582119905) respectively Where 119883
1119894119905and 119883
2119894119905are
observable characteristics 120576119882119894119905
and 120576119871
119894119905are unobservable char-
acteristics while 120573119905 120574119905 and 120582
119905are parameters In particular
120582119905is the parameter that relates the hours of work to the
number of children
6 International Journal of Population Research
Nonlabor income is usually assumed to be exogenous inthemicrosimulation literature However as stated in previoussections the expansion of social programs to families withchildren suggests a potential link between fertility decisionsand nonlabor income Accordingly we assume that nonlaborincome is given by
119884119873119871
ℎ119905= 119898ℎ119905+ 119892 (119867
ℎ119905) (5)
where119898ℎ119905is exogenous and119892(119867
ℎ119905) represents the component
of nonlabor income that depends on the number of childrenWe label the parameters of the fertility decisions at time 119905
as 120578119905and those at time 1199051015840 as 120578
1199051015840 A key step in themethodology
is to estimate the counterfactual number of children in agiven year 119905 if fertility outcomes were determined as in analternative year 119905
1015840 Once this term labeled as 119867ℎ119905(1205781199051015840) is
estimated three microsimulation exercises are carried out byreplacing this estimate in the household per capita incomeequation
The first exercise aims to capture the contribution of thechange in fertility parameters 120578 to the actual change in theincome distribution through the denominator of (1) Thesimulated per capita income is given by
119910119863
119894119905=
119884119871
ℎ119905+ 119884119873119871
ℎ119905
119873ℎ119905(1205781199051015840)
(6)
where 119873ℎ119905(1205781199051015840) = 119877
ℎ119905+ 119867ℎ119905(1205781199051015840) Superscript 119863 in the
simulated per capita income stands for the direct-size effectthat is the change in the income distribution due to changesin the number of household members among whom totalhousehold income should be distributed
The second exercise involves simulating the labor incomethat a household would have in year 119905 if parameters thatgovern fertility decisionswere those of year 1199051015840 In other wordsthe exercise consists in replacing the simulated number ofchildren 119867
ℎ119905(1205781199051015840) in the equation of individual labor income
(4) to obtain
119910119867
119894119905=
119884119871
ℎ119905(1205781199051015840) + 119884
119873119871
ℎ119905
119873ℎ119905
(7)
Superscript 119867 in the simulated per capita income stands forthe hours-size effect that is the contribution of the changein fertility parameters 120578 to the actual change in the incomedistribution through the indirect channel of affecting thehours-of-work decisions
We carry out a third exercise by simulating the coun-terfactual distribution arising from a change in nonlaborincome driven by changes in fertility decisions We refer tothe distributional impact of changes through this channelas the nonlabor-income effect and thus label the per capitaincome variable in (8) with the NL superscript
119910119873119871
119894119905=
119884119871
ℎ119905+ 119884119873119871
ℎ119905(1205781199051015840)
119873ℎ119905
(8)
Finally (9) considers the three channels simultaneously toobtain the total effect (superscript119879) of changes in the fertilityparameters
119910119879
119894119905=
119884119871
ℎ119905(1205781199051015840) + 119884
119873119871
ℎ119905(1205781199051015840)
119873ℎ119905(1205781199051015840)
(9)
We denote the per capita income distribution among indi-viduals in year 119905 as 119863
119905(119910119905) = (119910
1119905 1199102119905 119910
119899119905) where 119899
denotes the total number of individuals in the populationLet 119868(sdot) be a distributive indicator as the Gini coefficientor the poverty headcount ratio Then the contribution ofthe change in fertility parameters 120578 to the actual change inthe income distribution through the direct channel is givenby (10) The contributions through the other channels arecomputed analogously
119868 (119863119905(119910119905)) minus 119868 (119863
119905(119910119863
119905)) (10)
A shortcoming of the microsimulation methodology is thatdecomposition is path-dependent This means that takingyear 119905 or year 1199051015840 as the base year does not necessarily implythe same result Accordingly we perform the simulationsalternating the base year and report the average effect for eachexercise
In the following subsections we provide details concern-ing the specification and estimation of fertility labor incomeand nonlabor-income equations
32 Fertility Decisions We assume that the number of chil-dren in (3) follows a Poisson process with parameter 120583
ℎ119905
Formally
119867ℎ119905
sim Poisson (120583ℎ119905) with 120583
ℎ119905= 119864 (119867
ℎ119905| 119885ℎ119905) = exp (119885
1015840
ℎ119905120578119905) (11)
Then
Prob (119867ℎ119905
= 1198670) =
exp (minus120583ℎ119905) (120583ℎ119905)1198670
1198670
with 1198670= 0 1 2 (12)
Equation (12) is the Poisson regressionmodel fromwhich it ispossible to consistently estimate parameters 120578
119905by maximum
likelihood (it can be shown that consistency holds for themaximum likelihood estimators of 120578
119905as long as the real
distribution is any of the linear exponential family to whichthe Poisson distribution belongs provided that the condi-tional mean in (11) is correctly specified see eg Wooldridge[11]) We estimate separate models for two-parent householdand single-parent households For simplicity we denote theestimates by 120578
Once parameters 120578 are estimated we proceed to simulatechanges in the number of children Since the objective isto simulate these changes as a consequence of changes onlyin the parameters 120578 it is necessary to keep unobservablefactors fixed Therefore we characterize each household bythe quantile it occupies in the distribution of children ofyear 119905 Let 119865
120578119905|119885ℎ119905
(sdot) be the cumulative probability function ofa random variable that follows a Poisson distribution withparameter equal to exp(1198851015840
ℎ119905120578119905) and let 119902
ℎ119905be the quantile
for household ℎ at time 119905 that is 119865120578119905|119885ℎ119905
(119867ℎ119905) = 119902
ℎ119905 The
International Journal of Population Research 7
simulated number of children in household ℎ will be the onethat places it in the 119902
ℎ119905quantile of the distribution of children
with the relevant parameters of time 1199051015840(1205781199051015840) conditional to
the observable characteristics 119885ℎ119905 Formally the simulated
number of children in household ℎ at year 119905 using theestimated fertility parameters for year 1199051015840 is given by
119867ℎ119905(1205781199051015840) = 119865
minus1
1205781199051015840 |119885ℎ119905
∘ 119865120578119905|119885ℎ119905
(119867ℎ119905) (13)
33 Labor Incomes To estimate individual labor income in(4) we follow Gasparini et al [12] in assuming that bothwages and hours worked are determined in a reduced-formmodel of the labor market equilibrium
ln119908lowast
119894119905= 1198831015840
1119894119905120573119905+ 120576119882
119894119905 (14)
119871lowast
119894119905= 1198831015840
2119894119905120574119905+ 119867119894119905120582119905+ 120576119871
119894119905 (15)
with
if 119871lowast119894119905gt 0
119908119894119905= 119908lowast
119894119905
119871119894119905= 119871lowast
119894119905
if 119871lowast119894119905le 0
119908119894119905= 0
119871119894119905= 0
(120576119882
119894119905 120576119871
119894119905) sim 119873(0 0 120590
2
119882119905 1205902
119871119905 120588119905)
(16)
where119908lowast119894119905and119871lowast
119894119905are unobserved latent variables Vectors119883
1119894119905
and 1198832119894119905
include observable factors affecting hourly wagesand hours of work respectively while 119867
119894119905stands for the
number of children 120573119905 120574119905(vectors) and 120582
119905are parameters
to be estimated along with 1205902
119882119905 1205902119871119905 and 120588
119905
We estimate (14) by Heckmanrsquos maximum likelihoodmethod where a censored version of (15) is used as a selectionequation replacing hours of work by a binary variable thatindicates whether the individual works For estimation of(15) we use aTobitmodel (although this estimation strategy isnot fully efficient efficiency loss is not necessarily significantfor a given sample size and this alternative has certaincomputational advantages over a full information procedurefor more details see Gasparini et al [12]) We estimateseparate models for head of households and spouses
To simulate changes in individual labor incomes wereplace the observed number of children of individual 119894 atyear 119905 in (15) by the simulated number of children 119867
119905ℎ(1205781199051015840)
while assigning himher a wage based on (14) and an errorterm drawn from the bivariate distribution implicit in themodel
34 Nonlabor Incomes The measure of nonlabor incomeincludes pensions capital income and transfers Conditionalcash transfers are included as government transfers Wesimulate nonlabor incomes in an ad hoc manner taking
into account the characteristics of each countryrsquos socialprogram In those cases where it is not possible to identifythe recipients in the household survey we identify potentialbeneficiaries and estimate the amount of the transfers basedon the design of each program The general framework is asfollows nonlabor income of household ℎ at year 119905 is given by(5) where 119892(119867
ℎ119905) represents the part of nonlabor income that
depends on the number of children Typically cash transfersin poverty-alleviation programs vary with the number ofchildren in the family For example if the program has a fixedamount 120579
119905per child the nonlabor income becomes
119884119873119871
ℎ119905= 119898ℎ119905+ 120579119905119867ℎ119905 (17)
The simulated nonlabor income is calculated using thecounterfactual number of children in a given year 119905 if fertilityoutcomes were determined as in an alternative year 1199051015840
119884119873119871
ℎ119905(1205781199051015840) = 119898
ℎ119905+ 120579ℎ119905119867ℎ119905(1205781199051015840) (18)
Even though all programs have a similar setting to the onedescribed they differ in some aspects maximum number ofchildren age and in some cases how to identify the programrsquosrecipients The details are given in the Appendix We sim-ulate nonlabor incomes for Argentina Brazil Mexico andUruguay Even though the programs in Chile El Salvadorand Peru have similar characteristics it is not possible toestimate how the amount of transfers varies with the numberof children given the information in the household surveysFor these countries we assume that nonlabor income iscompletely exogenous
4 The Results
The results of the microsimulations regarding poverty arepresented inTable 2 Column (i) displays the observed changein the headcount ratio while the rest of the columns presentthe impact of the changes in fertility on poverty Column(v) shows the sum of all effects explored In order to betterunderstand the information in the table take the case ofBrazil as an example Column (i) shows that between 1990and 2012 the poverty headcount ratio in that country (usingthe line of US$ 4 a day) fell by around 25 points Thevalue in the last column has the following interpretationif fertility had been the only factor that changed duringthat period then the poverty headcount ratio in Brazilwould have fallen by 392 points This change is statisticallysignificant although not very big compared to the actual fallin poverty in that period The overall effect is the result ofa significant direct effect (minus374) and two smaller effects thatgo in different directions On the one hand the differentialreduction in fertility implied a greater increase in hours ofwork among more disadvantaged households which in turncontributed though very slightly to a further reduction inincome poverty On the other hand the falling patterns infertility among the poor ameliorated the poverty-decreasingimpact of the conditional cash transfer programs that aretargeted at families with children (mainly the Bolsa Familiain the 2000s) However this effect was quantitatively almostinsignificant
8 International Journal of Population Research
Table 2 Impact of fertility changes on poverty Changes in the poverty headcount ratio (US$ 4 line)
Country Period Observed change (i) EffectsDirect (ii) Hours (iii) NLI (iv) Total (v)
Argentina 1992ndash2012 minus1122 minus139 004 039 minus080(008) (002) (001) (001) (002)
Brazil 1990ndash2012 minus2516 minus374 minus017 014 minus392(003) (001) (000) (000) (001)
Chile 1990ndash2011 minus3324 minus156 018 mdash minus143(005) (001) (000) mdash (001)
El Salvador 1991ndash2010 minus1412 minus498 minus034 mdash minus531(006) (004) (001) mdash (005)
Mexico 1992ndash2012 minus350 minus435 minus030 025 minus443(010) (004) (002) (001) (004)
Peru 1997ndash2012 minus2320 minus383 minus019 mdash minus394(009) (003) (001) mdash (003)
Uruguay 1995ndash2012 minus156 minus008 000 002 minus005(003) (002) (000) (000) (002)
Source own calculations based on microdata from national household surveysNote All effects are significant at the 1 level The standard errors were calculated using bootstrap with 200 replications The values of each effect are averagesthat result from taking alternatively each year in the comparison as the base year The sample includes only households in which the head is between 25 and45 years old NLI nonlabor income
Table 3 Impact of fertility changes on inequality Changes in the Gini coefficient
Country Period Observed change (i) EffectsDirect (ii) Hours (iii) NLI (iv) Total (v)
Argentina 1992ndash2012 minus544 minus104 minus001 013 minus095(010) (001) (000) (000) (002)
Brazil 1990ndash2012 minus718 minus130 minus006 006 minus132(004) (000) (000) (000) (000)
Chile 1990ndash2011 minus178 minus075 004 mdash minus073(008) (001) (000) mdash (001)
El Salvador 1991ndash2010 minus768 minus130 minus007 mdash minus140(006) (002) (000) mdash (002)
Mexico 1992ndash2012 minus086 minus168 minus011 012 minus169(019) (002) (000) (000) (002)
Peru 1997ndash2012 minus732 minus192 minus006 mdash minus199(013) (001) (000) mdash (001)
Uruguay 1995ndash2012 minus203 minus012 minus001 002 minus013(004) (001) (000) (000) (001)
Source own calculations based on microdata from national household surveysNote All effects are significant at the 1 level The standard errors were calculated using bootstrap with 200 replications The values of each effect are averagesthat result from taking alternatively each year in the comparison as the base year The sample includes only households in which the head is between 25 and45 years old NLI nonlabor income
Changes in fertility patterns over the last two decades inLatin America have implied a reduction in income povertyThe impact is statistically significant and in some caseseconomically large For instance the estimated poverty-reduction effect was more than 4 points in Mexico andEl Salvador Most of the effect comes through the directeffect a reduction in fertility rates among the most dis-advantaged groups reduced family size and increased percapita income The hours-of-work effect is in most casespoverty reducing and the nonlabor-income effect is poverty
increasing but in both cases the estimated sizes aresmall
Table 3 shows the results of the simulations on incomeinequality measured by the Gini coefficient Inequality inBrazil as measured by the Gini coefficient fell by 718points between 1990 and 2012 If fertility had been the onlyfactor that changed in that period then the Gini coefficientwould have fallen by 132 points (column v) This change isstatistically significant representing around 18 of the actualreduction in inequality during that period The differential
International Journal of Population Research 9
reduction in fertility across socioeconomic groups in Brazilcontributed to the observed decline in inequality in the lasttwo decadesThe overall effect ofminus132 points in column (v) isthe result of a significant direct effect (minus130) and two smallereffects going in different directions
Some interesting general results emerge from Table 3First changes in fertility have implied a decline in incomeinequality The differential pattern in fertility across groupsexperienced in most Latin American countries over thelast decades translated into an equalizing impact on theincome distribution Second this effect is small althoughnot negligible To be sure demographic changes are not thecentral reason behind changes in income inequality but theyare statistically significant and economically relevant Onaverage fertility changes account for a fall of around onepoint in the Gini coefficient
Third most of the impact comes from the direct effectThe differential fall in fertility rates among socioeconomicgroups had a larger impact on the family size of poorerfamilies implying a proportionally larger increasing effect ontheir per capita incomes
Fourth the effect of fertility changes on hours of workand in turn on incomes is smaller and has different signsacross countries It is important to notice that a more intensereduction in fertility among the poor may be consistentwith a positive sign (inequality-increasing) for this effectThis could happen if the elasticity of hours of work withrespect to the number of children is higher among thenonpoor and if the extra hours worked triggered by the fallin fertility imply a household income increase proportionallylarger for the nonpoor Anyway with one exception thehours-of-work effect has a negative sign implying that theasymmetric reduction in fertility contributed to a reductionin income inequality by fostering higher participation in thelabor market among poor parents (mostly mothers)
Finally the effect through nonlabor incomes is alwaysinequality-increasing but small The fall in fertility amongthe poor implied a lower inequality-reducing impact of theconditional cash transfer programs that were introduced andexpanded in the region mainly during the 2000s The effecthowever is in most cases very small
5 Concluding Remarks
Poverty and inequality fell in Latin America over the lastdecades driven by a large number of factors In thispaper we highlight one factor that is usually overlooked inthe economic literature the impact of changes in fertilityThe average number of children fell in households fromall groups but especially among the more vulnerable afact that could have distributive implications We quantifythe consequences of this demographic pattern by applyingmicrosimulation techniquesThe results suggest that changesin fertility that took place over the last decades in LatinAmerica contributed to the reduction in income inequalityas well as in poverty The impacts found were in generalstatistically significant and economically relevant althoughnot very large Although surely there are more importantdeterminants behind the observed patterns in the income
distributions of the Latin American countries demographicfactors should not be overlooked
The paper has focused on three simple yet central chan-nels the change in the number of children (i) enlarges thefamily size and reduces per capita income (ii) affects thelabor supply decisions and (iii) modifies the income supportfrom social programs Certainly there are other potentiallyrelevant channels that are ignored in the paper For instancea reduction in the number of children per family couldimply higher inheritances and higher family spending oneducation and health per child and hence better incomeperspectives Also smaller families may be better positionedto take advantage of labor market opportunities for instancethrough migration Moreover lower fertility rates may beassociated with a reduction in the relative supply of unskilledlabor and hence an increase in its relative wage
Although in the short and medium run lower fertility inmore disadvantaged households may contribute to reducingpoverty and inequality in the long run the implicationsare more nuanced When children become adults they maycontribute to their parentsrsquo incomes a fall in fertility mayreduce incomes for the elderly especially in countries wherethe pension system is weak and ultimately contribute tohigher poverty through that intertemporal channel
Appendix
In this appendix we describe the issues regarding the simu-lations of the nonlabor incomes for each country
A Argentina Asignacioacuten Universal porHijo (AUH)
Argentina launched the conditional cash transfer programAUH in 2009 The beneficiaries are children under the ageof 18 in households whose members are unemployed orinformal workers The program provides a monthly cashsubsidy per child up to 5 children Since Argentinarsquos house-hold survey (EPH) does not include a question to identifyrecipients of this program we follow Garganta and Gasparini[13] assuming full take-up among those who qualify Thesimulation is made according to the following rule
If119867ℎ119905(1205781199051015840) + 119876
ℎ119905le 5 and119867
ℎ119905+ 119876ℎ119905
le 5
119884119873119871
ℎ119905(1205781199051015840) = 119884
119873119871
ℎ119905+ 120579119905(119867ℎ119905(1205781199051015840) minus 119867
ℎ119905) (A1)
If119867ℎ119905(1205781199051015840) + 119876
ℎ119905gt 5 and119867
ℎ119905+ 119876ℎ119905
le 5
119884119873119871
ℎ119905(1205781199051015840) = 119884
119873119871
ℎ119905+ 120579119905(5 minus 119867
ℎ119905minus 119876ℎ119905) (A2)
If119867ℎ119905(1205781199051015840) + 119876
ℎ119905le 5 and119867
ℎ119905+ 119876ℎ119905
gt 5
119884119873119871
ℎ119905(1205781199051015840) = 119884
119873119871
ℎ119905+ 120579119905(119867ℎ119905(1205781199051015840) + 119876
ℎ119905minus 5) (A3)
If119867ℎ119905(1205781199051015840) + 119876
ℎ119905gt 5 and119867
ℎ119905+ 119876ℎ119905
gt 5
119884119873119871
ℎ119905(1205781199051015840) = 119884
119873119871
ℎ119905 (A4)
10 International Journal of Population Research
where 119876ℎ119905
is the number of children aged 17 or 18(recall that we carry out the simulations for childrenaged 16 or younger) and 120579
119905is equal to AR$270 for
the third trimester of 2012 and AR$340 for the fourthtrimester
B Brazil Bolsa Familia Program (BFP)
The Bolsa Familia implemented in 2003 is the flagshipconditional cash transfer program in Brazil aimed at poorhouseholds The estimated nonlabor incomes in this case arecalculated as follows
If119867119905ℎ
gt 0 and beneficiary
119884119873119871
ℎ119905(1205781199051015840) = 119884
119873119871
119905ℎ+ 119872119905ℎ(119867119905ℎ(1205781199051015840)
119867119905ℎ
minus 1) (B1)
If119867119905ℎ
= 0119867119905ℎ(1205781199051015840) = 119896 and beneficiary
119884119873119871
ℎ119905(1205781199051015840) = 119884
119873119871
119905ℎ+ 119872119896
119905ℎminus 119897119905ℎ (B2)
with 119896 = 1 2 3 119872119905ℎ
being the amount of thetransfer119872119896
119905ℎbeing the average benefit of households
with 119896 children and 119897119905ℎbeing the standard benefit of
recipients without children
C Mexico Oportunidades
Oportunidades is the main antipoverty program of theMexican Government The size of the transfer depends onthe number of children among other characteristics of thehouseholdMexicorsquos ENIGH identifies the beneficiaries of theprogram and the amount of the subsidy that each householdreceives The simulation is made as follows
119884119873119871
ℎ119905(1205781199051015840) = 119884
119873119871
119905ℎ+ 119872119896
119905ℎminus 119872119905ℎ
if 119867119905ℎ(1205781199051015840) = 119896 119867
119905ℎ= 119896 and beneficiary
(C1)
where 119872119896119905ℎis the average transfer of a beneficiary household
with 119896 children and119872119905ℎis the original household transfer
D Uruguay Asignaciones Familiares
The Asignaciones Familiares program in Uruguay is targetedat the children of workers The amount of the transfer varieswith the householdrsquos income Since the size of the transferdepends on characteristics of the child such as the age thesimulation is carried out as follows
If119867119905ℎ
gt 0 and beneficiary
119884119873119871
ℎ119905(1205781199051015840) = 119884
119873119871
119905ℎ+ 119872119905ℎ(119867119905ℎ(1205781199051015840)
119867119905ℎ
minus 1) (D1)
If119867119905ℎ
= 0119867119905ℎ(1205781199051015840) = 119896 and beneficiary
119884119873119871
ℎ119905(1205781199051015840) = 119884
119873119871
119905ℎ+ 119872119896
119905ℎ (D2)
with 119896 = 1 2 3 119872119905ℎ
being the total amount ofthe transfer and 119872
119896
119905ℎbeing the average transfer of a
household with 119896 children
Disclosure
The information presented herein is solely the responsibilityof the authors and does not necessarily reflect the opinions ofthe aforementioned entities
Competing Interests
The authors declare that they have no competing interests
Acknowledgments
This paper was written in the framework of the projectEnhancing Womenrsquos Economic Empowerment throughBetter Policies in Latin America a joint initiative betweenCEDLAS-UNLP and CIEDUR with the support ofIDRC Canada The paper includes material from NicolasBadaraccorsquos Masters dissertation at Universidad Nacionalde La Plata The authors are grateful to Carolina RobinoGuillermo Cruces Jorge Paz Facundo Albornoz andseminar participants at AAEP (Posadas 2014) Network ofInequality and Poverty (UNGS 2014) and Conference onSocial Protection (IDRC and IPC Brasilia 2014) for valuablecomments and suggestions
References
[1] T W Guinnane ldquoThe historical fertility transition a guide foreconomistsrdquo Journal of Economic Literature vol 49 no 3 pp589ndash614 2011
[2] L E Jones and M Tertilt ldquoAn economic history of fertility inthe US 1826ndash1960rdquo NBERWorking Paper 12796 2006
[3] J Chackiel La Dinamica Demografica en America LatinaSerie Poblacion y Desarrollo No 52 Centro Latinoamericanoy Caribeno de Demografıa-Division de Poblacion de CEPAL2004
[4] ECLAC ldquoLa fecundidad en America Latina un descensoacelerado y heterogeneo con profundas transformacionesdemograficas y socialesrdquo in Fertility Latin American and theCaribbean Demographic Observatory no 5 Centro Latinoa-mericano y Caribeno de DemografıamdashDivision de Poblacionde CEPAL 2008
[5] ECLAC ldquoChapter II current situation and outlook for fertilityin Latin Americardquo in Social Panorama of Latin America 2011ECLAC 2011
[6] L Gasparini and M Marchionni Bridging Gender Gaps TheRise and Deceleration of Female Labor Force ParticipationCEDLAS-UNLP 2015
[7] L Gasparini and N Lustig ldquoThe rise and fall of incomeinequality in Latin Americardquo inThe Oxford Handbook of LatinAmerican Economics J Ocampo and J Ros Eds chapter 27 pp691ndash714 Oxford University Press Oxford UK 2011
[8] G Cornia Ed Falling Inequality in Latin America PolicyChanges and Lessons Oxford University Press 2014
International Journal of Population Research 11
[9] F Alvaredo and L Gasparini ldquoRecent trends in inequalityand poverty in developing countriesrdquo in Handbook of IncomeDistribution Atkinson and Bourguignon Eds vol 2 chapter 9pp 697ndash805 Elsevier New York NY USA 2015
[10] M Marchionni and L Gasparini ldquoTracing out the effects ofdemographic changes on the income distribution The caseof Greater Buenos Aires 1980ndash1998rdquo Journal of EconomicInequality vol 5 no 1 pp 97ndash114 2007
[11] J M Wooldridge Econometric Analysis of Cross Section andPanel Data MIT Press Boston Mass USA 2002
[12] L Gasparini M Marchionni and W Sosa Escudero ldquoChar-acterization of inequality changes through microeconometricdecompositions The case of Greater Buenos Airesrdquo in TheMicroeconomics of Income Distribution Dynamics in East Asiaand Latina America F Bourguignon F Ferreira and N LustigEds Oxford University Press New York NY USA 2004
[13] S Garganta and L Gasparini ldquoThe impact of a social programon labor informality the case of AUH in Argentinardquo Journal ofDevelopment Economics vol 115 pp 99ndash110 2015
Submit your manuscripts athttpwwwhindawicom
Child Development Research
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Education Research International
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Biomedical EducationJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Psychiatry Journal
ArchaeologyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
AnthropologyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Research and TreatmentSchizophrenia
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Urban Studies Research
Population ResearchInternational Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CriminologyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Aging ResearchJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
NursingResearch and Practice
Current Gerontologyamp Geriatrics Research
Hindawi Publishing Corporationhttpwwwhindawicom
Volume 2014
Sleep DisordersHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
AddictionJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Depression Research and TreatmentHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Geography Journal
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Research and TreatmentAutism
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Economics Research International
International Journal of Population Research 5
40
45
50
55
60
6519
9219
9419
9619
9820
0020
0220
0420
0620
0820
1020
12
40
45
50
55
60
65
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
40
45
50
55
60
65
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
40
45
50
55
60
65
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
40
45
50
55
60
65
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
40
45
50
55
60
65
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
40
45
50
55
60
65
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
Argentina Brazil Chile
El Salvador Mexico Peru
Uruguay
Figure 5 Income inequality Gini coefficient Source own calculations based onmicrodata fromnational household surveys Note inequalitycomputed over the distribution of household per capita income
fixed Although this is clearly an unrealistic scenario theestimations are still useful as a first approximation of theimpact especially considering the fact that the more ambi-tious alternative of estimating a general equilibriummodel isusually unfeasible
31 Microsimulations We carry out the simulation over thedistribution of household per capita income defined as
119910119894119905=
119884119871
ℎ119905+ 119884119873119871
ℎ119905
119873ℎ119905
forall119894 isin ℎ at time 119905 (1)
where 119894 indexes individuals ℎ households and 119905 time periods(years) 119884119871
ℎ119905denotes total labor income of household ℎ at time
119905 119884119873119871
ℎ119905labels nonlabor income and 119873
ℎ119905is the family size
which is the sum of the number of children up to 16 yearsold119867
ℎ119905and the rest of the household members 119877
ℎ119905
119873ℎ119905
= 119867ℎ119905+ 119877ℎ119905 (2)
Weneed to specify how119867ℎ119905is determined in order to simulate
the counterfactual distributions The literature argues that
fertility outcomes are the result of a process affected bycharacteristics of each spouse and of the household amongother factors This process can be formalized as
119867ℎ119905
= 119867 (119885ℎ119905 119890ℎ119905 120578119905) (3)
where 119885ℎ119905
is a vector of observable characteristics and119890ℎ119905
includes all unobservable characteristics that influencethe familyrsquos reproductive behavior while 120578
119905is the set of
parameters that govern fertility decisionsWe assume that labor income for individual 119894 is given by
119884119871
119894119905= 119908119894119905119871119894119905 (4)
where 119908119894119905
is the hourly wage rate and 119871119894119905
is the num-ber of hours worked by individual 119894 Wages and hoursof work are defined as 119908
119894119905= 119908(119883
1119894119905 120576119882
119894119905 120573119905) and 119871
119894119905=
119871(1198832119894119905 119867119894119905 120576119871
119894119905 120574119905 120582119905) respectively Where 119883
1119894119905and 119883
2119894119905are
observable characteristics 120576119882119894119905
and 120576119871
119894119905are unobservable char-
acteristics while 120573119905 120574119905 and 120582
119905are parameters In particular
120582119905is the parameter that relates the hours of work to the
number of children
6 International Journal of Population Research
Nonlabor income is usually assumed to be exogenous inthemicrosimulation literature However as stated in previoussections the expansion of social programs to families withchildren suggests a potential link between fertility decisionsand nonlabor income Accordingly we assume that nonlaborincome is given by
119884119873119871
ℎ119905= 119898ℎ119905+ 119892 (119867
ℎ119905) (5)
where119898ℎ119905is exogenous and119892(119867
ℎ119905) represents the component
of nonlabor income that depends on the number of childrenWe label the parameters of the fertility decisions at time 119905
as 120578119905and those at time 1199051015840 as 120578
1199051015840 A key step in themethodology
is to estimate the counterfactual number of children in agiven year 119905 if fertility outcomes were determined as in analternative year 119905
1015840 Once this term labeled as 119867ℎ119905(1205781199051015840) is
estimated three microsimulation exercises are carried out byreplacing this estimate in the household per capita incomeequation
The first exercise aims to capture the contribution of thechange in fertility parameters 120578 to the actual change in theincome distribution through the denominator of (1) Thesimulated per capita income is given by
119910119863
119894119905=
119884119871
ℎ119905+ 119884119873119871
ℎ119905
119873ℎ119905(1205781199051015840)
(6)
where 119873ℎ119905(1205781199051015840) = 119877
ℎ119905+ 119867ℎ119905(1205781199051015840) Superscript 119863 in the
simulated per capita income stands for the direct-size effectthat is the change in the income distribution due to changesin the number of household members among whom totalhousehold income should be distributed
The second exercise involves simulating the labor incomethat a household would have in year 119905 if parameters thatgovern fertility decisionswere those of year 1199051015840 In other wordsthe exercise consists in replacing the simulated number ofchildren 119867
ℎ119905(1205781199051015840) in the equation of individual labor income
(4) to obtain
119910119867
119894119905=
119884119871
ℎ119905(1205781199051015840) + 119884
119873119871
ℎ119905
119873ℎ119905
(7)
Superscript 119867 in the simulated per capita income stands forthe hours-size effect that is the contribution of the changein fertility parameters 120578 to the actual change in the incomedistribution through the indirect channel of affecting thehours-of-work decisions
We carry out a third exercise by simulating the coun-terfactual distribution arising from a change in nonlaborincome driven by changes in fertility decisions We refer tothe distributional impact of changes through this channelas the nonlabor-income effect and thus label the per capitaincome variable in (8) with the NL superscript
119910119873119871
119894119905=
119884119871
ℎ119905+ 119884119873119871
ℎ119905(1205781199051015840)
119873ℎ119905
(8)
Finally (9) considers the three channels simultaneously toobtain the total effect (superscript119879) of changes in the fertilityparameters
119910119879
119894119905=
119884119871
ℎ119905(1205781199051015840) + 119884
119873119871
ℎ119905(1205781199051015840)
119873ℎ119905(1205781199051015840)
(9)
We denote the per capita income distribution among indi-viduals in year 119905 as 119863
119905(119910119905) = (119910
1119905 1199102119905 119910
119899119905) where 119899
denotes the total number of individuals in the populationLet 119868(sdot) be a distributive indicator as the Gini coefficientor the poverty headcount ratio Then the contribution ofthe change in fertility parameters 120578 to the actual change inthe income distribution through the direct channel is givenby (10) The contributions through the other channels arecomputed analogously
119868 (119863119905(119910119905)) minus 119868 (119863
119905(119910119863
119905)) (10)
A shortcoming of the microsimulation methodology is thatdecomposition is path-dependent This means that takingyear 119905 or year 1199051015840 as the base year does not necessarily implythe same result Accordingly we perform the simulationsalternating the base year and report the average effect for eachexercise
In the following subsections we provide details concern-ing the specification and estimation of fertility labor incomeand nonlabor-income equations
32 Fertility Decisions We assume that the number of chil-dren in (3) follows a Poisson process with parameter 120583
ℎ119905
Formally
119867ℎ119905
sim Poisson (120583ℎ119905) with 120583
ℎ119905= 119864 (119867
ℎ119905| 119885ℎ119905) = exp (119885
1015840
ℎ119905120578119905) (11)
Then
Prob (119867ℎ119905
= 1198670) =
exp (minus120583ℎ119905) (120583ℎ119905)1198670
1198670
with 1198670= 0 1 2 (12)
Equation (12) is the Poisson regressionmodel fromwhich it ispossible to consistently estimate parameters 120578
119905by maximum
likelihood (it can be shown that consistency holds for themaximum likelihood estimators of 120578
119905as long as the real
distribution is any of the linear exponential family to whichthe Poisson distribution belongs provided that the condi-tional mean in (11) is correctly specified see eg Wooldridge[11]) We estimate separate models for two-parent householdand single-parent households For simplicity we denote theestimates by 120578
Once parameters 120578 are estimated we proceed to simulatechanges in the number of children Since the objective isto simulate these changes as a consequence of changes onlyin the parameters 120578 it is necessary to keep unobservablefactors fixed Therefore we characterize each household bythe quantile it occupies in the distribution of children ofyear 119905 Let 119865
120578119905|119885ℎ119905
(sdot) be the cumulative probability function ofa random variable that follows a Poisson distribution withparameter equal to exp(1198851015840
ℎ119905120578119905) and let 119902
ℎ119905be the quantile
for household ℎ at time 119905 that is 119865120578119905|119885ℎ119905
(119867ℎ119905) = 119902
ℎ119905 The
International Journal of Population Research 7
simulated number of children in household ℎ will be the onethat places it in the 119902
ℎ119905quantile of the distribution of children
with the relevant parameters of time 1199051015840(1205781199051015840) conditional to
the observable characteristics 119885ℎ119905 Formally the simulated
number of children in household ℎ at year 119905 using theestimated fertility parameters for year 1199051015840 is given by
119867ℎ119905(1205781199051015840) = 119865
minus1
1205781199051015840 |119885ℎ119905
∘ 119865120578119905|119885ℎ119905
(119867ℎ119905) (13)
33 Labor Incomes To estimate individual labor income in(4) we follow Gasparini et al [12] in assuming that bothwages and hours worked are determined in a reduced-formmodel of the labor market equilibrium
ln119908lowast
119894119905= 1198831015840
1119894119905120573119905+ 120576119882
119894119905 (14)
119871lowast
119894119905= 1198831015840
2119894119905120574119905+ 119867119894119905120582119905+ 120576119871
119894119905 (15)
with
if 119871lowast119894119905gt 0
119908119894119905= 119908lowast
119894119905
119871119894119905= 119871lowast
119894119905
if 119871lowast119894119905le 0
119908119894119905= 0
119871119894119905= 0
(120576119882
119894119905 120576119871
119894119905) sim 119873(0 0 120590
2
119882119905 1205902
119871119905 120588119905)
(16)
where119908lowast119894119905and119871lowast
119894119905are unobserved latent variables Vectors119883
1119894119905
and 1198832119894119905
include observable factors affecting hourly wagesand hours of work respectively while 119867
119894119905stands for the
number of children 120573119905 120574119905(vectors) and 120582
119905are parameters
to be estimated along with 1205902
119882119905 1205902119871119905 and 120588
119905
We estimate (14) by Heckmanrsquos maximum likelihoodmethod where a censored version of (15) is used as a selectionequation replacing hours of work by a binary variable thatindicates whether the individual works For estimation of(15) we use aTobitmodel (although this estimation strategy isnot fully efficient efficiency loss is not necessarily significantfor a given sample size and this alternative has certaincomputational advantages over a full information procedurefor more details see Gasparini et al [12]) We estimateseparate models for head of households and spouses
To simulate changes in individual labor incomes wereplace the observed number of children of individual 119894 atyear 119905 in (15) by the simulated number of children 119867
119905ℎ(1205781199051015840)
while assigning himher a wage based on (14) and an errorterm drawn from the bivariate distribution implicit in themodel
34 Nonlabor Incomes The measure of nonlabor incomeincludes pensions capital income and transfers Conditionalcash transfers are included as government transfers Wesimulate nonlabor incomes in an ad hoc manner taking
into account the characteristics of each countryrsquos socialprogram In those cases where it is not possible to identifythe recipients in the household survey we identify potentialbeneficiaries and estimate the amount of the transfers basedon the design of each program The general framework is asfollows nonlabor income of household ℎ at year 119905 is given by(5) where 119892(119867
ℎ119905) represents the part of nonlabor income that
depends on the number of children Typically cash transfersin poverty-alleviation programs vary with the number ofchildren in the family For example if the program has a fixedamount 120579
119905per child the nonlabor income becomes
119884119873119871
ℎ119905= 119898ℎ119905+ 120579119905119867ℎ119905 (17)
The simulated nonlabor income is calculated using thecounterfactual number of children in a given year 119905 if fertilityoutcomes were determined as in an alternative year 1199051015840
119884119873119871
ℎ119905(1205781199051015840) = 119898
ℎ119905+ 120579ℎ119905119867ℎ119905(1205781199051015840) (18)
Even though all programs have a similar setting to the onedescribed they differ in some aspects maximum number ofchildren age and in some cases how to identify the programrsquosrecipients The details are given in the Appendix We sim-ulate nonlabor incomes for Argentina Brazil Mexico andUruguay Even though the programs in Chile El Salvadorand Peru have similar characteristics it is not possible toestimate how the amount of transfers varies with the numberof children given the information in the household surveysFor these countries we assume that nonlabor income iscompletely exogenous
4 The Results
The results of the microsimulations regarding poverty arepresented inTable 2 Column (i) displays the observed changein the headcount ratio while the rest of the columns presentthe impact of the changes in fertility on poverty Column(v) shows the sum of all effects explored In order to betterunderstand the information in the table take the case ofBrazil as an example Column (i) shows that between 1990and 2012 the poverty headcount ratio in that country (usingthe line of US$ 4 a day) fell by around 25 points Thevalue in the last column has the following interpretationif fertility had been the only factor that changed duringthat period then the poverty headcount ratio in Brazilwould have fallen by 392 points This change is statisticallysignificant although not very big compared to the actual fallin poverty in that period The overall effect is the result ofa significant direct effect (minus374) and two smaller effects thatgo in different directions On the one hand the differentialreduction in fertility implied a greater increase in hours ofwork among more disadvantaged households which in turncontributed though very slightly to a further reduction inincome poverty On the other hand the falling patterns infertility among the poor ameliorated the poverty-decreasingimpact of the conditional cash transfer programs that aretargeted at families with children (mainly the Bolsa Familiain the 2000s) However this effect was quantitatively almostinsignificant
8 International Journal of Population Research
Table 2 Impact of fertility changes on poverty Changes in the poverty headcount ratio (US$ 4 line)
Country Period Observed change (i) EffectsDirect (ii) Hours (iii) NLI (iv) Total (v)
Argentina 1992ndash2012 minus1122 minus139 004 039 minus080(008) (002) (001) (001) (002)
Brazil 1990ndash2012 minus2516 minus374 minus017 014 minus392(003) (001) (000) (000) (001)
Chile 1990ndash2011 minus3324 minus156 018 mdash minus143(005) (001) (000) mdash (001)
El Salvador 1991ndash2010 minus1412 minus498 minus034 mdash minus531(006) (004) (001) mdash (005)
Mexico 1992ndash2012 minus350 minus435 minus030 025 minus443(010) (004) (002) (001) (004)
Peru 1997ndash2012 minus2320 minus383 minus019 mdash minus394(009) (003) (001) mdash (003)
Uruguay 1995ndash2012 minus156 minus008 000 002 minus005(003) (002) (000) (000) (002)
Source own calculations based on microdata from national household surveysNote All effects are significant at the 1 level The standard errors were calculated using bootstrap with 200 replications The values of each effect are averagesthat result from taking alternatively each year in the comparison as the base year The sample includes only households in which the head is between 25 and45 years old NLI nonlabor income
Table 3 Impact of fertility changes on inequality Changes in the Gini coefficient
Country Period Observed change (i) EffectsDirect (ii) Hours (iii) NLI (iv) Total (v)
Argentina 1992ndash2012 minus544 minus104 minus001 013 minus095(010) (001) (000) (000) (002)
Brazil 1990ndash2012 minus718 minus130 minus006 006 minus132(004) (000) (000) (000) (000)
Chile 1990ndash2011 minus178 minus075 004 mdash minus073(008) (001) (000) mdash (001)
El Salvador 1991ndash2010 minus768 minus130 minus007 mdash minus140(006) (002) (000) mdash (002)
Mexico 1992ndash2012 minus086 minus168 minus011 012 minus169(019) (002) (000) (000) (002)
Peru 1997ndash2012 minus732 minus192 minus006 mdash minus199(013) (001) (000) mdash (001)
Uruguay 1995ndash2012 minus203 minus012 minus001 002 minus013(004) (001) (000) (000) (001)
Source own calculations based on microdata from national household surveysNote All effects are significant at the 1 level The standard errors were calculated using bootstrap with 200 replications The values of each effect are averagesthat result from taking alternatively each year in the comparison as the base year The sample includes only households in which the head is between 25 and45 years old NLI nonlabor income
Changes in fertility patterns over the last two decades inLatin America have implied a reduction in income povertyThe impact is statistically significant and in some caseseconomically large For instance the estimated poverty-reduction effect was more than 4 points in Mexico andEl Salvador Most of the effect comes through the directeffect a reduction in fertility rates among the most dis-advantaged groups reduced family size and increased percapita income The hours-of-work effect is in most casespoverty reducing and the nonlabor-income effect is poverty
increasing but in both cases the estimated sizes aresmall
Table 3 shows the results of the simulations on incomeinequality measured by the Gini coefficient Inequality inBrazil as measured by the Gini coefficient fell by 718points between 1990 and 2012 If fertility had been the onlyfactor that changed in that period then the Gini coefficientwould have fallen by 132 points (column v) This change isstatistically significant representing around 18 of the actualreduction in inequality during that period The differential
International Journal of Population Research 9
reduction in fertility across socioeconomic groups in Brazilcontributed to the observed decline in inequality in the lasttwo decadesThe overall effect ofminus132 points in column (v) isthe result of a significant direct effect (minus130) and two smallereffects going in different directions
Some interesting general results emerge from Table 3First changes in fertility have implied a decline in incomeinequality The differential pattern in fertility across groupsexperienced in most Latin American countries over thelast decades translated into an equalizing impact on theincome distribution Second this effect is small althoughnot negligible To be sure demographic changes are not thecentral reason behind changes in income inequality but theyare statistically significant and economically relevant Onaverage fertility changes account for a fall of around onepoint in the Gini coefficient
Third most of the impact comes from the direct effectThe differential fall in fertility rates among socioeconomicgroups had a larger impact on the family size of poorerfamilies implying a proportionally larger increasing effect ontheir per capita incomes
Fourth the effect of fertility changes on hours of workand in turn on incomes is smaller and has different signsacross countries It is important to notice that a more intensereduction in fertility among the poor may be consistentwith a positive sign (inequality-increasing) for this effectThis could happen if the elasticity of hours of work withrespect to the number of children is higher among thenonpoor and if the extra hours worked triggered by the fallin fertility imply a household income increase proportionallylarger for the nonpoor Anyway with one exception thehours-of-work effect has a negative sign implying that theasymmetric reduction in fertility contributed to a reductionin income inequality by fostering higher participation in thelabor market among poor parents (mostly mothers)
Finally the effect through nonlabor incomes is alwaysinequality-increasing but small The fall in fertility amongthe poor implied a lower inequality-reducing impact of theconditional cash transfer programs that were introduced andexpanded in the region mainly during the 2000s The effecthowever is in most cases very small
5 Concluding Remarks
Poverty and inequality fell in Latin America over the lastdecades driven by a large number of factors In thispaper we highlight one factor that is usually overlooked inthe economic literature the impact of changes in fertilityThe average number of children fell in households fromall groups but especially among the more vulnerable afact that could have distributive implications We quantifythe consequences of this demographic pattern by applyingmicrosimulation techniquesThe results suggest that changesin fertility that took place over the last decades in LatinAmerica contributed to the reduction in income inequalityas well as in poverty The impacts found were in generalstatistically significant and economically relevant althoughnot very large Although surely there are more importantdeterminants behind the observed patterns in the income
distributions of the Latin American countries demographicfactors should not be overlooked
The paper has focused on three simple yet central chan-nels the change in the number of children (i) enlarges thefamily size and reduces per capita income (ii) affects thelabor supply decisions and (iii) modifies the income supportfrom social programs Certainly there are other potentiallyrelevant channels that are ignored in the paper For instancea reduction in the number of children per family couldimply higher inheritances and higher family spending oneducation and health per child and hence better incomeperspectives Also smaller families may be better positionedto take advantage of labor market opportunities for instancethrough migration Moreover lower fertility rates may beassociated with a reduction in the relative supply of unskilledlabor and hence an increase in its relative wage
Although in the short and medium run lower fertility inmore disadvantaged households may contribute to reducingpoverty and inequality in the long run the implicationsare more nuanced When children become adults they maycontribute to their parentsrsquo incomes a fall in fertility mayreduce incomes for the elderly especially in countries wherethe pension system is weak and ultimately contribute tohigher poverty through that intertemporal channel
Appendix
In this appendix we describe the issues regarding the simu-lations of the nonlabor incomes for each country
A Argentina Asignacioacuten Universal porHijo (AUH)
Argentina launched the conditional cash transfer programAUH in 2009 The beneficiaries are children under the ageof 18 in households whose members are unemployed orinformal workers The program provides a monthly cashsubsidy per child up to 5 children Since Argentinarsquos house-hold survey (EPH) does not include a question to identifyrecipients of this program we follow Garganta and Gasparini[13] assuming full take-up among those who qualify Thesimulation is made according to the following rule
If119867ℎ119905(1205781199051015840) + 119876
ℎ119905le 5 and119867
ℎ119905+ 119876ℎ119905
le 5
119884119873119871
ℎ119905(1205781199051015840) = 119884
119873119871
ℎ119905+ 120579119905(119867ℎ119905(1205781199051015840) minus 119867
ℎ119905) (A1)
If119867ℎ119905(1205781199051015840) + 119876
ℎ119905gt 5 and119867
ℎ119905+ 119876ℎ119905
le 5
119884119873119871
ℎ119905(1205781199051015840) = 119884
119873119871
ℎ119905+ 120579119905(5 minus 119867
ℎ119905minus 119876ℎ119905) (A2)
If119867ℎ119905(1205781199051015840) + 119876
ℎ119905le 5 and119867
ℎ119905+ 119876ℎ119905
gt 5
119884119873119871
ℎ119905(1205781199051015840) = 119884
119873119871
ℎ119905+ 120579119905(119867ℎ119905(1205781199051015840) + 119876
ℎ119905minus 5) (A3)
If119867ℎ119905(1205781199051015840) + 119876
ℎ119905gt 5 and119867
ℎ119905+ 119876ℎ119905
gt 5
119884119873119871
ℎ119905(1205781199051015840) = 119884
119873119871
ℎ119905 (A4)
10 International Journal of Population Research
where 119876ℎ119905
is the number of children aged 17 or 18(recall that we carry out the simulations for childrenaged 16 or younger) and 120579
119905is equal to AR$270 for
the third trimester of 2012 and AR$340 for the fourthtrimester
B Brazil Bolsa Familia Program (BFP)
The Bolsa Familia implemented in 2003 is the flagshipconditional cash transfer program in Brazil aimed at poorhouseholds The estimated nonlabor incomes in this case arecalculated as follows
If119867119905ℎ
gt 0 and beneficiary
119884119873119871
ℎ119905(1205781199051015840) = 119884
119873119871
119905ℎ+ 119872119905ℎ(119867119905ℎ(1205781199051015840)
119867119905ℎ
minus 1) (B1)
If119867119905ℎ
= 0119867119905ℎ(1205781199051015840) = 119896 and beneficiary
119884119873119871
ℎ119905(1205781199051015840) = 119884
119873119871
119905ℎ+ 119872119896
119905ℎminus 119897119905ℎ (B2)
with 119896 = 1 2 3 119872119905ℎ
being the amount of thetransfer119872119896
119905ℎbeing the average benefit of households
with 119896 children and 119897119905ℎbeing the standard benefit of
recipients without children
C Mexico Oportunidades
Oportunidades is the main antipoverty program of theMexican Government The size of the transfer depends onthe number of children among other characteristics of thehouseholdMexicorsquos ENIGH identifies the beneficiaries of theprogram and the amount of the subsidy that each householdreceives The simulation is made as follows
119884119873119871
ℎ119905(1205781199051015840) = 119884
119873119871
119905ℎ+ 119872119896
119905ℎminus 119872119905ℎ
if 119867119905ℎ(1205781199051015840) = 119896 119867
119905ℎ= 119896 and beneficiary
(C1)
where 119872119896119905ℎis the average transfer of a beneficiary household
with 119896 children and119872119905ℎis the original household transfer
D Uruguay Asignaciones Familiares
The Asignaciones Familiares program in Uruguay is targetedat the children of workers The amount of the transfer varieswith the householdrsquos income Since the size of the transferdepends on characteristics of the child such as the age thesimulation is carried out as follows
If119867119905ℎ
gt 0 and beneficiary
119884119873119871
ℎ119905(1205781199051015840) = 119884
119873119871
119905ℎ+ 119872119905ℎ(119867119905ℎ(1205781199051015840)
119867119905ℎ
minus 1) (D1)
If119867119905ℎ
= 0119867119905ℎ(1205781199051015840) = 119896 and beneficiary
119884119873119871
ℎ119905(1205781199051015840) = 119884
119873119871
119905ℎ+ 119872119896
119905ℎ (D2)
with 119896 = 1 2 3 119872119905ℎ
being the total amount ofthe transfer and 119872
119896
119905ℎbeing the average transfer of a
household with 119896 children
Disclosure
The information presented herein is solely the responsibilityof the authors and does not necessarily reflect the opinions ofthe aforementioned entities
Competing Interests
The authors declare that they have no competing interests
Acknowledgments
This paper was written in the framework of the projectEnhancing Womenrsquos Economic Empowerment throughBetter Policies in Latin America a joint initiative betweenCEDLAS-UNLP and CIEDUR with the support ofIDRC Canada The paper includes material from NicolasBadaraccorsquos Masters dissertation at Universidad Nacionalde La Plata The authors are grateful to Carolina RobinoGuillermo Cruces Jorge Paz Facundo Albornoz andseminar participants at AAEP (Posadas 2014) Network ofInequality and Poverty (UNGS 2014) and Conference onSocial Protection (IDRC and IPC Brasilia 2014) for valuablecomments and suggestions
References
[1] T W Guinnane ldquoThe historical fertility transition a guide foreconomistsrdquo Journal of Economic Literature vol 49 no 3 pp589ndash614 2011
[2] L E Jones and M Tertilt ldquoAn economic history of fertility inthe US 1826ndash1960rdquo NBERWorking Paper 12796 2006
[3] J Chackiel La Dinamica Demografica en America LatinaSerie Poblacion y Desarrollo No 52 Centro Latinoamericanoy Caribeno de Demografıa-Division de Poblacion de CEPAL2004
[4] ECLAC ldquoLa fecundidad en America Latina un descensoacelerado y heterogeneo con profundas transformacionesdemograficas y socialesrdquo in Fertility Latin American and theCaribbean Demographic Observatory no 5 Centro Latinoa-mericano y Caribeno de DemografıamdashDivision de Poblacionde CEPAL 2008
[5] ECLAC ldquoChapter II current situation and outlook for fertilityin Latin Americardquo in Social Panorama of Latin America 2011ECLAC 2011
[6] L Gasparini and M Marchionni Bridging Gender Gaps TheRise and Deceleration of Female Labor Force ParticipationCEDLAS-UNLP 2015
[7] L Gasparini and N Lustig ldquoThe rise and fall of incomeinequality in Latin Americardquo inThe Oxford Handbook of LatinAmerican Economics J Ocampo and J Ros Eds chapter 27 pp691ndash714 Oxford University Press Oxford UK 2011
[8] G Cornia Ed Falling Inequality in Latin America PolicyChanges and Lessons Oxford University Press 2014
International Journal of Population Research 11
[9] F Alvaredo and L Gasparini ldquoRecent trends in inequalityand poverty in developing countriesrdquo in Handbook of IncomeDistribution Atkinson and Bourguignon Eds vol 2 chapter 9pp 697ndash805 Elsevier New York NY USA 2015
[10] M Marchionni and L Gasparini ldquoTracing out the effects ofdemographic changes on the income distribution The caseof Greater Buenos Aires 1980ndash1998rdquo Journal of EconomicInequality vol 5 no 1 pp 97ndash114 2007
[11] J M Wooldridge Econometric Analysis of Cross Section andPanel Data MIT Press Boston Mass USA 2002
[12] L Gasparini M Marchionni and W Sosa Escudero ldquoChar-acterization of inequality changes through microeconometricdecompositions The case of Greater Buenos Airesrdquo in TheMicroeconomics of Income Distribution Dynamics in East Asiaand Latina America F Bourguignon F Ferreira and N LustigEds Oxford University Press New York NY USA 2004
[13] S Garganta and L Gasparini ldquoThe impact of a social programon labor informality the case of AUH in Argentinardquo Journal ofDevelopment Economics vol 115 pp 99ndash110 2015
Submit your manuscripts athttpwwwhindawicom
Child Development Research
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Education Research International
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Biomedical EducationJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Psychiatry Journal
ArchaeologyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
AnthropologyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Research and TreatmentSchizophrenia
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Urban Studies Research
Population ResearchInternational Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CriminologyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Aging ResearchJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
NursingResearch and Practice
Current Gerontologyamp Geriatrics Research
Hindawi Publishing Corporationhttpwwwhindawicom
Volume 2014
Sleep DisordersHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
AddictionJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Depression Research and TreatmentHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Geography Journal
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Research and TreatmentAutism
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Economics Research International
6 International Journal of Population Research
Nonlabor income is usually assumed to be exogenous inthemicrosimulation literature However as stated in previoussections the expansion of social programs to families withchildren suggests a potential link between fertility decisionsand nonlabor income Accordingly we assume that nonlaborincome is given by
119884119873119871
ℎ119905= 119898ℎ119905+ 119892 (119867
ℎ119905) (5)
where119898ℎ119905is exogenous and119892(119867
ℎ119905) represents the component
of nonlabor income that depends on the number of childrenWe label the parameters of the fertility decisions at time 119905
as 120578119905and those at time 1199051015840 as 120578
1199051015840 A key step in themethodology
is to estimate the counterfactual number of children in agiven year 119905 if fertility outcomes were determined as in analternative year 119905
1015840 Once this term labeled as 119867ℎ119905(1205781199051015840) is
estimated three microsimulation exercises are carried out byreplacing this estimate in the household per capita incomeequation
The first exercise aims to capture the contribution of thechange in fertility parameters 120578 to the actual change in theincome distribution through the denominator of (1) Thesimulated per capita income is given by
119910119863
119894119905=
119884119871
ℎ119905+ 119884119873119871
ℎ119905
119873ℎ119905(1205781199051015840)
(6)
where 119873ℎ119905(1205781199051015840) = 119877
ℎ119905+ 119867ℎ119905(1205781199051015840) Superscript 119863 in the
simulated per capita income stands for the direct-size effectthat is the change in the income distribution due to changesin the number of household members among whom totalhousehold income should be distributed
The second exercise involves simulating the labor incomethat a household would have in year 119905 if parameters thatgovern fertility decisionswere those of year 1199051015840 In other wordsthe exercise consists in replacing the simulated number ofchildren 119867
ℎ119905(1205781199051015840) in the equation of individual labor income
(4) to obtain
119910119867
119894119905=
119884119871
ℎ119905(1205781199051015840) + 119884
119873119871
ℎ119905
119873ℎ119905
(7)
Superscript 119867 in the simulated per capita income stands forthe hours-size effect that is the contribution of the changein fertility parameters 120578 to the actual change in the incomedistribution through the indirect channel of affecting thehours-of-work decisions
We carry out a third exercise by simulating the coun-terfactual distribution arising from a change in nonlaborincome driven by changes in fertility decisions We refer tothe distributional impact of changes through this channelas the nonlabor-income effect and thus label the per capitaincome variable in (8) with the NL superscript
119910119873119871
119894119905=
119884119871
ℎ119905+ 119884119873119871
ℎ119905(1205781199051015840)
119873ℎ119905
(8)
Finally (9) considers the three channels simultaneously toobtain the total effect (superscript119879) of changes in the fertilityparameters
119910119879
119894119905=
119884119871
ℎ119905(1205781199051015840) + 119884
119873119871
ℎ119905(1205781199051015840)
119873ℎ119905(1205781199051015840)
(9)
We denote the per capita income distribution among indi-viduals in year 119905 as 119863
119905(119910119905) = (119910
1119905 1199102119905 119910
119899119905) where 119899
denotes the total number of individuals in the populationLet 119868(sdot) be a distributive indicator as the Gini coefficientor the poverty headcount ratio Then the contribution ofthe change in fertility parameters 120578 to the actual change inthe income distribution through the direct channel is givenby (10) The contributions through the other channels arecomputed analogously
119868 (119863119905(119910119905)) minus 119868 (119863
119905(119910119863
119905)) (10)
A shortcoming of the microsimulation methodology is thatdecomposition is path-dependent This means that takingyear 119905 or year 1199051015840 as the base year does not necessarily implythe same result Accordingly we perform the simulationsalternating the base year and report the average effect for eachexercise
In the following subsections we provide details concern-ing the specification and estimation of fertility labor incomeand nonlabor-income equations
32 Fertility Decisions We assume that the number of chil-dren in (3) follows a Poisson process with parameter 120583
ℎ119905
Formally
119867ℎ119905
sim Poisson (120583ℎ119905) with 120583
ℎ119905= 119864 (119867
ℎ119905| 119885ℎ119905) = exp (119885
1015840
ℎ119905120578119905) (11)
Then
Prob (119867ℎ119905
= 1198670) =
exp (minus120583ℎ119905) (120583ℎ119905)1198670
1198670
with 1198670= 0 1 2 (12)
Equation (12) is the Poisson regressionmodel fromwhich it ispossible to consistently estimate parameters 120578
119905by maximum
likelihood (it can be shown that consistency holds for themaximum likelihood estimators of 120578
119905as long as the real
distribution is any of the linear exponential family to whichthe Poisson distribution belongs provided that the condi-tional mean in (11) is correctly specified see eg Wooldridge[11]) We estimate separate models for two-parent householdand single-parent households For simplicity we denote theestimates by 120578
Once parameters 120578 are estimated we proceed to simulatechanges in the number of children Since the objective isto simulate these changes as a consequence of changes onlyin the parameters 120578 it is necessary to keep unobservablefactors fixed Therefore we characterize each household bythe quantile it occupies in the distribution of children ofyear 119905 Let 119865
120578119905|119885ℎ119905
(sdot) be the cumulative probability function ofa random variable that follows a Poisson distribution withparameter equal to exp(1198851015840
ℎ119905120578119905) and let 119902
ℎ119905be the quantile
for household ℎ at time 119905 that is 119865120578119905|119885ℎ119905
(119867ℎ119905) = 119902
ℎ119905 The
International Journal of Population Research 7
simulated number of children in household ℎ will be the onethat places it in the 119902
ℎ119905quantile of the distribution of children
with the relevant parameters of time 1199051015840(1205781199051015840) conditional to
the observable characteristics 119885ℎ119905 Formally the simulated
number of children in household ℎ at year 119905 using theestimated fertility parameters for year 1199051015840 is given by
119867ℎ119905(1205781199051015840) = 119865
minus1
1205781199051015840 |119885ℎ119905
∘ 119865120578119905|119885ℎ119905
(119867ℎ119905) (13)
33 Labor Incomes To estimate individual labor income in(4) we follow Gasparini et al [12] in assuming that bothwages and hours worked are determined in a reduced-formmodel of the labor market equilibrium
ln119908lowast
119894119905= 1198831015840
1119894119905120573119905+ 120576119882
119894119905 (14)
119871lowast
119894119905= 1198831015840
2119894119905120574119905+ 119867119894119905120582119905+ 120576119871
119894119905 (15)
with
if 119871lowast119894119905gt 0
119908119894119905= 119908lowast
119894119905
119871119894119905= 119871lowast
119894119905
if 119871lowast119894119905le 0
119908119894119905= 0
119871119894119905= 0
(120576119882
119894119905 120576119871
119894119905) sim 119873(0 0 120590
2
119882119905 1205902
119871119905 120588119905)
(16)
where119908lowast119894119905and119871lowast
119894119905are unobserved latent variables Vectors119883
1119894119905
and 1198832119894119905
include observable factors affecting hourly wagesand hours of work respectively while 119867
119894119905stands for the
number of children 120573119905 120574119905(vectors) and 120582
119905are parameters
to be estimated along with 1205902
119882119905 1205902119871119905 and 120588
119905
We estimate (14) by Heckmanrsquos maximum likelihoodmethod where a censored version of (15) is used as a selectionequation replacing hours of work by a binary variable thatindicates whether the individual works For estimation of(15) we use aTobitmodel (although this estimation strategy isnot fully efficient efficiency loss is not necessarily significantfor a given sample size and this alternative has certaincomputational advantages over a full information procedurefor more details see Gasparini et al [12]) We estimateseparate models for head of households and spouses
To simulate changes in individual labor incomes wereplace the observed number of children of individual 119894 atyear 119905 in (15) by the simulated number of children 119867
119905ℎ(1205781199051015840)
while assigning himher a wage based on (14) and an errorterm drawn from the bivariate distribution implicit in themodel
34 Nonlabor Incomes The measure of nonlabor incomeincludes pensions capital income and transfers Conditionalcash transfers are included as government transfers Wesimulate nonlabor incomes in an ad hoc manner taking
into account the characteristics of each countryrsquos socialprogram In those cases where it is not possible to identifythe recipients in the household survey we identify potentialbeneficiaries and estimate the amount of the transfers basedon the design of each program The general framework is asfollows nonlabor income of household ℎ at year 119905 is given by(5) where 119892(119867
ℎ119905) represents the part of nonlabor income that
depends on the number of children Typically cash transfersin poverty-alleviation programs vary with the number ofchildren in the family For example if the program has a fixedamount 120579
119905per child the nonlabor income becomes
119884119873119871
ℎ119905= 119898ℎ119905+ 120579119905119867ℎ119905 (17)
The simulated nonlabor income is calculated using thecounterfactual number of children in a given year 119905 if fertilityoutcomes were determined as in an alternative year 1199051015840
119884119873119871
ℎ119905(1205781199051015840) = 119898
ℎ119905+ 120579ℎ119905119867ℎ119905(1205781199051015840) (18)
Even though all programs have a similar setting to the onedescribed they differ in some aspects maximum number ofchildren age and in some cases how to identify the programrsquosrecipients The details are given in the Appendix We sim-ulate nonlabor incomes for Argentina Brazil Mexico andUruguay Even though the programs in Chile El Salvadorand Peru have similar characteristics it is not possible toestimate how the amount of transfers varies with the numberof children given the information in the household surveysFor these countries we assume that nonlabor income iscompletely exogenous
4 The Results
The results of the microsimulations regarding poverty arepresented inTable 2 Column (i) displays the observed changein the headcount ratio while the rest of the columns presentthe impact of the changes in fertility on poverty Column(v) shows the sum of all effects explored In order to betterunderstand the information in the table take the case ofBrazil as an example Column (i) shows that between 1990and 2012 the poverty headcount ratio in that country (usingthe line of US$ 4 a day) fell by around 25 points Thevalue in the last column has the following interpretationif fertility had been the only factor that changed duringthat period then the poverty headcount ratio in Brazilwould have fallen by 392 points This change is statisticallysignificant although not very big compared to the actual fallin poverty in that period The overall effect is the result ofa significant direct effect (minus374) and two smaller effects thatgo in different directions On the one hand the differentialreduction in fertility implied a greater increase in hours ofwork among more disadvantaged households which in turncontributed though very slightly to a further reduction inincome poverty On the other hand the falling patterns infertility among the poor ameliorated the poverty-decreasingimpact of the conditional cash transfer programs that aretargeted at families with children (mainly the Bolsa Familiain the 2000s) However this effect was quantitatively almostinsignificant
8 International Journal of Population Research
Table 2 Impact of fertility changes on poverty Changes in the poverty headcount ratio (US$ 4 line)
Country Period Observed change (i) EffectsDirect (ii) Hours (iii) NLI (iv) Total (v)
Argentina 1992ndash2012 minus1122 minus139 004 039 minus080(008) (002) (001) (001) (002)
Brazil 1990ndash2012 minus2516 minus374 minus017 014 minus392(003) (001) (000) (000) (001)
Chile 1990ndash2011 minus3324 minus156 018 mdash minus143(005) (001) (000) mdash (001)
El Salvador 1991ndash2010 minus1412 minus498 minus034 mdash minus531(006) (004) (001) mdash (005)
Mexico 1992ndash2012 minus350 minus435 minus030 025 minus443(010) (004) (002) (001) (004)
Peru 1997ndash2012 minus2320 minus383 minus019 mdash minus394(009) (003) (001) mdash (003)
Uruguay 1995ndash2012 minus156 minus008 000 002 minus005(003) (002) (000) (000) (002)
Source own calculations based on microdata from national household surveysNote All effects are significant at the 1 level The standard errors were calculated using bootstrap with 200 replications The values of each effect are averagesthat result from taking alternatively each year in the comparison as the base year The sample includes only households in which the head is between 25 and45 years old NLI nonlabor income
Table 3 Impact of fertility changes on inequality Changes in the Gini coefficient
Country Period Observed change (i) EffectsDirect (ii) Hours (iii) NLI (iv) Total (v)
Argentina 1992ndash2012 minus544 minus104 minus001 013 minus095(010) (001) (000) (000) (002)
Brazil 1990ndash2012 minus718 minus130 minus006 006 minus132(004) (000) (000) (000) (000)
Chile 1990ndash2011 minus178 minus075 004 mdash minus073(008) (001) (000) mdash (001)
El Salvador 1991ndash2010 minus768 minus130 minus007 mdash minus140(006) (002) (000) mdash (002)
Mexico 1992ndash2012 minus086 minus168 minus011 012 minus169(019) (002) (000) (000) (002)
Peru 1997ndash2012 minus732 minus192 minus006 mdash minus199(013) (001) (000) mdash (001)
Uruguay 1995ndash2012 minus203 minus012 minus001 002 minus013(004) (001) (000) (000) (001)
Source own calculations based on microdata from national household surveysNote All effects are significant at the 1 level The standard errors were calculated using bootstrap with 200 replications The values of each effect are averagesthat result from taking alternatively each year in the comparison as the base year The sample includes only households in which the head is between 25 and45 years old NLI nonlabor income
Changes in fertility patterns over the last two decades inLatin America have implied a reduction in income povertyThe impact is statistically significant and in some caseseconomically large For instance the estimated poverty-reduction effect was more than 4 points in Mexico andEl Salvador Most of the effect comes through the directeffect a reduction in fertility rates among the most dis-advantaged groups reduced family size and increased percapita income The hours-of-work effect is in most casespoverty reducing and the nonlabor-income effect is poverty
increasing but in both cases the estimated sizes aresmall
Table 3 shows the results of the simulations on incomeinequality measured by the Gini coefficient Inequality inBrazil as measured by the Gini coefficient fell by 718points between 1990 and 2012 If fertility had been the onlyfactor that changed in that period then the Gini coefficientwould have fallen by 132 points (column v) This change isstatistically significant representing around 18 of the actualreduction in inequality during that period The differential
International Journal of Population Research 9
reduction in fertility across socioeconomic groups in Brazilcontributed to the observed decline in inequality in the lasttwo decadesThe overall effect ofminus132 points in column (v) isthe result of a significant direct effect (minus130) and two smallereffects going in different directions
Some interesting general results emerge from Table 3First changes in fertility have implied a decline in incomeinequality The differential pattern in fertility across groupsexperienced in most Latin American countries over thelast decades translated into an equalizing impact on theincome distribution Second this effect is small althoughnot negligible To be sure demographic changes are not thecentral reason behind changes in income inequality but theyare statistically significant and economically relevant Onaverage fertility changes account for a fall of around onepoint in the Gini coefficient
Third most of the impact comes from the direct effectThe differential fall in fertility rates among socioeconomicgroups had a larger impact on the family size of poorerfamilies implying a proportionally larger increasing effect ontheir per capita incomes
Fourth the effect of fertility changes on hours of workand in turn on incomes is smaller and has different signsacross countries It is important to notice that a more intensereduction in fertility among the poor may be consistentwith a positive sign (inequality-increasing) for this effectThis could happen if the elasticity of hours of work withrespect to the number of children is higher among thenonpoor and if the extra hours worked triggered by the fallin fertility imply a household income increase proportionallylarger for the nonpoor Anyway with one exception thehours-of-work effect has a negative sign implying that theasymmetric reduction in fertility contributed to a reductionin income inequality by fostering higher participation in thelabor market among poor parents (mostly mothers)
Finally the effect through nonlabor incomes is alwaysinequality-increasing but small The fall in fertility amongthe poor implied a lower inequality-reducing impact of theconditional cash transfer programs that were introduced andexpanded in the region mainly during the 2000s The effecthowever is in most cases very small
5 Concluding Remarks
Poverty and inequality fell in Latin America over the lastdecades driven by a large number of factors In thispaper we highlight one factor that is usually overlooked inthe economic literature the impact of changes in fertilityThe average number of children fell in households fromall groups but especially among the more vulnerable afact that could have distributive implications We quantifythe consequences of this demographic pattern by applyingmicrosimulation techniquesThe results suggest that changesin fertility that took place over the last decades in LatinAmerica contributed to the reduction in income inequalityas well as in poverty The impacts found were in generalstatistically significant and economically relevant althoughnot very large Although surely there are more importantdeterminants behind the observed patterns in the income
distributions of the Latin American countries demographicfactors should not be overlooked
The paper has focused on three simple yet central chan-nels the change in the number of children (i) enlarges thefamily size and reduces per capita income (ii) affects thelabor supply decisions and (iii) modifies the income supportfrom social programs Certainly there are other potentiallyrelevant channels that are ignored in the paper For instancea reduction in the number of children per family couldimply higher inheritances and higher family spending oneducation and health per child and hence better incomeperspectives Also smaller families may be better positionedto take advantage of labor market opportunities for instancethrough migration Moreover lower fertility rates may beassociated with a reduction in the relative supply of unskilledlabor and hence an increase in its relative wage
Although in the short and medium run lower fertility inmore disadvantaged households may contribute to reducingpoverty and inequality in the long run the implicationsare more nuanced When children become adults they maycontribute to their parentsrsquo incomes a fall in fertility mayreduce incomes for the elderly especially in countries wherethe pension system is weak and ultimately contribute tohigher poverty through that intertemporal channel
Appendix
In this appendix we describe the issues regarding the simu-lations of the nonlabor incomes for each country
A Argentina Asignacioacuten Universal porHijo (AUH)
Argentina launched the conditional cash transfer programAUH in 2009 The beneficiaries are children under the ageof 18 in households whose members are unemployed orinformal workers The program provides a monthly cashsubsidy per child up to 5 children Since Argentinarsquos house-hold survey (EPH) does not include a question to identifyrecipients of this program we follow Garganta and Gasparini[13] assuming full take-up among those who qualify Thesimulation is made according to the following rule
If119867ℎ119905(1205781199051015840) + 119876
ℎ119905le 5 and119867
ℎ119905+ 119876ℎ119905
le 5
119884119873119871
ℎ119905(1205781199051015840) = 119884
119873119871
ℎ119905+ 120579119905(119867ℎ119905(1205781199051015840) minus 119867
ℎ119905) (A1)
If119867ℎ119905(1205781199051015840) + 119876
ℎ119905gt 5 and119867
ℎ119905+ 119876ℎ119905
le 5
119884119873119871
ℎ119905(1205781199051015840) = 119884
119873119871
ℎ119905+ 120579119905(5 minus 119867
ℎ119905minus 119876ℎ119905) (A2)
If119867ℎ119905(1205781199051015840) + 119876
ℎ119905le 5 and119867
ℎ119905+ 119876ℎ119905
gt 5
119884119873119871
ℎ119905(1205781199051015840) = 119884
119873119871
ℎ119905+ 120579119905(119867ℎ119905(1205781199051015840) + 119876
ℎ119905minus 5) (A3)
If119867ℎ119905(1205781199051015840) + 119876
ℎ119905gt 5 and119867
ℎ119905+ 119876ℎ119905
gt 5
119884119873119871
ℎ119905(1205781199051015840) = 119884
119873119871
ℎ119905 (A4)
10 International Journal of Population Research
where 119876ℎ119905
is the number of children aged 17 or 18(recall that we carry out the simulations for childrenaged 16 or younger) and 120579
119905is equal to AR$270 for
the third trimester of 2012 and AR$340 for the fourthtrimester
B Brazil Bolsa Familia Program (BFP)
The Bolsa Familia implemented in 2003 is the flagshipconditional cash transfer program in Brazil aimed at poorhouseholds The estimated nonlabor incomes in this case arecalculated as follows
If119867119905ℎ
gt 0 and beneficiary
119884119873119871
ℎ119905(1205781199051015840) = 119884
119873119871
119905ℎ+ 119872119905ℎ(119867119905ℎ(1205781199051015840)
119867119905ℎ
minus 1) (B1)
If119867119905ℎ
= 0119867119905ℎ(1205781199051015840) = 119896 and beneficiary
119884119873119871
ℎ119905(1205781199051015840) = 119884
119873119871
119905ℎ+ 119872119896
119905ℎminus 119897119905ℎ (B2)
with 119896 = 1 2 3 119872119905ℎ
being the amount of thetransfer119872119896
119905ℎbeing the average benefit of households
with 119896 children and 119897119905ℎbeing the standard benefit of
recipients without children
C Mexico Oportunidades
Oportunidades is the main antipoverty program of theMexican Government The size of the transfer depends onthe number of children among other characteristics of thehouseholdMexicorsquos ENIGH identifies the beneficiaries of theprogram and the amount of the subsidy that each householdreceives The simulation is made as follows
119884119873119871
ℎ119905(1205781199051015840) = 119884
119873119871
119905ℎ+ 119872119896
119905ℎminus 119872119905ℎ
if 119867119905ℎ(1205781199051015840) = 119896 119867
119905ℎ= 119896 and beneficiary
(C1)
where 119872119896119905ℎis the average transfer of a beneficiary household
with 119896 children and119872119905ℎis the original household transfer
D Uruguay Asignaciones Familiares
The Asignaciones Familiares program in Uruguay is targetedat the children of workers The amount of the transfer varieswith the householdrsquos income Since the size of the transferdepends on characteristics of the child such as the age thesimulation is carried out as follows
If119867119905ℎ
gt 0 and beneficiary
119884119873119871
ℎ119905(1205781199051015840) = 119884
119873119871
119905ℎ+ 119872119905ℎ(119867119905ℎ(1205781199051015840)
119867119905ℎ
minus 1) (D1)
If119867119905ℎ
= 0119867119905ℎ(1205781199051015840) = 119896 and beneficiary
119884119873119871
ℎ119905(1205781199051015840) = 119884
119873119871
119905ℎ+ 119872119896
119905ℎ (D2)
with 119896 = 1 2 3 119872119905ℎ
being the total amount ofthe transfer and 119872
119896
119905ℎbeing the average transfer of a
household with 119896 children
Disclosure
The information presented herein is solely the responsibilityof the authors and does not necessarily reflect the opinions ofthe aforementioned entities
Competing Interests
The authors declare that they have no competing interests
Acknowledgments
This paper was written in the framework of the projectEnhancing Womenrsquos Economic Empowerment throughBetter Policies in Latin America a joint initiative betweenCEDLAS-UNLP and CIEDUR with the support ofIDRC Canada The paper includes material from NicolasBadaraccorsquos Masters dissertation at Universidad Nacionalde La Plata The authors are grateful to Carolina RobinoGuillermo Cruces Jorge Paz Facundo Albornoz andseminar participants at AAEP (Posadas 2014) Network ofInequality and Poverty (UNGS 2014) and Conference onSocial Protection (IDRC and IPC Brasilia 2014) for valuablecomments and suggestions
References
[1] T W Guinnane ldquoThe historical fertility transition a guide foreconomistsrdquo Journal of Economic Literature vol 49 no 3 pp589ndash614 2011
[2] L E Jones and M Tertilt ldquoAn economic history of fertility inthe US 1826ndash1960rdquo NBERWorking Paper 12796 2006
[3] J Chackiel La Dinamica Demografica en America LatinaSerie Poblacion y Desarrollo No 52 Centro Latinoamericanoy Caribeno de Demografıa-Division de Poblacion de CEPAL2004
[4] ECLAC ldquoLa fecundidad en America Latina un descensoacelerado y heterogeneo con profundas transformacionesdemograficas y socialesrdquo in Fertility Latin American and theCaribbean Demographic Observatory no 5 Centro Latinoa-mericano y Caribeno de DemografıamdashDivision de Poblacionde CEPAL 2008
[5] ECLAC ldquoChapter II current situation and outlook for fertilityin Latin Americardquo in Social Panorama of Latin America 2011ECLAC 2011
[6] L Gasparini and M Marchionni Bridging Gender Gaps TheRise and Deceleration of Female Labor Force ParticipationCEDLAS-UNLP 2015
[7] L Gasparini and N Lustig ldquoThe rise and fall of incomeinequality in Latin Americardquo inThe Oxford Handbook of LatinAmerican Economics J Ocampo and J Ros Eds chapter 27 pp691ndash714 Oxford University Press Oxford UK 2011
[8] G Cornia Ed Falling Inequality in Latin America PolicyChanges and Lessons Oxford University Press 2014
International Journal of Population Research 11
[9] F Alvaredo and L Gasparini ldquoRecent trends in inequalityand poverty in developing countriesrdquo in Handbook of IncomeDistribution Atkinson and Bourguignon Eds vol 2 chapter 9pp 697ndash805 Elsevier New York NY USA 2015
[10] M Marchionni and L Gasparini ldquoTracing out the effects ofdemographic changes on the income distribution The caseof Greater Buenos Aires 1980ndash1998rdquo Journal of EconomicInequality vol 5 no 1 pp 97ndash114 2007
[11] J M Wooldridge Econometric Analysis of Cross Section andPanel Data MIT Press Boston Mass USA 2002
[12] L Gasparini M Marchionni and W Sosa Escudero ldquoChar-acterization of inequality changes through microeconometricdecompositions The case of Greater Buenos Airesrdquo in TheMicroeconomics of Income Distribution Dynamics in East Asiaand Latina America F Bourguignon F Ferreira and N LustigEds Oxford University Press New York NY USA 2004
[13] S Garganta and L Gasparini ldquoThe impact of a social programon labor informality the case of AUH in Argentinardquo Journal ofDevelopment Economics vol 115 pp 99ndash110 2015
Submit your manuscripts athttpwwwhindawicom
Child Development Research
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Education Research International
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Biomedical EducationJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Psychiatry Journal
ArchaeologyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
AnthropologyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Research and TreatmentSchizophrenia
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Urban Studies Research
Population ResearchInternational Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CriminologyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Aging ResearchJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
NursingResearch and Practice
Current Gerontologyamp Geriatrics Research
Hindawi Publishing Corporationhttpwwwhindawicom
Volume 2014
Sleep DisordersHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
AddictionJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Depression Research and TreatmentHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Geography Journal
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Research and TreatmentAutism
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Economics Research International
International Journal of Population Research 7
simulated number of children in household ℎ will be the onethat places it in the 119902
ℎ119905quantile of the distribution of children
with the relevant parameters of time 1199051015840(1205781199051015840) conditional to
the observable characteristics 119885ℎ119905 Formally the simulated
number of children in household ℎ at year 119905 using theestimated fertility parameters for year 1199051015840 is given by
119867ℎ119905(1205781199051015840) = 119865
minus1
1205781199051015840 |119885ℎ119905
∘ 119865120578119905|119885ℎ119905
(119867ℎ119905) (13)
33 Labor Incomes To estimate individual labor income in(4) we follow Gasparini et al [12] in assuming that bothwages and hours worked are determined in a reduced-formmodel of the labor market equilibrium
ln119908lowast
119894119905= 1198831015840
1119894119905120573119905+ 120576119882
119894119905 (14)
119871lowast
119894119905= 1198831015840
2119894119905120574119905+ 119867119894119905120582119905+ 120576119871
119894119905 (15)
with
if 119871lowast119894119905gt 0
119908119894119905= 119908lowast
119894119905
119871119894119905= 119871lowast
119894119905
if 119871lowast119894119905le 0
119908119894119905= 0
119871119894119905= 0
(120576119882
119894119905 120576119871
119894119905) sim 119873(0 0 120590
2
119882119905 1205902
119871119905 120588119905)
(16)
where119908lowast119894119905and119871lowast
119894119905are unobserved latent variables Vectors119883
1119894119905
and 1198832119894119905
include observable factors affecting hourly wagesand hours of work respectively while 119867
119894119905stands for the
number of children 120573119905 120574119905(vectors) and 120582
119905are parameters
to be estimated along with 1205902
119882119905 1205902119871119905 and 120588
119905
We estimate (14) by Heckmanrsquos maximum likelihoodmethod where a censored version of (15) is used as a selectionequation replacing hours of work by a binary variable thatindicates whether the individual works For estimation of(15) we use aTobitmodel (although this estimation strategy isnot fully efficient efficiency loss is not necessarily significantfor a given sample size and this alternative has certaincomputational advantages over a full information procedurefor more details see Gasparini et al [12]) We estimateseparate models for head of households and spouses
To simulate changes in individual labor incomes wereplace the observed number of children of individual 119894 atyear 119905 in (15) by the simulated number of children 119867
119905ℎ(1205781199051015840)
while assigning himher a wage based on (14) and an errorterm drawn from the bivariate distribution implicit in themodel
34 Nonlabor Incomes The measure of nonlabor incomeincludes pensions capital income and transfers Conditionalcash transfers are included as government transfers Wesimulate nonlabor incomes in an ad hoc manner taking
into account the characteristics of each countryrsquos socialprogram In those cases where it is not possible to identifythe recipients in the household survey we identify potentialbeneficiaries and estimate the amount of the transfers basedon the design of each program The general framework is asfollows nonlabor income of household ℎ at year 119905 is given by(5) where 119892(119867
ℎ119905) represents the part of nonlabor income that
depends on the number of children Typically cash transfersin poverty-alleviation programs vary with the number ofchildren in the family For example if the program has a fixedamount 120579
119905per child the nonlabor income becomes
119884119873119871
ℎ119905= 119898ℎ119905+ 120579119905119867ℎ119905 (17)
The simulated nonlabor income is calculated using thecounterfactual number of children in a given year 119905 if fertilityoutcomes were determined as in an alternative year 1199051015840
119884119873119871
ℎ119905(1205781199051015840) = 119898
ℎ119905+ 120579ℎ119905119867ℎ119905(1205781199051015840) (18)
Even though all programs have a similar setting to the onedescribed they differ in some aspects maximum number ofchildren age and in some cases how to identify the programrsquosrecipients The details are given in the Appendix We sim-ulate nonlabor incomes for Argentina Brazil Mexico andUruguay Even though the programs in Chile El Salvadorand Peru have similar characteristics it is not possible toestimate how the amount of transfers varies with the numberof children given the information in the household surveysFor these countries we assume that nonlabor income iscompletely exogenous
4 The Results
The results of the microsimulations regarding poverty arepresented inTable 2 Column (i) displays the observed changein the headcount ratio while the rest of the columns presentthe impact of the changes in fertility on poverty Column(v) shows the sum of all effects explored In order to betterunderstand the information in the table take the case ofBrazil as an example Column (i) shows that between 1990and 2012 the poverty headcount ratio in that country (usingthe line of US$ 4 a day) fell by around 25 points Thevalue in the last column has the following interpretationif fertility had been the only factor that changed duringthat period then the poverty headcount ratio in Brazilwould have fallen by 392 points This change is statisticallysignificant although not very big compared to the actual fallin poverty in that period The overall effect is the result ofa significant direct effect (minus374) and two smaller effects thatgo in different directions On the one hand the differentialreduction in fertility implied a greater increase in hours ofwork among more disadvantaged households which in turncontributed though very slightly to a further reduction inincome poverty On the other hand the falling patterns infertility among the poor ameliorated the poverty-decreasingimpact of the conditional cash transfer programs that aretargeted at families with children (mainly the Bolsa Familiain the 2000s) However this effect was quantitatively almostinsignificant
8 International Journal of Population Research
Table 2 Impact of fertility changes on poverty Changes in the poverty headcount ratio (US$ 4 line)
Country Period Observed change (i) EffectsDirect (ii) Hours (iii) NLI (iv) Total (v)
Argentina 1992ndash2012 minus1122 minus139 004 039 minus080(008) (002) (001) (001) (002)
Brazil 1990ndash2012 minus2516 minus374 minus017 014 minus392(003) (001) (000) (000) (001)
Chile 1990ndash2011 minus3324 minus156 018 mdash minus143(005) (001) (000) mdash (001)
El Salvador 1991ndash2010 minus1412 minus498 minus034 mdash minus531(006) (004) (001) mdash (005)
Mexico 1992ndash2012 minus350 minus435 minus030 025 minus443(010) (004) (002) (001) (004)
Peru 1997ndash2012 minus2320 minus383 minus019 mdash minus394(009) (003) (001) mdash (003)
Uruguay 1995ndash2012 minus156 minus008 000 002 minus005(003) (002) (000) (000) (002)
Source own calculations based on microdata from national household surveysNote All effects are significant at the 1 level The standard errors were calculated using bootstrap with 200 replications The values of each effect are averagesthat result from taking alternatively each year in the comparison as the base year The sample includes only households in which the head is between 25 and45 years old NLI nonlabor income
Table 3 Impact of fertility changes on inequality Changes in the Gini coefficient
Country Period Observed change (i) EffectsDirect (ii) Hours (iii) NLI (iv) Total (v)
Argentina 1992ndash2012 minus544 minus104 minus001 013 minus095(010) (001) (000) (000) (002)
Brazil 1990ndash2012 minus718 minus130 minus006 006 minus132(004) (000) (000) (000) (000)
Chile 1990ndash2011 minus178 minus075 004 mdash minus073(008) (001) (000) mdash (001)
El Salvador 1991ndash2010 minus768 minus130 minus007 mdash minus140(006) (002) (000) mdash (002)
Mexico 1992ndash2012 minus086 minus168 minus011 012 minus169(019) (002) (000) (000) (002)
Peru 1997ndash2012 minus732 minus192 minus006 mdash minus199(013) (001) (000) mdash (001)
Uruguay 1995ndash2012 minus203 minus012 minus001 002 minus013(004) (001) (000) (000) (001)
Source own calculations based on microdata from national household surveysNote All effects are significant at the 1 level The standard errors were calculated using bootstrap with 200 replications The values of each effect are averagesthat result from taking alternatively each year in the comparison as the base year The sample includes only households in which the head is between 25 and45 years old NLI nonlabor income
Changes in fertility patterns over the last two decades inLatin America have implied a reduction in income povertyThe impact is statistically significant and in some caseseconomically large For instance the estimated poverty-reduction effect was more than 4 points in Mexico andEl Salvador Most of the effect comes through the directeffect a reduction in fertility rates among the most dis-advantaged groups reduced family size and increased percapita income The hours-of-work effect is in most casespoverty reducing and the nonlabor-income effect is poverty
increasing but in both cases the estimated sizes aresmall
Table 3 shows the results of the simulations on incomeinequality measured by the Gini coefficient Inequality inBrazil as measured by the Gini coefficient fell by 718points between 1990 and 2012 If fertility had been the onlyfactor that changed in that period then the Gini coefficientwould have fallen by 132 points (column v) This change isstatistically significant representing around 18 of the actualreduction in inequality during that period The differential
International Journal of Population Research 9
reduction in fertility across socioeconomic groups in Brazilcontributed to the observed decline in inequality in the lasttwo decadesThe overall effect ofminus132 points in column (v) isthe result of a significant direct effect (minus130) and two smallereffects going in different directions
Some interesting general results emerge from Table 3First changes in fertility have implied a decline in incomeinequality The differential pattern in fertility across groupsexperienced in most Latin American countries over thelast decades translated into an equalizing impact on theincome distribution Second this effect is small althoughnot negligible To be sure demographic changes are not thecentral reason behind changes in income inequality but theyare statistically significant and economically relevant Onaverage fertility changes account for a fall of around onepoint in the Gini coefficient
Third most of the impact comes from the direct effectThe differential fall in fertility rates among socioeconomicgroups had a larger impact on the family size of poorerfamilies implying a proportionally larger increasing effect ontheir per capita incomes
Fourth the effect of fertility changes on hours of workand in turn on incomes is smaller and has different signsacross countries It is important to notice that a more intensereduction in fertility among the poor may be consistentwith a positive sign (inequality-increasing) for this effectThis could happen if the elasticity of hours of work withrespect to the number of children is higher among thenonpoor and if the extra hours worked triggered by the fallin fertility imply a household income increase proportionallylarger for the nonpoor Anyway with one exception thehours-of-work effect has a negative sign implying that theasymmetric reduction in fertility contributed to a reductionin income inequality by fostering higher participation in thelabor market among poor parents (mostly mothers)
Finally the effect through nonlabor incomes is alwaysinequality-increasing but small The fall in fertility amongthe poor implied a lower inequality-reducing impact of theconditional cash transfer programs that were introduced andexpanded in the region mainly during the 2000s The effecthowever is in most cases very small
5 Concluding Remarks
Poverty and inequality fell in Latin America over the lastdecades driven by a large number of factors In thispaper we highlight one factor that is usually overlooked inthe economic literature the impact of changes in fertilityThe average number of children fell in households fromall groups but especially among the more vulnerable afact that could have distributive implications We quantifythe consequences of this demographic pattern by applyingmicrosimulation techniquesThe results suggest that changesin fertility that took place over the last decades in LatinAmerica contributed to the reduction in income inequalityas well as in poverty The impacts found were in generalstatistically significant and economically relevant althoughnot very large Although surely there are more importantdeterminants behind the observed patterns in the income
distributions of the Latin American countries demographicfactors should not be overlooked
The paper has focused on three simple yet central chan-nels the change in the number of children (i) enlarges thefamily size and reduces per capita income (ii) affects thelabor supply decisions and (iii) modifies the income supportfrom social programs Certainly there are other potentiallyrelevant channels that are ignored in the paper For instancea reduction in the number of children per family couldimply higher inheritances and higher family spending oneducation and health per child and hence better incomeperspectives Also smaller families may be better positionedto take advantage of labor market opportunities for instancethrough migration Moreover lower fertility rates may beassociated with a reduction in the relative supply of unskilledlabor and hence an increase in its relative wage
Although in the short and medium run lower fertility inmore disadvantaged households may contribute to reducingpoverty and inequality in the long run the implicationsare more nuanced When children become adults they maycontribute to their parentsrsquo incomes a fall in fertility mayreduce incomes for the elderly especially in countries wherethe pension system is weak and ultimately contribute tohigher poverty through that intertemporal channel
Appendix
In this appendix we describe the issues regarding the simu-lations of the nonlabor incomes for each country
A Argentina Asignacioacuten Universal porHijo (AUH)
Argentina launched the conditional cash transfer programAUH in 2009 The beneficiaries are children under the ageof 18 in households whose members are unemployed orinformal workers The program provides a monthly cashsubsidy per child up to 5 children Since Argentinarsquos house-hold survey (EPH) does not include a question to identifyrecipients of this program we follow Garganta and Gasparini[13] assuming full take-up among those who qualify Thesimulation is made according to the following rule
If119867ℎ119905(1205781199051015840) + 119876
ℎ119905le 5 and119867
ℎ119905+ 119876ℎ119905
le 5
119884119873119871
ℎ119905(1205781199051015840) = 119884
119873119871
ℎ119905+ 120579119905(119867ℎ119905(1205781199051015840) minus 119867
ℎ119905) (A1)
If119867ℎ119905(1205781199051015840) + 119876
ℎ119905gt 5 and119867
ℎ119905+ 119876ℎ119905
le 5
119884119873119871
ℎ119905(1205781199051015840) = 119884
119873119871
ℎ119905+ 120579119905(5 minus 119867
ℎ119905minus 119876ℎ119905) (A2)
If119867ℎ119905(1205781199051015840) + 119876
ℎ119905le 5 and119867
ℎ119905+ 119876ℎ119905
gt 5
119884119873119871
ℎ119905(1205781199051015840) = 119884
119873119871
ℎ119905+ 120579119905(119867ℎ119905(1205781199051015840) + 119876
ℎ119905minus 5) (A3)
If119867ℎ119905(1205781199051015840) + 119876
ℎ119905gt 5 and119867
ℎ119905+ 119876ℎ119905
gt 5
119884119873119871
ℎ119905(1205781199051015840) = 119884
119873119871
ℎ119905 (A4)
10 International Journal of Population Research
where 119876ℎ119905
is the number of children aged 17 or 18(recall that we carry out the simulations for childrenaged 16 or younger) and 120579
119905is equal to AR$270 for
the third trimester of 2012 and AR$340 for the fourthtrimester
B Brazil Bolsa Familia Program (BFP)
The Bolsa Familia implemented in 2003 is the flagshipconditional cash transfer program in Brazil aimed at poorhouseholds The estimated nonlabor incomes in this case arecalculated as follows
If119867119905ℎ
gt 0 and beneficiary
119884119873119871
ℎ119905(1205781199051015840) = 119884
119873119871
119905ℎ+ 119872119905ℎ(119867119905ℎ(1205781199051015840)
119867119905ℎ
minus 1) (B1)
If119867119905ℎ
= 0119867119905ℎ(1205781199051015840) = 119896 and beneficiary
119884119873119871
ℎ119905(1205781199051015840) = 119884
119873119871
119905ℎ+ 119872119896
119905ℎminus 119897119905ℎ (B2)
with 119896 = 1 2 3 119872119905ℎ
being the amount of thetransfer119872119896
119905ℎbeing the average benefit of households
with 119896 children and 119897119905ℎbeing the standard benefit of
recipients without children
C Mexico Oportunidades
Oportunidades is the main antipoverty program of theMexican Government The size of the transfer depends onthe number of children among other characteristics of thehouseholdMexicorsquos ENIGH identifies the beneficiaries of theprogram and the amount of the subsidy that each householdreceives The simulation is made as follows
119884119873119871
ℎ119905(1205781199051015840) = 119884
119873119871
119905ℎ+ 119872119896
119905ℎminus 119872119905ℎ
if 119867119905ℎ(1205781199051015840) = 119896 119867
119905ℎ= 119896 and beneficiary
(C1)
where 119872119896119905ℎis the average transfer of a beneficiary household
with 119896 children and119872119905ℎis the original household transfer
D Uruguay Asignaciones Familiares
The Asignaciones Familiares program in Uruguay is targetedat the children of workers The amount of the transfer varieswith the householdrsquos income Since the size of the transferdepends on characteristics of the child such as the age thesimulation is carried out as follows
If119867119905ℎ
gt 0 and beneficiary
119884119873119871
ℎ119905(1205781199051015840) = 119884
119873119871
119905ℎ+ 119872119905ℎ(119867119905ℎ(1205781199051015840)
119867119905ℎ
minus 1) (D1)
If119867119905ℎ
= 0119867119905ℎ(1205781199051015840) = 119896 and beneficiary
119884119873119871
ℎ119905(1205781199051015840) = 119884
119873119871
119905ℎ+ 119872119896
119905ℎ (D2)
with 119896 = 1 2 3 119872119905ℎ
being the total amount ofthe transfer and 119872
119896
119905ℎbeing the average transfer of a
household with 119896 children
Disclosure
The information presented herein is solely the responsibilityof the authors and does not necessarily reflect the opinions ofthe aforementioned entities
Competing Interests
The authors declare that they have no competing interests
Acknowledgments
This paper was written in the framework of the projectEnhancing Womenrsquos Economic Empowerment throughBetter Policies in Latin America a joint initiative betweenCEDLAS-UNLP and CIEDUR with the support ofIDRC Canada The paper includes material from NicolasBadaraccorsquos Masters dissertation at Universidad Nacionalde La Plata The authors are grateful to Carolina RobinoGuillermo Cruces Jorge Paz Facundo Albornoz andseminar participants at AAEP (Posadas 2014) Network ofInequality and Poverty (UNGS 2014) and Conference onSocial Protection (IDRC and IPC Brasilia 2014) for valuablecomments and suggestions
References
[1] T W Guinnane ldquoThe historical fertility transition a guide foreconomistsrdquo Journal of Economic Literature vol 49 no 3 pp589ndash614 2011
[2] L E Jones and M Tertilt ldquoAn economic history of fertility inthe US 1826ndash1960rdquo NBERWorking Paper 12796 2006
[3] J Chackiel La Dinamica Demografica en America LatinaSerie Poblacion y Desarrollo No 52 Centro Latinoamericanoy Caribeno de Demografıa-Division de Poblacion de CEPAL2004
[4] ECLAC ldquoLa fecundidad en America Latina un descensoacelerado y heterogeneo con profundas transformacionesdemograficas y socialesrdquo in Fertility Latin American and theCaribbean Demographic Observatory no 5 Centro Latinoa-mericano y Caribeno de DemografıamdashDivision de Poblacionde CEPAL 2008
[5] ECLAC ldquoChapter II current situation and outlook for fertilityin Latin Americardquo in Social Panorama of Latin America 2011ECLAC 2011
[6] L Gasparini and M Marchionni Bridging Gender Gaps TheRise and Deceleration of Female Labor Force ParticipationCEDLAS-UNLP 2015
[7] L Gasparini and N Lustig ldquoThe rise and fall of incomeinequality in Latin Americardquo inThe Oxford Handbook of LatinAmerican Economics J Ocampo and J Ros Eds chapter 27 pp691ndash714 Oxford University Press Oxford UK 2011
[8] G Cornia Ed Falling Inequality in Latin America PolicyChanges and Lessons Oxford University Press 2014
International Journal of Population Research 11
[9] F Alvaredo and L Gasparini ldquoRecent trends in inequalityand poverty in developing countriesrdquo in Handbook of IncomeDistribution Atkinson and Bourguignon Eds vol 2 chapter 9pp 697ndash805 Elsevier New York NY USA 2015
[10] M Marchionni and L Gasparini ldquoTracing out the effects ofdemographic changes on the income distribution The caseof Greater Buenos Aires 1980ndash1998rdquo Journal of EconomicInequality vol 5 no 1 pp 97ndash114 2007
[11] J M Wooldridge Econometric Analysis of Cross Section andPanel Data MIT Press Boston Mass USA 2002
[12] L Gasparini M Marchionni and W Sosa Escudero ldquoChar-acterization of inequality changes through microeconometricdecompositions The case of Greater Buenos Airesrdquo in TheMicroeconomics of Income Distribution Dynamics in East Asiaand Latina America F Bourguignon F Ferreira and N LustigEds Oxford University Press New York NY USA 2004
[13] S Garganta and L Gasparini ldquoThe impact of a social programon labor informality the case of AUH in Argentinardquo Journal ofDevelopment Economics vol 115 pp 99ndash110 2015
Submit your manuscripts athttpwwwhindawicom
Child Development Research
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Education Research International
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Biomedical EducationJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Psychiatry Journal
ArchaeologyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
AnthropologyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Research and TreatmentSchizophrenia
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Urban Studies Research
Population ResearchInternational Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CriminologyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Aging ResearchJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
NursingResearch and Practice
Current Gerontologyamp Geriatrics Research
Hindawi Publishing Corporationhttpwwwhindawicom
Volume 2014
Sleep DisordersHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
AddictionJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Depression Research and TreatmentHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Geography Journal
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Research and TreatmentAutism
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Economics Research International
8 International Journal of Population Research
Table 2 Impact of fertility changes on poverty Changes in the poverty headcount ratio (US$ 4 line)
Country Period Observed change (i) EffectsDirect (ii) Hours (iii) NLI (iv) Total (v)
Argentina 1992ndash2012 minus1122 minus139 004 039 minus080(008) (002) (001) (001) (002)
Brazil 1990ndash2012 minus2516 minus374 minus017 014 minus392(003) (001) (000) (000) (001)
Chile 1990ndash2011 minus3324 minus156 018 mdash minus143(005) (001) (000) mdash (001)
El Salvador 1991ndash2010 minus1412 minus498 minus034 mdash minus531(006) (004) (001) mdash (005)
Mexico 1992ndash2012 minus350 minus435 minus030 025 minus443(010) (004) (002) (001) (004)
Peru 1997ndash2012 minus2320 minus383 minus019 mdash minus394(009) (003) (001) mdash (003)
Uruguay 1995ndash2012 minus156 minus008 000 002 minus005(003) (002) (000) (000) (002)
Source own calculations based on microdata from national household surveysNote All effects are significant at the 1 level The standard errors were calculated using bootstrap with 200 replications The values of each effect are averagesthat result from taking alternatively each year in the comparison as the base year The sample includes only households in which the head is between 25 and45 years old NLI nonlabor income
Table 3 Impact of fertility changes on inequality Changes in the Gini coefficient
Country Period Observed change (i) EffectsDirect (ii) Hours (iii) NLI (iv) Total (v)
Argentina 1992ndash2012 minus544 minus104 minus001 013 minus095(010) (001) (000) (000) (002)
Brazil 1990ndash2012 minus718 minus130 minus006 006 minus132(004) (000) (000) (000) (000)
Chile 1990ndash2011 minus178 minus075 004 mdash minus073(008) (001) (000) mdash (001)
El Salvador 1991ndash2010 minus768 minus130 minus007 mdash minus140(006) (002) (000) mdash (002)
Mexico 1992ndash2012 minus086 minus168 minus011 012 minus169(019) (002) (000) (000) (002)
Peru 1997ndash2012 minus732 minus192 minus006 mdash minus199(013) (001) (000) mdash (001)
Uruguay 1995ndash2012 minus203 minus012 minus001 002 minus013(004) (001) (000) (000) (001)
Source own calculations based on microdata from national household surveysNote All effects are significant at the 1 level The standard errors were calculated using bootstrap with 200 replications The values of each effect are averagesthat result from taking alternatively each year in the comparison as the base year The sample includes only households in which the head is between 25 and45 years old NLI nonlabor income
Changes in fertility patterns over the last two decades inLatin America have implied a reduction in income povertyThe impact is statistically significant and in some caseseconomically large For instance the estimated poverty-reduction effect was more than 4 points in Mexico andEl Salvador Most of the effect comes through the directeffect a reduction in fertility rates among the most dis-advantaged groups reduced family size and increased percapita income The hours-of-work effect is in most casespoverty reducing and the nonlabor-income effect is poverty
increasing but in both cases the estimated sizes aresmall
Table 3 shows the results of the simulations on incomeinequality measured by the Gini coefficient Inequality inBrazil as measured by the Gini coefficient fell by 718points between 1990 and 2012 If fertility had been the onlyfactor that changed in that period then the Gini coefficientwould have fallen by 132 points (column v) This change isstatistically significant representing around 18 of the actualreduction in inequality during that period The differential
International Journal of Population Research 9
reduction in fertility across socioeconomic groups in Brazilcontributed to the observed decline in inequality in the lasttwo decadesThe overall effect ofminus132 points in column (v) isthe result of a significant direct effect (minus130) and two smallereffects going in different directions
Some interesting general results emerge from Table 3First changes in fertility have implied a decline in incomeinequality The differential pattern in fertility across groupsexperienced in most Latin American countries over thelast decades translated into an equalizing impact on theincome distribution Second this effect is small althoughnot negligible To be sure demographic changes are not thecentral reason behind changes in income inequality but theyare statistically significant and economically relevant Onaverage fertility changes account for a fall of around onepoint in the Gini coefficient
Third most of the impact comes from the direct effectThe differential fall in fertility rates among socioeconomicgroups had a larger impact on the family size of poorerfamilies implying a proportionally larger increasing effect ontheir per capita incomes
Fourth the effect of fertility changes on hours of workand in turn on incomes is smaller and has different signsacross countries It is important to notice that a more intensereduction in fertility among the poor may be consistentwith a positive sign (inequality-increasing) for this effectThis could happen if the elasticity of hours of work withrespect to the number of children is higher among thenonpoor and if the extra hours worked triggered by the fallin fertility imply a household income increase proportionallylarger for the nonpoor Anyway with one exception thehours-of-work effect has a negative sign implying that theasymmetric reduction in fertility contributed to a reductionin income inequality by fostering higher participation in thelabor market among poor parents (mostly mothers)
Finally the effect through nonlabor incomes is alwaysinequality-increasing but small The fall in fertility amongthe poor implied a lower inequality-reducing impact of theconditional cash transfer programs that were introduced andexpanded in the region mainly during the 2000s The effecthowever is in most cases very small
5 Concluding Remarks
Poverty and inequality fell in Latin America over the lastdecades driven by a large number of factors In thispaper we highlight one factor that is usually overlooked inthe economic literature the impact of changes in fertilityThe average number of children fell in households fromall groups but especially among the more vulnerable afact that could have distributive implications We quantifythe consequences of this demographic pattern by applyingmicrosimulation techniquesThe results suggest that changesin fertility that took place over the last decades in LatinAmerica contributed to the reduction in income inequalityas well as in poverty The impacts found were in generalstatistically significant and economically relevant althoughnot very large Although surely there are more importantdeterminants behind the observed patterns in the income
distributions of the Latin American countries demographicfactors should not be overlooked
The paper has focused on three simple yet central chan-nels the change in the number of children (i) enlarges thefamily size and reduces per capita income (ii) affects thelabor supply decisions and (iii) modifies the income supportfrom social programs Certainly there are other potentiallyrelevant channels that are ignored in the paper For instancea reduction in the number of children per family couldimply higher inheritances and higher family spending oneducation and health per child and hence better incomeperspectives Also smaller families may be better positionedto take advantage of labor market opportunities for instancethrough migration Moreover lower fertility rates may beassociated with a reduction in the relative supply of unskilledlabor and hence an increase in its relative wage
Although in the short and medium run lower fertility inmore disadvantaged households may contribute to reducingpoverty and inequality in the long run the implicationsare more nuanced When children become adults they maycontribute to their parentsrsquo incomes a fall in fertility mayreduce incomes for the elderly especially in countries wherethe pension system is weak and ultimately contribute tohigher poverty through that intertemporal channel
Appendix
In this appendix we describe the issues regarding the simu-lations of the nonlabor incomes for each country
A Argentina Asignacioacuten Universal porHijo (AUH)
Argentina launched the conditional cash transfer programAUH in 2009 The beneficiaries are children under the ageof 18 in households whose members are unemployed orinformal workers The program provides a monthly cashsubsidy per child up to 5 children Since Argentinarsquos house-hold survey (EPH) does not include a question to identifyrecipients of this program we follow Garganta and Gasparini[13] assuming full take-up among those who qualify Thesimulation is made according to the following rule
If119867ℎ119905(1205781199051015840) + 119876
ℎ119905le 5 and119867
ℎ119905+ 119876ℎ119905
le 5
119884119873119871
ℎ119905(1205781199051015840) = 119884
119873119871
ℎ119905+ 120579119905(119867ℎ119905(1205781199051015840) minus 119867
ℎ119905) (A1)
If119867ℎ119905(1205781199051015840) + 119876
ℎ119905gt 5 and119867
ℎ119905+ 119876ℎ119905
le 5
119884119873119871
ℎ119905(1205781199051015840) = 119884
119873119871
ℎ119905+ 120579119905(5 minus 119867
ℎ119905minus 119876ℎ119905) (A2)
If119867ℎ119905(1205781199051015840) + 119876
ℎ119905le 5 and119867
ℎ119905+ 119876ℎ119905
gt 5
119884119873119871
ℎ119905(1205781199051015840) = 119884
119873119871
ℎ119905+ 120579119905(119867ℎ119905(1205781199051015840) + 119876
ℎ119905minus 5) (A3)
If119867ℎ119905(1205781199051015840) + 119876
ℎ119905gt 5 and119867
ℎ119905+ 119876ℎ119905
gt 5
119884119873119871
ℎ119905(1205781199051015840) = 119884
119873119871
ℎ119905 (A4)
10 International Journal of Population Research
where 119876ℎ119905
is the number of children aged 17 or 18(recall that we carry out the simulations for childrenaged 16 or younger) and 120579
119905is equal to AR$270 for
the third trimester of 2012 and AR$340 for the fourthtrimester
B Brazil Bolsa Familia Program (BFP)
The Bolsa Familia implemented in 2003 is the flagshipconditional cash transfer program in Brazil aimed at poorhouseholds The estimated nonlabor incomes in this case arecalculated as follows
If119867119905ℎ
gt 0 and beneficiary
119884119873119871
ℎ119905(1205781199051015840) = 119884
119873119871
119905ℎ+ 119872119905ℎ(119867119905ℎ(1205781199051015840)
119867119905ℎ
minus 1) (B1)
If119867119905ℎ
= 0119867119905ℎ(1205781199051015840) = 119896 and beneficiary
119884119873119871
ℎ119905(1205781199051015840) = 119884
119873119871
119905ℎ+ 119872119896
119905ℎminus 119897119905ℎ (B2)
with 119896 = 1 2 3 119872119905ℎ
being the amount of thetransfer119872119896
119905ℎbeing the average benefit of households
with 119896 children and 119897119905ℎbeing the standard benefit of
recipients without children
C Mexico Oportunidades
Oportunidades is the main antipoverty program of theMexican Government The size of the transfer depends onthe number of children among other characteristics of thehouseholdMexicorsquos ENIGH identifies the beneficiaries of theprogram and the amount of the subsidy that each householdreceives The simulation is made as follows
119884119873119871
ℎ119905(1205781199051015840) = 119884
119873119871
119905ℎ+ 119872119896
119905ℎminus 119872119905ℎ
if 119867119905ℎ(1205781199051015840) = 119896 119867
119905ℎ= 119896 and beneficiary
(C1)
where 119872119896119905ℎis the average transfer of a beneficiary household
with 119896 children and119872119905ℎis the original household transfer
D Uruguay Asignaciones Familiares
The Asignaciones Familiares program in Uruguay is targetedat the children of workers The amount of the transfer varieswith the householdrsquos income Since the size of the transferdepends on characteristics of the child such as the age thesimulation is carried out as follows
If119867119905ℎ
gt 0 and beneficiary
119884119873119871
ℎ119905(1205781199051015840) = 119884
119873119871
119905ℎ+ 119872119905ℎ(119867119905ℎ(1205781199051015840)
119867119905ℎ
minus 1) (D1)
If119867119905ℎ
= 0119867119905ℎ(1205781199051015840) = 119896 and beneficiary
119884119873119871
ℎ119905(1205781199051015840) = 119884
119873119871
119905ℎ+ 119872119896
119905ℎ (D2)
with 119896 = 1 2 3 119872119905ℎ
being the total amount ofthe transfer and 119872
119896
119905ℎbeing the average transfer of a
household with 119896 children
Disclosure
The information presented herein is solely the responsibilityof the authors and does not necessarily reflect the opinions ofthe aforementioned entities
Competing Interests
The authors declare that they have no competing interests
Acknowledgments
This paper was written in the framework of the projectEnhancing Womenrsquos Economic Empowerment throughBetter Policies in Latin America a joint initiative betweenCEDLAS-UNLP and CIEDUR with the support ofIDRC Canada The paper includes material from NicolasBadaraccorsquos Masters dissertation at Universidad Nacionalde La Plata The authors are grateful to Carolina RobinoGuillermo Cruces Jorge Paz Facundo Albornoz andseminar participants at AAEP (Posadas 2014) Network ofInequality and Poverty (UNGS 2014) and Conference onSocial Protection (IDRC and IPC Brasilia 2014) for valuablecomments and suggestions
References
[1] T W Guinnane ldquoThe historical fertility transition a guide foreconomistsrdquo Journal of Economic Literature vol 49 no 3 pp589ndash614 2011
[2] L E Jones and M Tertilt ldquoAn economic history of fertility inthe US 1826ndash1960rdquo NBERWorking Paper 12796 2006
[3] J Chackiel La Dinamica Demografica en America LatinaSerie Poblacion y Desarrollo No 52 Centro Latinoamericanoy Caribeno de Demografıa-Division de Poblacion de CEPAL2004
[4] ECLAC ldquoLa fecundidad en America Latina un descensoacelerado y heterogeneo con profundas transformacionesdemograficas y socialesrdquo in Fertility Latin American and theCaribbean Demographic Observatory no 5 Centro Latinoa-mericano y Caribeno de DemografıamdashDivision de Poblacionde CEPAL 2008
[5] ECLAC ldquoChapter II current situation and outlook for fertilityin Latin Americardquo in Social Panorama of Latin America 2011ECLAC 2011
[6] L Gasparini and M Marchionni Bridging Gender Gaps TheRise and Deceleration of Female Labor Force ParticipationCEDLAS-UNLP 2015
[7] L Gasparini and N Lustig ldquoThe rise and fall of incomeinequality in Latin Americardquo inThe Oxford Handbook of LatinAmerican Economics J Ocampo and J Ros Eds chapter 27 pp691ndash714 Oxford University Press Oxford UK 2011
[8] G Cornia Ed Falling Inequality in Latin America PolicyChanges and Lessons Oxford University Press 2014
International Journal of Population Research 11
[9] F Alvaredo and L Gasparini ldquoRecent trends in inequalityand poverty in developing countriesrdquo in Handbook of IncomeDistribution Atkinson and Bourguignon Eds vol 2 chapter 9pp 697ndash805 Elsevier New York NY USA 2015
[10] M Marchionni and L Gasparini ldquoTracing out the effects ofdemographic changes on the income distribution The caseof Greater Buenos Aires 1980ndash1998rdquo Journal of EconomicInequality vol 5 no 1 pp 97ndash114 2007
[11] J M Wooldridge Econometric Analysis of Cross Section andPanel Data MIT Press Boston Mass USA 2002
[12] L Gasparini M Marchionni and W Sosa Escudero ldquoChar-acterization of inequality changes through microeconometricdecompositions The case of Greater Buenos Airesrdquo in TheMicroeconomics of Income Distribution Dynamics in East Asiaand Latina America F Bourguignon F Ferreira and N LustigEds Oxford University Press New York NY USA 2004
[13] S Garganta and L Gasparini ldquoThe impact of a social programon labor informality the case of AUH in Argentinardquo Journal ofDevelopment Economics vol 115 pp 99ndash110 2015
Submit your manuscripts athttpwwwhindawicom
Child Development Research
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Education Research International
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Biomedical EducationJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Psychiatry Journal
ArchaeologyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
AnthropologyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Research and TreatmentSchizophrenia
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Urban Studies Research
Population ResearchInternational Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CriminologyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Aging ResearchJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
NursingResearch and Practice
Current Gerontologyamp Geriatrics Research
Hindawi Publishing Corporationhttpwwwhindawicom
Volume 2014
Sleep DisordersHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
AddictionJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Depression Research and TreatmentHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Geography Journal
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Research and TreatmentAutism
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Economics Research International
International Journal of Population Research 9
reduction in fertility across socioeconomic groups in Brazilcontributed to the observed decline in inequality in the lasttwo decadesThe overall effect ofminus132 points in column (v) isthe result of a significant direct effect (minus130) and two smallereffects going in different directions
Some interesting general results emerge from Table 3First changes in fertility have implied a decline in incomeinequality The differential pattern in fertility across groupsexperienced in most Latin American countries over thelast decades translated into an equalizing impact on theincome distribution Second this effect is small althoughnot negligible To be sure demographic changes are not thecentral reason behind changes in income inequality but theyare statistically significant and economically relevant Onaverage fertility changes account for a fall of around onepoint in the Gini coefficient
Third most of the impact comes from the direct effectThe differential fall in fertility rates among socioeconomicgroups had a larger impact on the family size of poorerfamilies implying a proportionally larger increasing effect ontheir per capita incomes
Fourth the effect of fertility changes on hours of workand in turn on incomes is smaller and has different signsacross countries It is important to notice that a more intensereduction in fertility among the poor may be consistentwith a positive sign (inequality-increasing) for this effectThis could happen if the elasticity of hours of work withrespect to the number of children is higher among thenonpoor and if the extra hours worked triggered by the fallin fertility imply a household income increase proportionallylarger for the nonpoor Anyway with one exception thehours-of-work effect has a negative sign implying that theasymmetric reduction in fertility contributed to a reductionin income inequality by fostering higher participation in thelabor market among poor parents (mostly mothers)
Finally the effect through nonlabor incomes is alwaysinequality-increasing but small The fall in fertility amongthe poor implied a lower inequality-reducing impact of theconditional cash transfer programs that were introduced andexpanded in the region mainly during the 2000s The effecthowever is in most cases very small
5 Concluding Remarks
Poverty and inequality fell in Latin America over the lastdecades driven by a large number of factors In thispaper we highlight one factor that is usually overlooked inthe economic literature the impact of changes in fertilityThe average number of children fell in households fromall groups but especially among the more vulnerable afact that could have distributive implications We quantifythe consequences of this demographic pattern by applyingmicrosimulation techniquesThe results suggest that changesin fertility that took place over the last decades in LatinAmerica contributed to the reduction in income inequalityas well as in poverty The impacts found were in generalstatistically significant and economically relevant althoughnot very large Although surely there are more importantdeterminants behind the observed patterns in the income
distributions of the Latin American countries demographicfactors should not be overlooked
The paper has focused on three simple yet central chan-nels the change in the number of children (i) enlarges thefamily size and reduces per capita income (ii) affects thelabor supply decisions and (iii) modifies the income supportfrom social programs Certainly there are other potentiallyrelevant channels that are ignored in the paper For instancea reduction in the number of children per family couldimply higher inheritances and higher family spending oneducation and health per child and hence better incomeperspectives Also smaller families may be better positionedto take advantage of labor market opportunities for instancethrough migration Moreover lower fertility rates may beassociated with a reduction in the relative supply of unskilledlabor and hence an increase in its relative wage
Although in the short and medium run lower fertility inmore disadvantaged households may contribute to reducingpoverty and inequality in the long run the implicationsare more nuanced When children become adults they maycontribute to their parentsrsquo incomes a fall in fertility mayreduce incomes for the elderly especially in countries wherethe pension system is weak and ultimately contribute tohigher poverty through that intertemporal channel
Appendix
In this appendix we describe the issues regarding the simu-lations of the nonlabor incomes for each country
A Argentina Asignacioacuten Universal porHijo (AUH)
Argentina launched the conditional cash transfer programAUH in 2009 The beneficiaries are children under the ageof 18 in households whose members are unemployed orinformal workers The program provides a monthly cashsubsidy per child up to 5 children Since Argentinarsquos house-hold survey (EPH) does not include a question to identifyrecipients of this program we follow Garganta and Gasparini[13] assuming full take-up among those who qualify Thesimulation is made according to the following rule
If119867ℎ119905(1205781199051015840) + 119876
ℎ119905le 5 and119867
ℎ119905+ 119876ℎ119905
le 5
119884119873119871
ℎ119905(1205781199051015840) = 119884
119873119871
ℎ119905+ 120579119905(119867ℎ119905(1205781199051015840) minus 119867
ℎ119905) (A1)
If119867ℎ119905(1205781199051015840) + 119876
ℎ119905gt 5 and119867
ℎ119905+ 119876ℎ119905
le 5
119884119873119871
ℎ119905(1205781199051015840) = 119884
119873119871
ℎ119905+ 120579119905(5 minus 119867
ℎ119905minus 119876ℎ119905) (A2)
If119867ℎ119905(1205781199051015840) + 119876
ℎ119905le 5 and119867
ℎ119905+ 119876ℎ119905
gt 5
119884119873119871
ℎ119905(1205781199051015840) = 119884
119873119871
ℎ119905+ 120579119905(119867ℎ119905(1205781199051015840) + 119876
ℎ119905minus 5) (A3)
If119867ℎ119905(1205781199051015840) + 119876
ℎ119905gt 5 and119867
ℎ119905+ 119876ℎ119905
gt 5
119884119873119871
ℎ119905(1205781199051015840) = 119884
119873119871
ℎ119905 (A4)
10 International Journal of Population Research
where 119876ℎ119905
is the number of children aged 17 or 18(recall that we carry out the simulations for childrenaged 16 or younger) and 120579
119905is equal to AR$270 for
the third trimester of 2012 and AR$340 for the fourthtrimester
B Brazil Bolsa Familia Program (BFP)
The Bolsa Familia implemented in 2003 is the flagshipconditional cash transfer program in Brazil aimed at poorhouseholds The estimated nonlabor incomes in this case arecalculated as follows
If119867119905ℎ
gt 0 and beneficiary
119884119873119871
ℎ119905(1205781199051015840) = 119884
119873119871
119905ℎ+ 119872119905ℎ(119867119905ℎ(1205781199051015840)
119867119905ℎ
minus 1) (B1)
If119867119905ℎ
= 0119867119905ℎ(1205781199051015840) = 119896 and beneficiary
119884119873119871
ℎ119905(1205781199051015840) = 119884
119873119871
119905ℎ+ 119872119896
119905ℎminus 119897119905ℎ (B2)
with 119896 = 1 2 3 119872119905ℎ
being the amount of thetransfer119872119896
119905ℎbeing the average benefit of households
with 119896 children and 119897119905ℎbeing the standard benefit of
recipients without children
C Mexico Oportunidades
Oportunidades is the main antipoverty program of theMexican Government The size of the transfer depends onthe number of children among other characteristics of thehouseholdMexicorsquos ENIGH identifies the beneficiaries of theprogram and the amount of the subsidy that each householdreceives The simulation is made as follows
119884119873119871
ℎ119905(1205781199051015840) = 119884
119873119871
119905ℎ+ 119872119896
119905ℎminus 119872119905ℎ
if 119867119905ℎ(1205781199051015840) = 119896 119867
119905ℎ= 119896 and beneficiary
(C1)
where 119872119896119905ℎis the average transfer of a beneficiary household
with 119896 children and119872119905ℎis the original household transfer
D Uruguay Asignaciones Familiares
The Asignaciones Familiares program in Uruguay is targetedat the children of workers The amount of the transfer varieswith the householdrsquos income Since the size of the transferdepends on characteristics of the child such as the age thesimulation is carried out as follows
If119867119905ℎ
gt 0 and beneficiary
119884119873119871
ℎ119905(1205781199051015840) = 119884
119873119871
119905ℎ+ 119872119905ℎ(119867119905ℎ(1205781199051015840)
119867119905ℎ
minus 1) (D1)
If119867119905ℎ
= 0119867119905ℎ(1205781199051015840) = 119896 and beneficiary
119884119873119871
ℎ119905(1205781199051015840) = 119884
119873119871
119905ℎ+ 119872119896
119905ℎ (D2)
with 119896 = 1 2 3 119872119905ℎ
being the total amount ofthe transfer and 119872
119896
119905ℎbeing the average transfer of a
household with 119896 children
Disclosure
The information presented herein is solely the responsibilityof the authors and does not necessarily reflect the opinions ofthe aforementioned entities
Competing Interests
The authors declare that they have no competing interests
Acknowledgments
This paper was written in the framework of the projectEnhancing Womenrsquos Economic Empowerment throughBetter Policies in Latin America a joint initiative betweenCEDLAS-UNLP and CIEDUR with the support ofIDRC Canada The paper includes material from NicolasBadaraccorsquos Masters dissertation at Universidad Nacionalde La Plata The authors are grateful to Carolina RobinoGuillermo Cruces Jorge Paz Facundo Albornoz andseminar participants at AAEP (Posadas 2014) Network ofInequality and Poverty (UNGS 2014) and Conference onSocial Protection (IDRC and IPC Brasilia 2014) for valuablecomments and suggestions
References
[1] T W Guinnane ldquoThe historical fertility transition a guide foreconomistsrdquo Journal of Economic Literature vol 49 no 3 pp589ndash614 2011
[2] L E Jones and M Tertilt ldquoAn economic history of fertility inthe US 1826ndash1960rdquo NBERWorking Paper 12796 2006
[3] J Chackiel La Dinamica Demografica en America LatinaSerie Poblacion y Desarrollo No 52 Centro Latinoamericanoy Caribeno de Demografıa-Division de Poblacion de CEPAL2004
[4] ECLAC ldquoLa fecundidad en America Latina un descensoacelerado y heterogeneo con profundas transformacionesdemograficas y socialesrdquo in Fertility Latin American and theCaribbean Demographic Observatory no 5 Centro Latinoa-mericano y Caribeno de DemografıamdashDivision de Poblacionde CEPAL 2008
[5] ECLAC ldquoChapter II current situation and outlook for fertilityin Latin Americardquo in Social Panorama of Latin America 2011ECLAC 2011
[6] L Gasparini and M Marchionni Bridging Gender Gaps TheRise and Deceleration of Female Labor Force ParticipationCEDLAS-UNLP 2015
[7] L Gasparini and N Lustig ldquoThe rise and fall of incomeinequality in Latin Americardquo inThe Oxford Handbook of LatinAmerican Economics J Ocampo and J Ros Eds chapter 27 pp691ndash714 Oxford University Press Oxford UK 2011
[8] G Cornia Ed Falling Inequality in Latin America PolicyChanges and Lessons Oxford University Press 2014
International Journal of Population Research 11
[9] F Alvaredo and L Gasparini ldquoRecent trends in inequalityand poverty in developing countriesrdquo in Handbook of IncomeDistribution Atkinson and Bourguignon Eds vol 2 chapter 9pp 697ndash805 Elsevier New York NY USA 2015
[10] M Marchionni and L Gasparini ldquoTracing out the effects ofdemographic changes on the income distribution The caseof Greater Buenos Aires 1980ndash1998rdquo Journal of EconomicInequality vol 5 no 1 pp 97ndash114 2007
[11] J M Wooldridge Econometric Analysis of Cross Section andPanel Data MIT Press Boston Mass USA 2002
[12] L Gasparini M Marchionni and W Sosa Escudero ldquoChar-acterization of inequality changes through microeconometricdecompositions The case of Greater Buenos Airesrdquo in TheMicroeconomics of Income Distribution Dynamics in East Asiaand Latina America F Bourguignon F Ferreira and N LustigEds Oxford University Press New York NY USA 2004
[13] S Garganta and L Gasparini ldquoThe impact of a social programon labor informality the case of AUH in Argentinardquo Journal ofDevelopment Economics vol 115 pp 99ndash110 2015
Submit your manuscripts athttpwwwhindawicom
Child Development Research
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Education Research International
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Biomedical EducationJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Psychiatry Journal
ArchaeologyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
AnthropologyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Research and TreatmentSchizophrenia
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Urban Studies Research
Population ResearchInternational Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CriminologyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Aging ResearchJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
NursingResearch and Practice
Current Gerontologyamp Geriatrics Research
Hindawi Publishing Corporationhttpwwwhindawicom
Volume 2014
Sleep DisordersHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
AddictionJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Depression Research and TreatmentHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Geography Journal
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Research and TreatmentAutism
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Economics Research International
10 International Journal of Population Research
where 119876ℎ119905
is the number of children aged 17 or 18(recall that we carry out the simulations for childrenaged 16 or younger) and 120579
119905is equal to AR$270 for
the third trimester of 2012 and AR$340 for the fourthtrimester
B Brazil Bolsa Familia Program (BFP)
The Bolsa Familia implemented in 2003 is the flagshipconditional cash transfer program in Brazil aimed at poorhouseholds The estimated nonlabor incomes in this case arecalculated as follows
If119867119905ℎ
gt 0 and beneficiary
119884119873119871
ℎ119905(1205781199051015840) = 119884
119873119871
119905ℎ+ 119872119905ℎ(119867119905ℎ(1205781199051015840)
119867119905ℎ
minus 1) (B1)
If119867119905ℎ
= 0119867119905ℎ(1205781199051015840) = 119896 and beneficiary
119884119873119871
ℎ119905(1205781199051015840) = 119884
119873119871
119905ℎ+ 119872119896
119905ℎminus 119897119905ℎ (B2)
with 119896 = 1 2 3 119872119905ℎ
being the amount of thetransfer119872119896
119905ℎbeing the average benefit of households
with 119896 children and 119897119905ℎbeing the standard benefit of
recipients without children
C Mexico Oportunidades
Oportunidades is the main antipoverty program of theMexican Government The size of the transfer depends onthe number of children among other characteristics of thehouseholdMexicorsquos ENIGH identifies the beneficiaries of theprogram and the amount of the subsidy that each householdreceives The simulation is made as follows
119884119873119871
ℎ119905(1205781199051015840) = 119884
119873119871
119905ℎ+ 119872119896
119905ℎminus 119872119905ℎ
if 119867119905ℎ(1205781199051015840) = 119896 119867
119905ℎ= 119896 and beneficiary
(C1)
where 119872119896119905ℎis the average transfer of a beneficiary household
with 119896 children and119872119905ℎis the original household transfer
D Uruguay Asignaciones Familiares
The Asignaciones Familiares program in Uruguay is targetedat the children of workers The amount of the transfer varieswith the householdrsquos income Since the size of the transferdepends on characteristics of the child such as the age thesimulation is carried out as follows
If119867119905ℎ
gt 0 and beneficiary
119884119873119871
ℎ119905(1205781199051015840) = 119884
119873119871
119905ℎ+ 119872119905ℎ(119867119905ℎ(1205781199051015840)
119867119905ℎ
minus 1) (D1)
If119867119905ℎ
= 0119867119905ℎ(1205781199051015840) = 119896 and beneficiary
119884119873119871
ℎ119905(1205781199051015840) = 119884
119873119871
119905ℎ+ 119872119896
119905ℎ (D2)
with 119896 = 1 2 3 119872119905ℎ
being the total amount ofthe transfer and 119872
119896
119905ℎbeing the average transfer of a
household with 119896 children
Disclosure
The information presented herein is solely the responsibilityof the authors and does not necessarily reflect the opinions ofthe aforementioned entities
Competing Interests
The authors declare that they have no competing interests
Acknowledgments
This paper was written in the framework of the projectEnhancing Womenrsquos Economic Empowerment throughBetter Policies in Latin America a joint initiative betweenCEDLAS-UNLP and CIEDUR with the support ofIDRC Canada The paper includes material from NicolasBadaraccorsquos Masters dissertation at Universidad Nacionalde La Plata The authors are grateful to Carolina RobinoGuillermo Cruces Jorge Paz Facundo Albornoz andseminar participants at AAEP (Posadas 2014) Network ofInequality and Poverty (UNGS 2014) and Conference onSocial Protection (IDRC and IPC Brasilia 2014) for valuablecomments and suggestions
References
[1] T W Guinnane ldquoThe historical fertility transition a guide foreconomistsrdquo Journal of Economic Literature vol 49 no 3 pp589ndash614 2011
[2] L E Jones and M Tertilt ldquoAn economic history of fertility inthe US 1826ndash1960rdquo NBERWorking Paper 12796 2006
[3] J Chackiel La Dinamica Demografica en America LatinaSerie Poblacion y Desarrollo No 52 Centro Latinoamericanoy Caribeno de Demografıa-Division de Poblacion de CEPAL2004
[4] ECLAC ldquoLa fecundidad en America Latina un descensoacelerado y heterogeneo con profundas transformacionesdemograficas y socialesrdquo in Fertility Latin American and theCaribbean Demographic Observatory no 5 Centro Latinoa-mericano y Caribeno de DemografıamdashDivision de Poblacionde CEPAL 2008
[5] ECLAC ldquoChapter II current situation and outlook for fertilityin Latin Americardquo in Social Panorama of Latin America 2011ECLAC 2011
[6] L Gasparini and M Marchionni Bridging Gender Gaps TheRise and Deceleration of Female Labor Force ParticipationCEDLAS-UNLP 2015
[7] L Gasparini and N Lustig ldquoThe rise and fall of incomeinequality in Latin Americardquo inThe Oxford Handbook of LatinAmerican Economics J Ocampo and J Ros Eds chapter 27 pp691ndash714 Oxford University Press Oxford UK 2011
[8] G Cornia Ed Falling Inequality in Latin America PolicyChanges and Lessons Oxford University Press 2014
International Journal of Population Research 11
[9] F Alvaredo and L Gasparini ldquoRecent trends in inequalityand poverty in developing countriesrdquo in Handbook of IncomeDistribution Atkinson and Bourguignon Eds vol 2 chapter 9pp 697ndash805 Elsevier New York NY USA 2015
[10] M Marchionni and L Gasparini ldquoTracing out the effects ofdemographic changes on the income distribution The caseof Greater Buenos Aires 1980ndash1998rdquo Journal of EconomicInequality vol 5 no 1 pp 97ndash114 2007
[11] J M Wooldridge Econometric Analysis of Cross Section andPanel Data MIT Press Boston Mass USA 2002
[12] L Gasparini M Marchionni and W Sosa Escudero ldquoChar-acterization of inequality changes through microeconometricdecompositions The case of Greater Buenos Airesrdquo in TheMicroeconomics of Income Distribution Dynamics in East Asiaand Latina America F Bourguignon F Ferreira and N LustigEds Oxford University Press New York NY USA 2004
[13] S Garganta and L Gasparini ldquoThe impact of a social programon labor informality the case of AUH in Argentinardquo Journal ofDevelopment Economics vol 115 pp 99ndash110 2015
Submit your manuscripts athttpwwwhindawicom
Child Development Research
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Education Research International
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Biomedical EducationJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Psychiatry Journal
ArchaeologyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
AnthropologyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Research and TreatmentSchizophrenia
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Urban Studies Research
Population ResearchInternational Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CriminologyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Aging ResearchJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
NursingResearch and Practice
Current Gerontologyamp Geriatrics Research
Hindawi Publishing Corporationhttpwwwhindawicom
Volume 2014
Sleep DisordersHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
AddictionJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Depression Research and TreatmentHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Geography Journal
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Research and TreatmentAutism
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Economics Research International
International Journal of Population Research 11
[9] F Alvaredo and L Gasparini ldquoRecent trends in inequalityand poverty in developing countriesrdquo in Handbook of IncomeDistribution Atkinson and Bourguignon Eds vol 2 chapter 9pp 697ndash805 Elsevier New York NY USA 2015
[10] M Marchionni and L Gasparini ldquoTracing out the effects ofdemographic changes on the income distribution The caseof Greater Buenos Aires 1980ndash1998rdquo Journal of EconomicInequality vol 5 no 1 pp 97ndash114 2007
[11] J M Wooldridge Econometric Analysis of Cross Section andPanel Data MIT Press Boston Mass USA 2002
[12] L Gasparini M Marchionni and W Sosa Escudero ldquoChar-acterization of inequality changes through microeconometricdecompositions The case of Greater Buenos Airesrdquo in TheMicroeconomics of Income Distribution Dynamics in East Asiaand Latina America F Bourguignon F Ferreira and N LustigEds Oxford University Press New York NY USA 2004
[13] S Garganta and L Gasparini ldquoThe impact of a social programon labor informality the case of AUH in Argentinardquo Journal ofDevelopment Economics vol 115 pp 99ndash110 2015
Submit your manuscripts athttpwwwhindawicom
Child Development Research
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Education Research International
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Biomedical EducationJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Psychiatry Journal
ArchaeologyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
AnthropologyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Research and TreatmentSchizophrenia
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Urban Studies Research
Population ResearchInternational Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CriminologyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Aging ResearchJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
NursingResearch and Practice
Current Gerontologyamp Geriatrics Research
Hindawi Publishing Corporationhttpwwwhindawicom
Volume 2014
Sleep DisordersHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
AddictionJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Depression Research and TreatmentHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Geography Journal
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Research and TreatmentAutism
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Economics Research International
Submit your manuscripts athttpwwwhindawicom
Child Development Research
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Education Research International
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Biomedical EducationJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Psychiatry Journal
ArchaeologyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
AnthropologyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Research and TreatmentSchizophrenia
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Urban Studies Research
Population ResearchInternational Journal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
CriminologyJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Aging ResearchJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
NursingResearch and Practice
Current Gerontologyamp Geriatrics Research
Hindawi Publishing Corporationhttpwwwhindawicom
Volume 2014
Sleep DisordersHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
AddictionJournal of
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Depression Research and TreatmentHindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Geography Journal
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Research and TreatmentAutism
Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014
Economics Research International
