4Ai& t Tunisia Social Aspects of Development · 2016. 7. 15. · Report No. 295-TUN F 4Ai& ¾A t...

151
Report No. 295-TUN F 4Ai& ¾A t Tunisia Social Aspects of Development June 18, 1980 Europe, Middle East and North Africa Regional Office Country Programs Department II FOR OFFICIALUSEONLY Document of the World Bank This document has a restricted distribution andmay be used by recipients only in theperformance of their official duties. Its contents may not otherwise bedisclosed withoutWorldBark authorization. Public Disclosure Authorized Public Disclosure Authorized Public Disclosure Authorized Public Disclosure Authorized Public Disclosure Authorized Public Disclosure Authorized Public Disclosure Authorized Public Disclosure Authorized

Transcript of 4Ai& t Tunisia Social Aspects of Development · 2016. 7. 15. · Report No. 295-TUN F 4Ai& ¾A t...

  • Report No. 295-TUN F 4Ai& ¾A t

    TunisiaSocial Aspects of Development

    June 18, 1980

    Europe, Middle East and North Africa Regional OfficeCountry Programs Department II

    FOR OFFICIAL USE ONLY

    Document of the World Bank

    This document has a restricted distribution and may be used by recipientsonly in the performance of their official duties. Its contents may nototherwise be disclosed without World Bark authorization.

    Pub

    lic D

    iscl

    osur

    e A

    utho

    rized

    Pub

    lic D

    iscl

    osur

    e A

    utho

    rized

    Pub

    lic D

    iscl

    osur

    e A

    utho

    rized

    Pub

    lic D

    iscl

    osur

    e A

    utho

    rized

    Pub

    lic D

    iscl

    osur

    e A

    utho

    rized

    Pub

    lic D

    iscl

    osur

    e A

    utho

    rized

    Pub

    lic D

    iscl

    osur

    e A

    utho

    rized

    Pub

    lic D

    iscl

    osur

    e A

    utho

    rized

  • CURE17NCY EQUIVALENTS

    Currency Unit Dinar = 1000 Millimes

    Dinars per US Dollar

    Period End of Period Period Average

    1970 0.5208 0.52081971 0.4807 0.51571972 0.4840 0.47721973 0.4451 0.42001974 0.4065 0.43651975 0.4253 0.40231976 0.4309 0.42881977 0.4121 0.42901978 0.4034 0.41621979 0.3959 0.4065

    Source: Iff, International Financial Statistics, April. 1980.

  • FOR OFFICIAL USE ONLY

    TUNISIA

    SOCIAL ASPECTS OF DEVELOPMENT

    Table of Contents

    Page No.

    SUMMARY AND CONCLUSIONS .............................. i-xi

    I. INTRODUCTION ......................................... 1

    II. GENERAL BACKGROUND ...... ............................. 3

    A. Development Objectives ..... ...................... 3B. Macro-Economic Performance ......... . .............. 5C. Social Achievements and Failures ................. 7

    III. THE POVERTY CONCEPT ...................... ..... ........ . 9

    IV. SOCIAL PROGRAMS ......... .. .. ............................ .... 14

    A. Social Services . ..................... .. ......... . 151. Education ................ .................... 162. Health .................. ..................... 223. Family Planning ............. .. ............... 274. Water Supply and Sewerage ................... 285. Nutrition . ................................... 296. Housing .............................................. 32

    B. Incomes Policy ....... ................... *... 351. Minimum Wages .......................... . 362. Price Support ...... .......................... 373. Social Security .............................. 434. Tax Policy .................................................. 43

    C. Regional and Rural Development ................. . 461. The Rural Development Fund ................... 462. Regional Development Poles ......... .. ........ 473. Other Regional Development Measures .......... 48

    Annex 1 - INCOME DISTRIBUTION: STRUCTURE AND TRENDSI. Distribution ObjectivesII. Recent DevelopmentsIII. Conclusions

    Annex 2 - TAX INCIDENCE IN TUNISIAI. IntroductionII. Method and Problems of Analysis

    This report was prepared by Werner M. Schelzig. Annex 2 is based on a contri-bution by Walter B. Stolber.

    This document has a restricted distribution and may be used by recipients only in the performanceof their official duties. Its contents may not otherwise be disclosed without World Bank authorization.

  • Table of Contents (Continued)

    III. Tax IncidenceA. Selection of TaxesB. Shifting AssumptionsC. Result: Quantitative Estimates of Tax Incidence

    IV. Summary and RecommendationAppendix 1 - Details on Incidence Assumptions of Specific Taxes

    Annex 3 - THE UNEMPLOYMENT PROBLEMI. IntroductionII. The Definition

    III. The ProblemIV. The Characteristics of the UnemployedV. The CausesVI. Conclusion

    STATISTICAL APPENDIX

    BIBLIOGRAPHY

  • SUMMARY AND CONCLUSIONS

    GENERAL

    i. This report analyzes the social aspects of development in Tunisia.Its main concern is with what is being done to cover the basic needs of thepopulation. The report proceeds essentially on a pretheoretical stage ofanalysis in which evidence is collected and examined to gain a basic under-standing of the salient features of the issues involved. At this stage,no efforts are being made to develop formalistic determination theoriesbased on behavioristic hypotheses. This will have to await further study.The main purpose of this report is to provide the Tunisian authorities andthe World Bank with useful background material for better assessing theexisting social situation, to evaluate past success and failures, and toderive the lessons to be learned from past experience with a view to makingpolicy adjustments.

    ii. The merit of this report lies in bringing a wealth of social informa-tion together and reviewing the effect of existing policies on the targetgroups. Invariably, a study such as this one gives rise to many additionalquestions about further details and related facts and figures. An initialeffort, like the present one, grappling with many different aspects of a verycomplex subject, simply cannot be exhaustive. Nor can the report claim tohave instant solutions for all issues identified. A great deal more analysisis needed to farthom the social interrelationships and to device a balancedset of policies to deal with them. Nonetheless, a number of conclusions arederived here and suggestions are made which should improve the existing socialsituation.

    SOCIAL PERFORMANCE

    iii. In general, Tunisia's social performance is very impressive. Fewcountries in the developing world have undertaken the degree of deliberateeffort towards social change that Tunisia has since independence in 1956.With far reaching legal and educational reforms, rapid economic growthand a comprehensive set of policy measures, the authorities encouragedmassive transformation of society. This created a shared commitment tomodernization itself in all quarters of society. A pragmatic approach waschosen to tackle the problems of social and economic development. Thecontinuity of political leadership, the willingness of the authoritiesto introduce structural changes in the economic system, and the determina-tion shown in pursuit of development have greatly contributed to creatingan environment in which the economy could prosper as planned. And fromthe very beginning the authorities were concerned with the social aspectsof development as well. As a result, Tunisia has come a long way in itsprogress towards economic and social development, the coverage of basicneeds of its population and reducing absolute poverty.

    iv. Overall social progress as measured by just about every thinkablesocial indicator is very good, certainly fully commensurate with Tunisia's percapita income level and often much better than that of many countries withcomparable per capita income levels. This is a direct outcome of rapidly

  • - ii -

    rising per capita income growth sustained over long periods and supplementedwith comprehensive social programs and policies. About 15 percent of GDP isbeing devoted to social programs. This is very much by international standardswhere a level of 10 percent is generally regarded as being very substantial.The fact that much of the expenditures devoted to these programs are biasedtowards the upper income groups, mostly in urban centers, is not surprising.This is so in most other countries too. However, by allocating the socialservices predominantly to those who pay for them through taxation, the flow ofbenefits to the poor is reduced to a small, although by no means negligible,trickle. While performance was indeed good, there remains still plenty ofroom for further improvement. While absolute poverty is diminishing rapidly,there is little cause for complacency. Absolute poverty, as defined in thisreport, remains a problem affecting one out of six people in Tunisia. Rapidoverall growth did not create sufficient opportunities for everyone. Compre-hensive redistribution policies did not reach the target groups by as much asneeded. This means that the great resources allocated to social programs arenot properly targetted to reach the poor.

    TOWARDS A SOCIAL STRATEGY

    v. The basic social issue therefore is to use available resourcesmore efficiently through improving the delivery system for basic needs. Also,looking at sectoral and regional distribution data, available informationclearly shows that the differences between groups are much higher than dif-ferences within these groups. Thus, simply ensuring mobility would havedramatic effects on overall income distribution. A pre-condition for suchmobility is that the supporting policy measures are not contradicting eachother as they can be shown to be in the case of employment creation andprice subsidies. Employment creation and an efficient price system may wellbe the key to solving the remaining social problems as well.

    vi. This report argues in favor of expressing the social objectives interms of consumption of basic needs goods and services rather than in termsof income distribution as such, in order not to disrupt market incentives forprivate savings, investment and work efforts. The programs to achieve basicneeds coverage do exist in Tunisia and are supported by incomes policies.Although commendable they are not without flaws, however. Increased effortsto improve the existing redistributive and basic needs measures are needed.In addition, it appears worthwhile to consider some redistribution of wealthas well. For example, the public ownership of land and enterprises, which nowgenerally contributes little to social welfare, is probably counterproductive.The authorities may wish to rethink their position on the relevance of largepublic holdings at the current stage of economic development. Through repri-vatization it might be possible to redistribute wealth without hurting any-body in particular. Such redistribution of assets was an important cause forsuccessful social performance in some countries (notably Korea and Taiwan)which in the international community generally serve as examples of how othercountries might structure their distributional measures.

  • - iii~ -

    vii. In view of the great unemployment problem (Annex III) facing thecountry, it is obvious that accelerated creation of productive employmentmust be the leading edge in Tunisia's efforts to improve the social situation.Past experience has shown that this cannot be achieved within the existingpolicy structure which is biased towards capital intensive technologies.In the future, much more stress needs to be put on the growth if employmentopportunities in the small scale agricultural, industrial, construction,trade and other sectors. This presupposes an appropriate policy environment.The main elements creating such an environment are the removal of creditconstraints, introducing flexible interest rates, modifying legislation thatdiscourages labor using technologies and the introduction of realistic pricesfor the major market variables from foreign exchange to agricultural products.Also, rural and urban development efforts cannot result in long term struc-tural adjustments unless they move away from general transfers and subsidiestoward self-help and productive activities. The use of simple, low cost solu-tions to preventive and curative health care, water supply, sewerage, shelter,and nutrition should be actively favored by the authorities. This requiresmuch closer coordination among the institutions involved in the various fieldsof social development as is currently the case.

    viii. While the general lines of a basic needs strategy in Tunisia canthus be readily spelled out, much more work is needed to arrive at specificpolicies and projects. It seems to be clear, however, that under the presentcircumstances the social targets are not being tackled efficiently. While itis certainly true that there is a core of activities and instruments whichreinforce each other in accomplishing their social objectives (family planning,nutritional education, social security), there are others which need to bemuch more refined (education shifts towards vocational training, preventivehealth care, rural development, employment creation incentives) to meet theirends, and others again which ought to be abandoned because in their presentform they actually pose obstacles to reaching the objectives (housing sub-sidies, price policies, indiscriminate subsidies to university students,investment incentives favoring capital intensive technologies). In addition,some new activities worth considering might include more general reliance onsite-and-services projects to promote better housing, appropriate supportmeasures for small scale industries, a new institution to promote small scalefarming patterned after FOPRODI and a domestic labor agency patterned afterOTTEEP. Furthermore, some type of national peace corps might be considered,perhaps to be incorporated into the National Work Corps, which would requireall university graduates, especially medical doctors, to serve at least 12months in outlying areas and in social projects before becoming eligible forworking elsewhere. Similarly, central government officials should be rotatedroutinely to local government positions to become eligible for promotion.Finally, the existing provisions for the National Work Crops should be usedactively to provide informal training to the young unemployed. The resourcesto finance such a vast practical training program could be easily mobilized byre-directing funds currently used to subsidize university students.

  • - iv -

    SPECIFIC POLICY ASPECTS

    ix. Existing social policies are comprehensive although not always welldirected. Firstly, there are social services geared towards the basic needsof the population in the following form:

    (i) free education;(ii) free health services and family planning;(iii) public housing, water supply and sewerage;(iv) education in nutrition.

    Secondly, the authorities pursue income policies through:(i) minimum wage legislation;(ii) price support programs for basic consumer goods;(iii) the social security fund;(iv) tax policies.

    Finally, there are separate regional and rural development programs like:(i) the rural development fund;(ii) the regional development poles;

    (iii) FOPRODI and the 1974 investment code as amended in 1978;(iv) land agencies and offices.

    This report shows that this existing set of social policies has some seriousflaws. The following paragraphs summarize the report's major findings in thisrespect and outline how to increase their impact on the poverty group.

    Education

    x. The authorities have always paid special attention to educationas a basic ingredient for economic development and a means for ensuringsocial progress. Education expenditures rank as the most important itemamong budgetary expenditures on social services. The authorities have builtup a comprehensive education system that provides free access to all students,although education is not compulsory. Yet, the education system does not meetthe needs of the economy. Unemployment among graduates is very high. Thismeans that the effect of about 60 percent of the substantial social expendi-ture is not as beneficial to the economy and to the recipients as commonlythought by government. This is especially true for the impact of educationexpenditure on the poor. The high opportunity cost of schooling and the smallimpact of education on productivity tends to keep children of the poorerstrata, and especially females, out of school. The benefits thus accruemostly to the upper income groups in predominantly urban areas. This impliesa need for better coordination of education and society's need, i.e. foreducation having a more direct impact on employability and lifetime earnings,and for drawing children from poorer groups into the system. Instead ofallocating sizeable resources available to intensive programs focussed essen-tially on the urban upper income group, much more should be done in way ofextensive programs to help the poor. And if informal education in Tunisia isindeed as important as shown in this report, then every effort should be madeto render it more efficient. The National Work Corps might be much moreintensively used for extensive training purposes. Most of the funds now spenton cash stipends for university students could be much more productively allo-cated to vocational training.

  • -v-

    Health

    xi. The second most important share of social expenditures goes topublic health services. The overall beneficial effects of the health systemis convincingly demonstrated in the vital statistics. However, the avail-ability of ae health services appears to be strongly correlated to the percapita income of various regions. There are great interregional differencesin the per capita supply of hospital beds, physicians and nursing personnel.Also, since regional and general hospitals are located in the urban centers,there are great urban-rural differentials in available medical resources.There is nothing wrong with the latter, provided the referral system on whichthe health care program is built, is indeed working as intended. However,the capacity of Tunisia's referral structure is not optimally distributedacross regions and districts. The rural poor, in particular, are excludedfrom ready access to the health system through the referral system. This andsevere congestion at the primary clinic level, combined with inadequatecapacity utilization at the district and regional hospitals, suggest a needfor a redirection of health expenditures to the base services. Also, Tunisianhealth services concentrate on curative medicine and only marginally allow forpreventive care. For public health expenditures to become truly effective,greater integration of the preventive and curative medicine is desirable.In addition, there is an apparent duplication of efforts in the public andprivate sector to deliver health services to the population with a third linenow being created by the Social Security Fund building its own hospitals.This is not in the best interest of the country. If indeed public hospitalsprovide inferior service, as it is claimed by the proponents of the twoalternative systems, then it is high time to improve the public system ratherthan to supplement it with additional ones. Finally, infant mortality is veryhigh, especially in the rural sector. About half of all deaths in Tunisia arechildren up to four years old. This is an untenable situation and greatestefforts need to be undertaken to help the most helpless part of the population,the children between zero and four years old.

    Nutrition

    xii. Although the share of the population unable to cover its minimumnutritional requirements is decreasing, malnutrition is still prominent,especially when nutritional components other than calories and proteins areconsidered. It is interesting to note that nutritional deficiencies extendacross income groups and can be found even in the groups with the highestexpenditure levels. Also, nutritional levels vary greatly in intra-familydistribution with very young children and pregnant and lactating women beingat the bottom of the scale. The stubbornly high infant mortality rate isclosely linked to nutritional deficiencies. The Nutritional Institute hasidentified eight major nutritional problems, namely (i) reduced linear growthin young children and immaturity at birth, (ii) high incidence of cardio-vascular illness, (iii) anemia, (iv) caries, (v) vitamin C deficiency, (vi)shortage of riboflavin, (vii) rickets, and (viii) goitre. Improvements couldbe achieved through (i) raising nutritional awareness through educationalprograms, (ii) food fortification through additives (e.g. lysine), (iii)

  • -vi -

    institutional feeding programs, and (iv) food rations replacing the currentlyexisting food subsidies which are distributed indiscriminately across incomegroups. Reaching proper nutritional standards for all will have importantoverall health effects, in reducing anemia (with its clinical effects ofanorexia, asthama, immaturity at birth and anthropometric and mental defi-ciency), marasmum, infections and rickets. Far reaching results may beobtained from expenditures on nutritional education and preventive medicine.The expenditures involved would be very minor compared to those currentlyspent on curative health and infrastructure.

    Housing

    xiii. Tunisia faces a very difficult housing situation. Most of theexisting housing stock is of very poor quality. In addition, demand by faroutstrips supply. The government has taken a leading role in increasingsupply, especially since poor housing has been shown to be an important causefor the high incidence of pulmonary and respiratory diseases rampant inTunisia. Subsidized conventional housing is the cornerstone of public build-ing programs, with subsidies being equivalent to about a third of the capitalvalue of the units built. However, these subsidies generally did not reachthe most deserving target groups. Even the most economical subsidized urbanpublic housing units remained too expensive to be affordable to the lowesthalf of the urban population. In addition, until 1976 such units representedonly about 13 percent of all housing units built. The rural poor are hardlyconsidered at all in these housing schemes, except maybe via rural developmentfund expenditures. The rural poor continue to live essentially in traditionalshelters. The subsidized housing programs of the past were mainly directedto the middle income groups with household incomes substantially above thepoverty limit and more than twice that of workers receiving minimum wages.Possible solutions lie in a rigorous redistribution of public housing funds tothe target groups, especially through the development of sites-and-services,as this would minimize cash requirements of individuals, which now seem to bean important bottleneck. Another policy aspect is the need to abolish rentcontrols in cities. Currently, urban rents of private apartments constructedbefore 1960 are frozen at their original levels plus an annual increase offive percent. This is one of the major reasons why no private sector invest-ment is forthcoming in apartment buildings, why the existing stock is ingeneral disrepair, and why new households cannot find adequate housing.

    Price Support

    xiv. The Government maintains a pervasive system of price controls.Basic foodstuffs for instance are subsidized by the Price Equalization Fund(PEF). Available information indicates that the price subsidy programcarried out by the PEF benefits the middle and upper income groups relativelyand absolutely more than the poor and the urban population more than therural one. The system is thus regressive and unjust. Furthermore, it iscertainly not cost efficient as the available funds allocated do not reach the

  • - vii -

    target groups as desired. For example, preliminary calculations indicate thatthe funds now spent on the PEF would suffice to provide all malnourishedTunisians in the lower income groups with a supplementary diet covering theirbasic food needs. Since the same outlays are far from achieving this underthe existing system, the majority of the resources devoted to it are misallo-cated. This is even more serious as the consumption subsidies seem to favorincreased consumption of imported foodstuffs and generally have been pushingtowards increased private consumption which in view of the country's savingsrequirement was certainly not warranted.

    xv. The price support measures respresent an intervention in the finalproduct market for certain wage goods to maintain the purchasing power essen-tially of the urban population. The rural population, with a high degree ofauto-consumption of basic foodstuffs, is largely by-passed. As there isno intended selectivity between income groups in the provision of these sub-sidies, in theory the financial transfers involved measure both the cost andthe benefits to the economy. That the PEF is mostly financed through ear-marked taxes on olive oil, petroleum products, alcohol and "luxury products",generally paid by the more well-to-do, is of secondary importance. What isimportant is that the transfers of the PEF pre-empt public revenues fromalternative uses while releasing private resources for other disposal. Hence,public savings are probably foregone to increase private consumption. Sincethere is a premium on public savings, the social cost of the subsidy isgreater than the actual transfer. The social benefits are shown to be regres-sive. This characteristic is inevitable in all undifferentiated subsidyprograms. It is the price at which administrative convenience is bought.

    xvi. The report argues against maintaining the PEF in its present form.The authorities' policy of administrative determination of prices is unsettl-ing economically and largely ineffective socially. This is not to say thatfood subsidies ought to be abolished suddenly once and for all. To thecontrary: food subsidies certainly have a role to play in the social policyscheme of Tunisia as important nutritional and health considerations are atstake, i.e. some very basic needs indeed. However, the program needs muchbetter targeting to be efficient. A possible alternative would be foodrations or food stamps distributed to the needy, allowing them to cover theirbasic nutritional needs with some very basic foodstuffs especially designedto reach this goal. The government would then be free to devote most of theresources now allocated to the PEF to productive investments thus increasingproduction which would also be helped by letting market forces find theappropriate price level.

    Redistributive Tax Incidence

    xvii. Since the very beginning of Tunisian planning of social targets andpolicies, taxation was identified as an important instrument for redistribut-ing income. However, despite the assertions of successive Plans that equityconsiderations and income redistribution should be an important feature of thetax system, the tax authorities appear essentially to regard themselves as taxcollectors, with taxes being viewed generally as financing instruments only.The overall fiscal pressure in Tunisia is now, as it has been for the last ten

  • - viii -

    years, high by usual international comparisons based on per capita GDP anddegree of openness of the economy. However, the tax system as such is regres-sive, with the main burden falling on wage and salary earners in the middleincome groups whose taxes are withheld at the source. Although incomes stillare redistributed, the redistributive effect is very mild. The main factorsof inequity are tax evasion by existing tax payers and the substantial numberof potential tax payers who do not even file tax returns. Most of these arelikely to be in the upper income groups.

    xviii. Increasing the government's tax collection capacity appears to be apriority areas for action, for instance through computerizatrion of tax files,training of inspectors, and abolition of a number of minor taxes which costmore to collect than they yield. Specific redistributive issues on therevenue side are (a) the taxation of the higher agricultural incomes under theprogressive income tax (these incomes go presently essentially untaxed) and(b) the tax exemptions granted to purchases of stocks and bonds (while thebasic idea of favoring financial savings to develop a capital market is sound,it remains true that because of traditional reluctance towards this kind ofinvestment middle and low income groups do not take advantage of this taxexemptions). In addition, the existing tax system shows excessively con-fiscatory features like the marginal rates of the progressive income tax whichreaches more than 80 percent in the upper income bracket. They should bereduced since they represent essentially powerful incentive to tax evasion.

    Rural and Regional Development

    xix. Tunisia is characterized by great regional disparities in generaland substantial urban-rural differences in income levels in particular. Theauthorities essentially rely on three types of rural and regional developmentinstruments. First, there is a Rural Development Fund (RDF) through whichrural development projects are being financed, mostly of a social infrastruc-ture type. Second, regional development poles are being developed centeredaround exploiting a region's specific resource. Third, a number of decentral-ization measures have been introduced providing incentives to private investorsto settle away from Tunis and the coast.

    xx. There are two major problems in finding a satisfactory approach toaddressing regional disparities. First, the scarcity of exploitable resources,unfavorable climatic conditions, the greatly uneven distribution of land hold-ings and thin infrastructure, make production based rural development projectsdifficult to implement. Second, much of the population of the deprived areaslive scattered over a vast area. This makes it difficult to bring in socialservices and other infrastructure. Hence, there is a certain preoccupation ofthe authorities' programs on urban areas, where more people can be reachedwith the limited financial resources available through projects addressingbasic needs.

    xxi. The weaknesses of the present RDF programs, stem largely from thelack of planning mechanisms and institutions which would integrate local,regional and national development efforts. Without an overall rural develop-ment framework, it is doubtful whether: (i) the scarce existing resources are

  • - ix -

    being optimally developed and with externalities being fully realized; (ii)the balance of productive and infrastructure investments is optimal withregard to employment creation; (iii) more benefits could not be reaped fromRDF programs by focusing investments on those areas with greater resourcepotential than that available in the less endowed areas; and (iv) effectivedevelopment policies on prices, wages, social transfers, tax credit, landreform could not be improved so as to provide greater impetus to rural devel-opment. While there is need to continue and expand the RDF, (and in partic-ular, participation of local and regional authorities in project design andimplementation), emphasis should also be given to rural development planning atnational and regional levels to (i) identify development bottlenecks in exist-ing institutions and policies, (ii) establish and adopt remedial measures, andset development priorities among sectors and regions; (iii) balance farm andnon-farm investments among regions and identify multisectoral integratedinvestment packages; and (iv) better allocate funds for rural developmentchanneled through the main Tunisian lending institutions and bilateral andmulti-lateral assistance. The essence of all this is that serious efforts areneeded to give a better directional approach to the rural development effortsthat ought to be more directly targetted to reaching rural poor and helpingthem to help themselves.

    xxii. Three regional development programs have been quite successfullycarried out: the development of the Gabes industrial zone, the Djerba tourismdevelopment and the establishment of the Tunis District Authority. In thesecases the authorities came to grips with a variety of problems that weresolved to the benefit of the areas concerned. These examples also show thatthe Government's regional development policies face a basic dilemma: shouldthe development effort be directed to regions where the resources, infrastruc-ture and the people are (i.e. the already relatively developed stretches alongthe coast) or should investments be made in the interior where developablenatural resources are scarce and the necessary infrastructure investmentswould benefit only a small fraction of the total population. In addition, itis very likely that outmigration would increase once government invests inhuman capital creation in the outlying areas.

    xxiii. It is imperative that some equitable solution be found. This wouldrequire in the first instance the Government to consolidate the approach toregional development on the basis of a detailed inventory of the existingresources by region. Second, an opportunity cost analysis would then have toassess the development potential of these resources with due attention beingpaid to the social implications. Finally, investment opportunities would beranked according to priorities and sectoral planning targets would be distri-buted according to regional priorities. However, this would require a con-siderable commitment of human capital and the creation of an appropriateinstitutional and incentive framework. The costs of such a program need tobe assessed against its potential benefits. Economically, the development ofregional development poles based on directly productive investments is mostattractive. Through employment creation incomes would be generated that wouldhave multiplier effects throughout the region. Supplementary social transferscould be used to counteract remaining (or resulting) social imbalances.

  • Unemployment

    xxiv. Unemployment in Tunisia is a multi-facetted problem. One of itsmost striking aspects is that government policies are often counterproductivewhen viewed with respect to their actual impact on employment creation. Mini-mum wage laws as they exist in Tunisia and arbitrarily raising minimum wagesby more than concurrent productivity increases, have serious implications forthe efficient allocation of labor, for output and growth. High effectiveprotection and favorable lending rates introduced a further bias in favor oflabor saving technologies. This puts heavy strains on the balance of paymentsas well. In addition, domestic prices increased thus putting the cost burdenon the consumer, especially that emanating from the generally cost-inefficientoperation of public enterprises. Their very high investment neither wasreflected in equivalent job creation nor in commensurate output growth. Thepervasive system of price controls did, of course, not succeed in stemminginflationary pressures which just manifested themselves in phenomena otherthan immediate increases in consumer prices, for instance in the high andrising current account deficit of the balance of payments and in subsidiesof the Price Equalization Fund and rising unemployment.

    xxv. The labor market as such is working quite well. There definitely isa response of workers to economic stimuli, as shown by internal and externalmigration for instance. However, the labor market is greatly segmented:geographically, by skill and by legal access age in the modern sector. Suchsegmentation implies rationing of high wage employment opportunities. However,the modern sector is not able to absorb all job seekers in productive employ-ment at the required rate. This is at the base of the existing Tunisianunemployment and social problems.

    xxvi. The segmentation of the labor markets led to the observed scarcityof labor in some unprotected sectors and to excess supply and consequentunderutilization of labor in protected sectors. The bias towards capitalintensive technologies increased the demand for specially skilled workers,which could not readily be satisfied. This induced large scale educationefforts. Socially, these soon turned out to be unproductive as many graduatescould not find jobs. Severe mismatches of skills, education levels and jobrequirements were the result. Overall, education has little effect on findinga job in Tunisia. Generally, labor is relatively more expensive than capital.Not only do firms tend to choose the more capital intensive of existing tech-nologies, but more importantly investors tend to direct the search for newtechniques toward increasing capital intensity of production and away fromlabor using ones. This is not where Tunisia's priorities and comparativeadvantages lie. This is very impressively demonstrated also by foreigninvestment in Tunisia under the export promotion law of 1972. These enter-prises which function outside the sphere of domestic economic policies aremuch more labor intensive on average than domestic enterprises and they arehighly competitive.

  • - xi -

    xxvii. As long as the policy bias against labor using technologies exists,the grave Tunisian unemployment problem will remain a structural, not acyclical phenomenon. This has very serious implications for the socialtargets of the government, especially for income distribution. As most ofthe unemployed are young people, savings cannot be their means of subsistance.Even for the not so young unemployed savings are generally non-existent.There is no government scheme of unemployment benefits. The unemployed thussurvive with the help of family or other personal transfers. Often this takesthe form of work sharing and thus disguised unemployment. Or else, peoplehave to emigrate. This used to be an important safety valve for unemploymentpressures, with very beneficial effects for the Tunisian economy in generaland for basic needs and income distribution in particular. However, withthe main receiving labor markets abroad being all but closed to additionalTunisian surplus labor, the outlook for continued emigration is rather dim.

    xxviii. Without in any way wanting to influence the conclusions and recom-mendations of the forthcoming comprehensive study of the ILO/UNDP on theTunisian unemployment problem, there can be little doubt that the authoritieswill have to come to grips with this most important issue. This will have tobegin with measures to revise economic policies which currently are stronglybiased in favor of capital intensive production. Special efforts need to bedirected to increase the productivity of the marginal people in agriculture,services and crafts, which form the "reserve army" of the unemployed. Ruraland urban development directed at productive investments are a possiblesolution. This would also be the key to improving the coverage of basic needsof the target group and income distribution. The existing system of pervasiveadministrative controls and subsidies cannot succeed in changing the struc-tural employment difficulties. In fact, some of these measures might even bethe initial cause for widespread unemployment.

    OUTLOOK

    xxix. The measures suggested in this report can only be accomplished instages, but a national plan should be formulated stating how and when thiscan happen, what are the costs involved and how it can be ensured that thebenefits indeed would reach the target groups. The forthcoming Sixth Develop-ment Plan (1982-86) could become the basic policy paper for such an attempt.A special social planning commission should be established without delay tofocus on basic needs and poverty alleviation through measures that would notentail abrupt and massive disturbances of the economy and with a clear under-standing of the costs involved. Tunisia has the resources to abolish absolutepoverty by the year 2000. It also has a chance to do this by the end of theforthcoming plan in 1986. Much suffering could be spared to many poor peopleif appropriate action were taken now. Substantial increases in social welfarecould be achieved simply by reallocating resources already spent on socialprograms, targetting them more directly to the poverty group.

  • TUNISIA

    SOCIAL ASPECTS OF DEVELOPMENT

    I. INTRODUCTION

    1. It is a generally held view that the promotion of social welfare isthe ultimate goal of economic development. The functional relationship ofeconomic growth on one hand and the generation of social welfare, equitableincome distribution and poverty alleviation on the other hand, are subjectto much controversy in a large and growing body of literature. 1/ Existingtheories run the gamut from the view that rapid growth will in due coursealso lead to the achievement of higher living standards of all people, to thebelief that growth and equity are mutually exclusive targets, since growth (atleast in a market economy) tended to benefit only a select few. Between thesetwo extreme views are those which hold that growth with equity can only beachieved through regulatory state intervention of varying intensity. Just howintense this intervention should be is influenced by ideological preferences.Finally, there are those who bypass the issue of growth and equity and insteadpropose the fulfillment of basic needs of the population as the foremost goalof the development process. 2/

    2. As purely theoretical approaches allow for a number of plausiblesolutions to the welfare problem, the issue has become an empirical one. How-ever, methodological and data weaknesses are such that empirical biases cannota priori be ruled out and the controversy is very much alive. Nonetheless,the majority of the available studies seems to indicate that the benefits ofdevelopment in most cases have been distributed quite unevenly and in partic-ular have reached the poor only to some limited extent. 3/ It thus has becomewidely accepted that economic growth generally bypasses the poor. If theirreal incomes increase at all, they typically increase more slowly than theaverage. Often there are no deliberate policies to narrow the gap amongincome groups. If such policies do exist they are just as often either badlyadministered, or ill conceived or both and inefficiencies abound. If theresulting politically and socially explosive and from a humanitarian viewuntenable trend towards growing numbers of poor in many LDCs is to be changed,appropriate policies have to be worked out that are directly oriented towards

    1/ For a review of the literature see W.R. Cline, "Distribution and Develop-ment," Journal of Development Economics, 1, nr. 4, 1975 pp. 359-402.

    2/ J. Fei, G. Ranis and F. Stewart, "Basic Needs: A Framework for Analysis,"World Bank, 1979; Paul Streeten, "From Growth to Basic Needs," Financeand Development, September 1979.

    3/ Montek S. Ahluwalia, Nicholas G. Carter, Hollis B. Chenery: "Growth andPoverty in Developing Countries", in H. Chenery, Structural Change andDevelopment Policy, New York 1979, pp. 456-499.

  • - 2 -

    target groups and whose effects can be readily monitored. As each LDC facesmore or less unique growth and social problems, individual country studies areneeded to cope with the specific issues rather than highly aggregate crosscountry analysis whose operational impact is at best indirect. 1/

    3. This paper addresses the specific issues of social development inTunisia by analyzing past performance and specifying existing shortcomings, toprepare the ground for formulating an appropriate policy mix. The principalfocus is on (i) what the social objectives are, (ii) what has been done toimplement social targets, (iii) what were the consequences of policy inter-vention and (iv) how the results can be evaluated with regard to efficiencycriteria. The paper attempts to summarize important aspects of social devel-opment, with a view to strengthen the emerging dialogue and giving directionto future work orientation.

    4. Past work in the field is sketchy. Its main weaknesses are that noreliable intertemporal comparisons could be made and that it relied on datagathered during the '60s. 2/ Since then basic economic changes have takenplace. However, with the publication in 1978 of the second household consump-tion budget survey of 1975, a comparable and exhaustive data base becameavailable that allows to draw inferences on income distribution and povertyincidence and their changes over time. 3/

    5. The quality of the data generated by the two household surveys is

    good. The surveys were based on completely drawn and stratified randomsamples. Random samples cannot guarantee that responses of certain groupsare complete, especially those of the highest and lowest groups. However, theprofessionalism of the staff of the National Statistical Institute and thegood attitude of respondents to the interviewers contributed to a high degreeof confidence in the results. Also, the two surveys followed the same method-ology and there is no apparent reason to suspect differences in the qualityof data over time. The surveys explain 79 percent and 83 percent of privateconsumption in 1966 and 1975, respectively, as it is reflected in the nationalaccounts. Auto-consumption of foodstuffs is included, but imputed rent forowner-used housing is not. This may account for some of the discrepancy.Overall, however, these are excellent results as far as consumer surveys go,

    1/ Frances Stewart: "Country Experience in Providing for Basic Needs",Finance and Development, December 1979, pp 23-26.

    2/ The most comprehensive work in this field is provided by Russel A. Stoneand John Simmons "Change in Tunisia: Studies in the Social Sciences",Albany, N.Y. 1976.

    3/ The two available comprehensive household surveys are: Institut Nationalde la Statistique "La consommation et les depenses des menages en Tunisie1965-68", Tunis 1968, and "Enguete nationale sur le budget et la consomma-tion des menages 1975", Tunis 1978.

  • - 3 -

    which in most countries typically explain a much lower percentage of privateconsumption. On the whole, Tunisian data compare favorably with those inother LDCs. And in any case, the findings presented below are not sensitiveto small variations in the data. This paper is mostly concerned with changesin differences within groups rather than with precise level of the indicatorsused. The quality of the existing data base is sufficient tu make the infer-ences drawn in this paper.

    6. This paper then is a first attempt to bring the new information to-gether, analyze it and identify major issues. As such one has to be verycareful about the inferences drawn which may or may not stand up to closescrutiny in light of additional data that will become available following thenext household survey planned for 1980, which will reflect important socialpolicy changes introduced during the current plan period (1977-81). In viewof the remaining uncertainties, the paper primarily aims to identify issuesrather than to give final results. These issues will be taken up in theongoing dialogue with the authorities. This paper contributes to thisdialogue through (i) collection of data; (ii) definition of target groups;(iii) description of current programs and policies; (iv) setting of evaluationcriteria; (v) presenting tentative evaluation; and (vi) suggesting possiblepolicy choices. However, because of obvious limitations in the data base andthe remaining uncertainties, the statements that can be made in this paperwill become progressively weaker as one moves from (i) to (vi).

    II. GENERAL BACKGROUND

    A. Development Objectives

    7. Tunisia's main development objective is progressive improvement inthe living standard of its population through a self-sustained process of fasteconomic growth coupled with equitable distribution of the fruits of develop-ment among individuals, regions and generations. As early as in 1961, quan-titative social targets were established in the first ten year perspectivePlan. 1/ At that time the planners considered that it took D 50 per personper year to lead a "decent" life, although it was never specified how thisfigure was derived. In subsequent planning documents and studies this figurebecame regarded as the "poverty line" against which social development perform-ance was evaluated. The objective of the first perspective plan was to assurethat every Tunisian should receive a minimum income above this poverty levelby 1974. For 1971, the end of that perspective planning period, a targetincome distribution was to be reached which specified a per capita minimumincome level of D 45 (in constant 1961 prices). Although this target was notreached (see Annex I for a detailed discussion), it is striking that a devel-oping country like Tunisia has set detailed quantitative targets on poverty

    1/ Republique Tunisienne: "Perspectives decennales de developpement,1962-71"; pp. 73-77.

  • -4-

    alleviation and income distribution nearly two decades ago. This rathertypical concern about social balance and human needs continues to characterizeTunisian economic policies to this day as witnessed by subsequent plans andpolicy declarations. However, the quantitative implications of such policieswere never again as clearly spelt out as in the first perspective plan. Theobjective still remains to bring every Tunisian above the perceived minimumannual income level of D 50 in constant 1961 prices which by 1979 has inflatedto about D 100 in current prices.

    8. Rapid economic growth coupled with special attention to employmentcreation and demographic control measures formed the basis of the long-termstrategy announced by the first perspective plan to reach the basic socialobjective. The planners recognized that reaching the social objectives throughexclusive reliance on market processes, i.e. generation of primary incomesthrough employment creation alone, would take too long, given the resourcesconstraint and the slow trickle-down effect. After all, equitable income dis-tribution and satisfaction of basic needs are quite different things and notnecessarily mutually dependent. On the other hand, the planners recognizedthat measures designed to redistribute incomes may have counterproductiveeffects on much needed private entrepreneural initiative and growth. Attempt-ing to strike a balance between these needs, successive plans pushed for rapideconomic growth and employment creation to generate incomes, while suggestingappropriate fiscal measures, minimum wages legislation, social services andrural development to redistribute primary incomes to the lower income groups.To safeguard their real incomes, price support measures of basic consumergoods were also included in the social policy package. This overall strategyremains unchanged to this day as reiterated in the current Fifth DevelopmentPlan (1977-81). 1/

    9. Such a comprehensively planned approach to poverty and distribu-tional issues is, of course, commendable. The issue is whether Tunisia's"politique concertee" of growth with equity has indeed succeeded in alleviat-ing poverty and in improving income distribution. It would also be interest-ing to know whether Tunisia succeeded in striking the planned balance betweengrowth, social justice and stimulation of private initiative. Another issuerelates to the cost of achieving the desired ends and how efficiently theprograms have been implemented. These issues will be taken up below.

    1/ Republique Tunisienne, Ve Plan de developpement economigue et social,1977-1981; esp. pp. 11-38; 151-205; and 337-433.

  • - 5 -

    B. Marco-Economic Performance 1/

    10. Tunisia is poorly equiped with natural resources. The limitedresource base, dependence on a small domestic market, an unusually long seriesof poor crop years and fundamental political-economic changes limited annualreal GDP growth per capita during the 60's to 2.2 percent. At the end of 1969reforms were made changing policy orientations from collectivism and inward-looking investment strategy towards an export oriented economy. Investmentstrategy shifted from infrastructure orientation towards directly productiveindustrial development. Public sector predominance made room for increasedprivate initiative. The economy picked up momentum and average annual percapita GDP during 1970-78 increased by 6.1 percent in real terms. This putTunisia among the ten best performers among member countries of the World Bank.Besides responsible policies pursued by the authorities, exceptional factorslike good weather, rapid improvements in the terms of trade following thechange in world market prices for petroleum and phosphates in 1973/74, buoyantworkers' remittances and a sharp upswing in tourism strongly contributed tothis excellent growth performance. However, structural weaknesses of theTunisian economy continued to exist as reflected in wide swings of annualgrowth, depending on weather and other exogenous developments (e.g. exportdemand). For instance, annual real GDP growth was 11 percent in 1971, 0.4percent in 1973, 10 percent in 1975 and 4.3 percent in 1978. Also, the quiterespectable per capita GNP in 1978 of $998 in current prices, which putsTunisia among the middle income countries, hides important interpersonal andinter-regional distributional imbalances. This is caused partly by greatdifferences among various regions in the endowment with natural resources,productivity levels and grave unemployment problems.

    11. The Tunisian economy is charaterized by distinct differences inregional development and great variability of productivity among sectors.This is caused by great differences in soil, water and climatic conditions invarious parts of the country, a skewed distribution of population and invest-ment and historical factors. Most of Tunisia's population and economicactivity are concentrated along the coast, especially in a few big urbancenters. This engenders great income disparities among the different regions(para. 14) as well as rural-urban migration resulting in certain diseconomiesof rapid urban growth.

    12. One of the most striking characteristics of the Tunisian economy iswidespread unemployment and underemployment. 2/ This is probably the mostimportant social and political problem facing the Tunisian authorities today.In 1978, some 16 percent of the labor force, i.e. a quarter million people,were unemployed. Considering the underemployed, especially in agriculture and

    1/ For detailed analysis of the Tunisian economy see the World Bank's eco-nomic reports, e.g. Economic Position and Prospects of Tunisia, A Reviewof the Fifth Development Plan 1977-81, May 1977 and updating memoranda.

    2/ A closer look at the unemployment problem is provided in Annex III.

  • services, the effective overall unemployment rate is closer to 30 percent.Also, there are enormous differences in rural and urban unemployment andgenerally among regions. However, employment data are not very reliable andthen there is the conceptual problem of the meaning of unemployment in adeveloping country like Tunisia where unemployment benefits and functioninglabor exchanges do not exist. Hence, to be unemployed or underemployed meansthat some resources are available to those affected on which they can fallback upon, be it own savings or support from relatives. In any case, it wouldbe wrong to assume that all those that are not permanently employed belong tothe poorest strata of the economy.

    13. About half of the unemployed are young people seeking their firstjob. The pressure on the job market is increasing despite the sustainedefforts in emloyment creation by the authorities. There are four majorreasons for this. First, those born during the high birthrate years duringthe 1960's are now coming of working age. Second, retention rates of schoolsare declining. Third, great numbers of women are joining the labor force.Fourth, emigration of workers to Europe has all but subsided. Althoughrenewed emigration to Libya in 1977-78 was temporarily able to compensate forthe loss of the European market, this might not be possible in future. In1979, for example, there was no net emigration. Hence, the labor force isofficially projected to grow at an average annual rate of 3.2 percent duringthe next 10 years, compared to population growth of only 2.3 percent. In viewof the large backlog of unemployment and the existing resources and marketconstraints to rapidly increasing employment, Tunisia faces a structuralunemployment problem for the years to come. This has, of course, far reachingimplications for reaching the social objective. 1/

    14. A three-pronged strategy has been chosen to deal with the unemploy-ment problems. (i) The government is making efforts in stemming populationgrowth through actively promoting family planning. However, much time isneeded before the family planning efforts of today result in slowing growthof the labor force. This is at best a long term possibility. (ii) Anotherapproach taken by the authorities consists in increasing investment and produc-tion to absorb more labor. In view of the relatively small resource base, thelimited absorbtive capacity for a large number of comparatively small labor-intensive investment projects, marketing difficulties and sometimes counter-productive incentives for employment creation, this is at best a medium termsolution. Meanwhile, the government seeks to disseminate some temporary relief,essentially through creating a National Work Corps 2/ and absorbing people in

    1/ Labor absorption and productivity issues might well be at the base ofunderstanding the links between growth, distribution and basic needs inTunisia. However, this paper will not cover exhaustively the issuesinvolved awaiting the outcome of a special study currently being preparedunder UNDP/ILO auspices in Tunisia.

    2/ All unemployed male Tunisians between 18 and 30 years old may be draftedinto the Corps for 12 months. They receive rudimentary on the job train-ing and are sent to work in priority sectors and geographic zones thatsuffer from labor shortages. They continue essentially the works under-taken by the "chantiers de lutte centre le sons developpement", a type offood for work program.

  • -7-

    public enterprises and public administration. In order to improve the posi-tion of low income earners the government maintains large subsidization andtransfer programs and the supply of public services free of charge. Thequestion is whether these measures indeed reach the target groups, whether thegiven effects are obtained with least cost and without imparing productivityand incentives of the recipients.

    C. Social Achievements and Failures

    15. Since Independence Tunisia has pursued an economic growth policythat was characterized by deliberate distributional and social measures. Forexample, during 1970-78 Tunisia spent about 16 percent of GDP on varioussocial programs. This is a very high level by international standards andmight be compared to other countries' high expenditure for defense, which inTunisia is very low (1 percent of GDP). Both have high opportunity costs, butsocial expenditures are surely less wasteful. As a result, apart from itsexcellent overall growth record (para. 10), Tunisia has made impressive socialgains. By 1976, primary school enrollment had reached 66 percent and secondaryenrollment 18 percent of the relevant age groups. Education is free throughuniversity including books, room and board, and clothing for needy students.Cash stipends at university level are the rule. A family planning program hasbeen introduced, which since 1973 has met with substantial success. The birthrate declined from 43.8 per thousand in 1966 to 35.4 per thousand in 1978.The health care program was expanded rapidly and services are essentially pro-vided free of charge to the needy. Infant mortality declined from 125.0 perthousand in 1966 to 85.0 per thousand in 1978. Attempts have been made bythe various plans through a number of measures to grant every Tunisian a"decent" living. The share of the population below the Tunisian poverty cut-off line of D 50 in constant 1961 prices declined from 73 percent in 1961 to42 percent in 1971 and to 37 percent in 1975. The share of the populationwith less than nutritionally acceptable calorie intake declined from 30percent in 1966 to about 14 percent by 1975.

    16. From the relatively high per capita GDP level, its very substantialgrowth, and the social achievements mentioned above, a rather glowing pictureon social performance appears to be emerging. In fact, recent calculationsby World Bank staff indicate that, based on cross section data, incomecorrected social indicators on life expectancy, fertility and literacy in1970 were better than expected, but were worse for infant mortality. 1/ Inthe meantime, substantial improvements in reducing infant mortality have beenmade. This pushes social achievements in all four fields beyond what would beexpected at Tunisia's current income level.

    1/ Paul Isenman: Income Corrected Social Indicators, August 9, 1978. Also,the social indicators included in Tables 17-24 of the World Bank's "WorldDevelopment Report 1979", Washington 1979, pp 158-173, show that Tunisia'ssocial achievements generally compare well to those of some other middleincome countries. Figures for selected countries of the same generalarea are included in Statistical Appendix Table 17.

  • -8-

    17. But the question arises whether these general indicators indeed aresufficient to judge performance, as they may hide a much more complex distri-butional reality. The distributional incidence of the various programsmentioned above is not at all obvious. This is underlined by the data con-tained in the 1966 and 1975 consumption surveys. While these surveys indicatethat in the inter-censoral period real per capita spending increased rapidly,they also indicate large and growing regional and interpersonal disparities.Thus, in 1975 the average per capita expenditure in the poorest region (theNorth West) amounted to only 38 percent of that of the richest region (GreaterTunis) and to 67 percent of the national average (Table 1). There is asimilarly pronounced disparity between the urban and rural population. From1966 to 1975 average real spending per person in big cities increased by 53percent and in other urban areas by 65 percent, whereas that in rural areasincreased by 41 percent. In addition, the 56 percent increase in nationwideannual spending glosses over the worsening distribution of that expenditure byregion also shown in Table 1.

    Table 1: AVERAGE ANNUAL EXPENDITURE PER PERSON BY LOCATION(in dinars; constant 1975 prices)

    1966 1975 1966 1975

    Greater Tunis 153 260 Big Cities 152 233North East 89 136 Other Urban Areas 92 152North West 80 98 Rural Areas 75 106Center 92 143South 88 117

    National Average 94 147 National Average 94 147

    Source: INS, Consumption Surveys, 1966 and 1975

    18. Besides the inter-regional differentials, the interpersonal distri-bution is also rather skewed. Thus, 68 percent of the population in 1975spent less than the national average. The richest 20 percent of the popula-tion in 1975 spent 50 percent of total consumption expenditure and the poorest20 percent a mere 5 percent. While rapid economic growth in Tunisia during1967-75 on average was reflected in rapidly increasing standards of living,distributional inequalities worsened, with the top 5 percent of the populationaccumulating shares at the expense of everybody else. The Gini coefficientincreased from 0.40 in 1966 to 0.44 in 1975, with the inequality remainingstrongest in the big cities, but rising most rapidly in rural areas. Whilethis increase is statistically significant, it is not all that important.In any case, it would be premature to draw conclusions from this informationon social performance in general without careful analysis of what actuallyhappened to those that are at the bottom of the distributional scale. Theycertainly were losing shares and thus were relatively worse off, but this isnot the issue. The issue really is whether the well being of the poor hasimproved or not. To evaluate properly the effects of growth and social pro-grams, more information about income distribution and existing social programs

  • -9-

    is needed. Above all, a benchmark needs to be established against whichpoverty incidence and poverty alleviation can be measured (see Chapter III).Distributional aspects proper and some of the likely causes for changes inincome distribution are briefly discussed in Annex I.

    III. THE POVERTY CONCEPT

    19. What actually happened to the lowest quintiles' absolute positioncannot readily be gleaned from the distributional data. It is, of course, atruism that the poor are poorer than the rich. And that the rich are gettingricher is a basic fact of the accumulation process on which economic growth isbased. The issue is whether the poor too are getting "richer" or - to cast itin less emotional terms - whether the poor are better off today than they wereyesterday in spite of their worsening relative position.

    20. In order to resolve this issue a benchmark is needed against whichto measure and which can be used to test hypotheses. This benchmark is theabsolute poverty limit. The literature engages in lengthy discussions on theappropriate definition of absolute poverty. 1/ There is no reason to enterthat disussion here. Instead, an absolute poverty measure based on the localcost of minimum nutritional and non-food requirements (clothing, shelter) isadopted for the purposes of this paper. Hence, the absolute poor are thosepeople below a certain income level whose basic human needs cannot be readilyfulfilled. As such they cannot lead fully productive lifes and by definitionwould eventually die of starvation and/or exposure depending on how far belowthe poverty threshold they are. Because of differences in rural and urbandietary and other consumption habits, auto-consumption and consumer prices,different absolute poverty levels are defined according to the environment thepoor are in. The relevant Tunisian urban and rural poverty thresholds for1966 and 1975 are derived in Table 2 using a methodology generally followed bythe World Bank.

    1/ For example see T.N. Srinivasan, "Poverty: Some Measurement Problems",Proceedings of the 41st Session of the International Statistical Insti-tute, New Dehli, December 1977, World Bank Reprint Series, Nr 77.

  • - 10 -

    Table 2: ABSOLUTE POVERTY THRESHOLD(In Dinars)

    1966 1975Urban Rural Urban Rural

    1. Food expenditure ofthe 20th percentile(per person per month) 2.385 1.629 3.416 2.333

    2. Adjusted expenditureto meet nutritionalrequirements 2.273 1.302 3.414 1.956

    3. Cost of nutritionallyadequate cereal-baseddiet (per person/month) 1.298 0.878 1.950 1.320

    4. Non-Food expendituresof 20th percentile(per person/month) 2.352 1.063 3.166 1.416

    5. Adjusted non-foodexpenditure 2.350 0.892 3.163 1.187

    6. Absolute Poverty Line(2) + (5)

    (i) per person/month 4.623 2.194 6.577 3.143(ii) per person/year 55.476 26.328 78.924 37.716

    Source: Bank staff calculations based on data reported in the INS ConsumptionSurveys 1966 and 1975.

    21. This methodology calculates the nutritional element of the povertyline based on a typical food basket of the low income groups, e.g. that ofthe 20th percentile of the household expenditure distribution. The caloriecontent and the cost of this diet are then determined and adjusted to FAOstandards. Although, theoretically, a diet based on cereals, with somesupplementation to provide adequate balance may be nutritionally adequate,this may not be acceptable to the population. Table 2 reproduces such adiet only as an indicator of different food standards according to environment.Unfortunately, no ready reference point equivalent to the nutritional require-ment is available to calculate non-food needs. Thus, the non-food expenditureof the lowest quintile is again taken as a bench mark. Adjustments are madeaccording to the views expressed by officials in the Statistical Office asto their conception of absolute necessity of such spending. While thismethodology of calculating absolute poverty thresholds is not a really satis-factory approach, it is a simple one and at least catches variations in pricesand conceived needs in different environments. It is safe to conclude thatthe probability is high that those people with expenditures below the proposedpoverty threshold are indeed absolute poor.

  • - 11 -

    22. Comparing the poverty thresholds derived in Table 2 to the distri-butional figures provided by the two consumption surveys, it appears that theincidence of absolute povercy in the population declined from 27 percent in1966 to 20 percent in 1975 in urban areas, and from 20 percent to 15 percentin rural areas, respectively. For Tunisia as a whole the percentage of thepopulation below the poverty threshold declined from 23 percent in 1966 to 17percent in 1975. The number of absolute poor declined by 6 percent overall,from 1,029,000 to 971,000, respectively, during a period of comparativelyrapid population growth of 2.3 percent. Hence it may be concluded that theperiod of rapid economic growth in the intercensoral years indeed benefittedthe poor as well. It is interesting to note, however, that the number ofabsolute poor living in urban areas increased, while the number of rural poormarkedly declined in the wake of internal and external migration.

    Table 3: POPULATION IN ABSOLUTE POVERTY

    1966 1975Percentage Percentage

    Absolute poor in absolute Absolute poor in absolute(in '000) poverty (in '000) poverty

    Urban 473 27.0 532 20.0Rural 556 20.0 439 15.0

    TOTAL 1,029 22.7 971 17.4

    Source: Staff estimates based on 1966 and 1975 household consumption survey.

    23. As mentioned above (paragraph 7), the Tunisian authorities alsowork with a normative income level below which a "decent" life is considerednot possible. According to their adopted definition, the population livingbelow that desirable level declined from 73 percent in 1961 to 54 percent in1965 and 37 percent in 1975. Hence, while the target of the first prospectiveten year plan of providing every Tunisian with a "decent" life by 1974 has notbeen reached since the cut-off chosen might simply have been too high, therewas clearly a strong move in the right direction in that the standard ofliving of many of the poor improved with rapid economic growth.

    24. Another indicator pointing in the same direction is that the groupof the population unable to cover their minimum daily caloric intake require-ment declined from 30 percent in 1966 to 14 percent in 1975. This latterpercentage corresponds closely with the nationwide absolute poverty estimatefor 1975 advanced in this paper and may thus be used as a check for therobustness of its adopted absolute poverty measure. It is interesting to notethat the percentage of the population not being able to cover their dailycaloric needs has been halved during the intercensoral period. The availablefigures on caloric intake in 1975 also show that 18 percent of the populationin the big cities are unable to meet their caloric requirements compared toonly 9 percent in rural areas. This once again underlines our earlier findingthat the majority of the absolute poor in 1975 can be found in the cities.

  • - 12 -

    25. Once it has been established that in 1975 about 17 percent of thepopulation or approximately one million people lived in absolute poverty, anumber of further details are needed before measures can be devised to helpthese people to improve their lot, which after all is the ultimate goal ofsuch analysis. For instance, one would have to know who the poor are, wherethey are and why they are poor. This would give a poverty profile againstwhich past and future social measures could be evaluated. Such a povertyprofile is developed next.

    26. According to the 1975 consumption survey most of the poor are likelyto be unskilled seasonal workers in agriculture and industry, small farmers,artisans and petty traders, all of whom have average annual spending levelssubstantially below the national average (Table 4). It is striking to note,that in 1975, Government employees who account for 10.2 percent of the popula-tion, accounted for 20 percent of total expenditures, whereas farmers and farmworkers who make up 40 percent of the population, spent 29 percent of totalexpenditures. Unskilled workers are also disproportionately represented:while they make up 2.9 percent of the population they spent but 1.2 percent ofthe total. Table 4 also shows that the poor can be found in all occupations,self-employed and workers alike. They are people in the mainstream of eco-nomic activity, but with low productivity and/or large families. It is veryinteresting to note that the differences between groups as measured by theGini coefficients are much greater than those within groups, which highlightsthe importance of skill differentials and the relative social immobilitybetween groups.

    Table 4: AVERAGE CONSUMPTION EXPENDITURE ACCORDING TOOCCUPATION OF HEAD OF HOUSEHOLD, 1975

    Avg. Exp. per Percentage of Householdsperson p.a. Ginni in Expenditure Group(in dinars) Coefficient

  • - 13 -

    27. The next issue is where the poor are. It has been shown above thatthe incidence of poverty is relatively higher in urban areas (20 percent) thanin rural areas (15 percent). In absolute terms 532,000 or 55 percent of theabsolute poor in 1975 lived in urban and 439,000 or 45 percent in rural areas.These figures can be further disaggregated according to the five major economiczones that are defined in the Fifth Development Plan (1977-81) and that alsoare instrumental in defining the Government's regionalization incentives asoutlined in the 1974 Investment Code. These five regions are: Greater Tunis(the governorate of Tunis), the North East (governorates of Zaghouan, Nabeul,Bizerte), the North West (governorates of Beja, Jendouba, Le Kef, Siliana),the Center (governorates of Sousse, Monastir, Mahdia, Kairouan, Kasserine) andthe South (governorates of Sfax, Gafsa, Medenine, Gabes, Sidi Bouzid).

    28. The great differences between average spending levels in theseregions has been shown above (paragraph 16). Unfortunately, the consumptionsurvey does not publish data on the regional distribution of the rural andurban population and its expenditure. However, from the data on per capitaspending levels by region one can extrapolate a regional approximation ofabsolute poverty incidence (Table 5). Accordingly, the incidence of absolutepoverty appears to be highest in the North West where approximately one thirdof this target group is concentrated, and in the South. It should be notedthat the Gini coefficient for the North West shows the most unequal distribu-tion (except for Tunis) and the least unequal distribution for the South.Hence, while the people in the South are "uniformly poor", poverty in theNorth West is contrasted with substantial affluence. This is also and espe-cially true for Tunis and very likely for the other big cities (Bizerte, Sfax,Sousse) as well.

    Table 5: DISTRIBUTION OF THE ABSOLUTE POOR BY REGION, 1975 /1

    Percentage Distribution Absolute Numbers Regionalof the Absolute Poor (in 1,000) Ginni

    Rural Urban Total Rural Urban Total Coefficient

    Greater Tunis - 18 10 - 94 94 0.43North East 12 20 16 53 106 159 0.38North West 40 23 31 175 125 300 0.41Center 21 19 20 92 101 193 0.40South 27 20 23 119 106 225 0.36

    TOTAL 100 100 100 439 532 971 0.44

    /1 These figures were derived by taking the distribution of spending belowurban and rural poverty levels by region from the 1975 survey and theresulting percentage distribution was applied to the total nationalfigures of rural and urban poor.

    Source: INS, Enquete nationale sur le budget et la consommation des menages,1975, Tunis 1978.

  • - 14 -

    29. The heavy incidence of absolute poor in the predominantly agricul-tural North West is a phenomenon which cannot be readily explained by anextremely poor agricultural resource base as it can be done for instance inmost parts of the South and the Center. The key lies in the unequal distribu-tion of the available resources. The North West has some of the best acreage(Medjerda Valley) and climatic conditions in Tunisia. Howevet, land holdingsare highly unequally distributed with the good bottom land being held by a fewand the poor mountain land being shared by many. Apart from agriculture thereis little else offering gainful employment. The structural unemployment prob-lem in this region is accentuated by the predominant large farming operationsusing farm machines substituting for labor. In view of the very limitedpermanent employment generation and the relatively poor equipment with socialinfrastructure, the North West has traditionally been an area with great out-migration.

    30. What is being done to improve the lot of the poor? What are theresources allocated? What is the potential for increased production andincomes for the poor in their present locations? What effect do governmentpolicies have on income distribution? Do the results correspond to the aspira-tions of the policy makers? These are the questions taken up in the followingchapter.

    IV. SOCIAL PROGRAMS

    31. This chapter tries to come to grips with the variety of effortsundertaken by the authorities (i) to alleviate poverty and (ii) to improveincome distribution. In pursuing these goals the government uses a three-

    pronged approach. Firstly, there are social services geared towards the basicneeds of the population in the following form:

    (i) free education;(ii) free health services and family planning;(iii) public housing, water supply and sewerage.

    Secondly, the authorities pursue income policies through:

    (i) minimum wage legislation;(ii) price support programs for basic consumer goods;(iii) the social security fund;(iv) tax policies.

    Finally, there are separate regional and rural development programs like:

    (i) the rural development fund;(ii) the regional development poles;(iii) FOPRODI and the 1974 investment code as amended in 1978;(iv) land agencies and offices.

  • - 15 -

    The issues are what each set of programs does to reach the target groups, whatresources are devoted to the various undertakings, and who benefits most fromsuch programs. The discussion is essentially limited to what happened duringthe intercensorial period (1967-1975), since only for this period relativelycomplete data are available. This is somewhat unfortunate, because importantstructural changes have taken place since 1975. Whenever possSble, data andpolicies up to 1978 were included in the discussion to convey the nature ofthe changes. However, until the results of the forthcoming third consumptionsurvey to be undertaken in 1980 become available, a general framework ismissing in which recent social developments can be incorporated and comparedto earlier ones.

    A. Social Services

    32. Social services play an important part in the effective income ofthe Tunisian population. Table 6 summarizes the trends of social expenditureincluded in the central government's budget. These figures are impressive,reflecting the concern of the authorities for covering the basic needs of thepopulation in education, health and housing. Three important observations canbe made. First, per capita expenditure in real terms on social servicesincreased by 5.3 percent p.a. during 1967-1975, compared to a per capita GDPgrowth of 5.1 percent during the same period. Second, the share of socialservices in private consumption increased very slowly. Third, the share ofsocial expenditure in total budgetary outlays declined. These three observa-tions could be interpreted to imply that the authorities consider the relativelevel of social expenditure in total budgetary expenditure as being satis-factory as an overall saturation level has been reached. With an equivalentof social service consumption of 14 percent of private consumption this is notsurprising, especially since it excludes the expenditures of the price equal-ization fund and of the social security system. The issue is, however, howthese expenditures are distributed by recipient income groups.

    Table 6: TRENDS IN GOVERNMENT SOCIAL SERVICES EXPENDITURE /1

    1962 1966 1975 1978

    Social Expenditures per capita (in dinar)- in current prices 6.7 11.3 25.6 37.3- in constant 1972 prices 9.5 13.5 21.4 26.1

    Social Expenditures in % of GDP 7.7 9.3 11.0 9.1Social Expenditures in % of budget 22.3 28.5 24.7 25.1Social Expenditure in % of private consumption 12.6 13.7 13.8 14.4

    /1 Including general Government's capital expenditure on education, healthand housing and current expenditure by the Ministries of Education,Health and Social Affairs. These figures exclude the expenditures ofthe price equalization fund and of the social security system.

    Source: Ministry of Planning

  • - 16 -

    33. If, for example, the poor had received in 1975 the average D 25.6of social expenditure per capita, this would have amounted to 68 percent ofthe rural poverty income threshold and to 32 percent of the urban one. Ifthe system of social expenditure were such as to favor the poor, then thesepercentages would have been been higher. However, a priori it cannot beexcluded that the poor did not participate in the benefits emanating fromsocial expenditure to the desired extent (i.e. overproportionately or at leastproportionately). If this was the case, social expenditure by the Governmentwould have contributed to the worsening trend in income distribution observedin Annex I. It is therefore very important to know the incidence of socialexpenditure, with due regard to the financing side of the budget as well (taxincidence).

    34. Unfortunately the time and manpower available for the completionof the present analysis did not allow for detailed work on overall budgetaryexpenditure incidence. The important question, who benefits most from socialexpenditure and why, thus can be treated only in a preliminary and impre-sionistic way. This is all the more regrettable as methodologies have recentlybecome available which are readily transferable to Tunisian circumstances. 1/It is hoped that future work could include appropriate surveys needed to cometo grips with the expenditure incidence issue in its entirety. For the timebeing, however, the present work had to content itself with already existing,albeit incomplete, data. The inferences drawn from them are to be taken witha grain of salt. If anything, they show that much more detailed informationis needed before specific measures can be devised to increase the efficiencyof the social system by targetting it more directly to the poverty group.

    1. Education

    35. The authorities have always paid special attention to education asa basic ingredient for economic development and a means for ensuring socialprogress. This is reflected in the Government budget where education expendi-tures rank as the most important item among budgetary expenditures on socialservices, accounting for some 60 percent of total (Table 7). Average realexpenditure per student increased by 38 percent during 1967-75. The authori-ties have built up a comprehensive education system that provides ready accessto all students although education is not compulsory. Education is freethrough university. The government pursues a policy of "Democratization ofEducation" through cash stipends at all levels of education. At the primarylevel needy students receive free learning materials and clothing. At thesecondary level nearly 80 percent of students receive cash stipends. Nearlyall university students have scholarships of currently about D 36 per month,an amount approaching the current monthly minimum wage of an industriallaborer. Since 1966 enrollment ratios have been growing at higher educationlevels, although primary enrollment ratios are below the 1966 level. The

    1/ Marcelo Selowsky, Who Benefits from Public Expenditure? A Case Study ofColumbia, New York, 1979; Jacob Meerman, Public Expenditure in Malaysia,Who Benefits and Why?, New York, 1979.

  • - 17 -

    quality of primary education is improving with better endowment with teachersand schools. However, adult literacy has not risen significantly (Table 7),with large regional differences remaining. In the poorest, i.e. rural,governorates (Kairouan, Beja, Jendouba, Kasserine, Sidi Bouzid, Mahdia) adultliteracy is below 25 percent, whereas it is only 65 percent in Tunis.

    Table 7: EDUCATION INDICATORS

    1962 1966 1975 1978

    Budgetary Expenditure on education inpercent of total social expenditure 58.0 63.0 60.0 62.0

    Expenditure on education per student(in const. 1972 dinars) 46.3 44.5 61.2 61.5

    Enrollment ratiosprimary: /1 total 67.0 72.0 66.0 68.0

    female 32.0 36.0 39.0 41.0

    Secondary: total 12.0 15.0 18.0 18.5female 6.0 8.0 12.0 13.0

    University: 0.2 1.3 3.5 4.0

    Pupil-teacher ratioprimary 61.0 55.0 39.7 39.2secondary 16.0 22.0 21.0 22.5

    Inhabitants per student 826 458 184 150

    Adult literacy rate /2 31.0 32.1 45.1 46.0

    /1 Excluding repeaters./2 Of population 10 years and older.

    Source: Ministry of Planning

    36. Overall, the figures of Table 7 seem to indicate successful quantita-tive performance 1/ especially if viewed together with other countries in thegeneral area. The question is, however, whether the education system is

    1/ For details see the forthcoming Education Sector Memorandum which concludesthat the formal education system in Tunisia is well developed quantita-tively, although there still exist serious qualitative problems like lowinternal efficiency and the general lack of relevance for employment.The paper also describes the measures currently being implemented toovercome some of the systems' weaknesses.

  • 18

    equitable and what its impact on employment and lifetime earnings is. Theseare two critical questions which have not yet been treated exhaustively in theTunisian context. One might argue that poorer households with more childrenwould be the prime beneficiaries of the free education system. However,there is no country in the world where this is borne out by the facts andTunisia is no exception. There are out-of-pocket costs involved in educatinga student in the different categories and this cost is not equal, althougheducation as such is free. Household income foregone in sending children toschool carries different weights in different budgets. Also, the cost togovernment is by no means equal. It is known that seeing a student throughsecondary education in 1975 was about 8 times more expensive than seeing himthrough primary education and university education was 41 times more expensive.

    37. The question then arises whether the poor participate equally in allcategories of education. As might be expected based on international experi-ence this is not the case (para. 38). The opportunity cost of sending childrento school is high, as indicated by low enrollment ratios, high dropout ratesand a high degree of seasonal absence of primary school children in ruralareas (para. 39) as the school year does not correspond to the agriculturalyear. Also, traditional attitudes are more pronounced among lower socio-economic groups thus preventing a larger amount of females from going toschool. In addition, access to schools in the poorer regions of the countryis physically difficult. All this would indicate that the poorer strata ofthe population, especially in rural areas, probably participate proportionallyless in the educational services provided by the government than the rest ofthe population. This is especially noticeable for the female school population.

    38. Equity in the education sector means a proper distributionbetween the different social groups of benefits and costs over the long run.Detailed information on intergeneration social mobility and student numbersby socio-economic origin are not readily available. However, for 1970 theMinistry of Education made data available on the social situation of studentsin the three levels of education (Table 8). Although no such survey data areavailable for subsequent years existing figures show that the proportion ofstudents getting some higher education with fathers of higher socio-economicstatus is about 18 times higher than their share in the primary school popula-tion. Students originating in the lowest income group hold only about onethird of the places at the university level that would be expected if educa-tion was distributed equitably. These figures imply that about 30 percentof total educational expenditures of the budget in 1970 accrued to the twohighest socio-economic groups. However, their share in the total school agepopulation was probably considerably lower, as indicated by their share in thetotal population of about 13 percent. The generous system of cash stipends,especially at the university level, thus introduces a substantial bias infavor of the more well-to-do groups. The issue also arises whether favoringa large academic population is in the economic interest of the country.

  • - 19 -

    Table 8: SOCIO-ECONOMIC STATUS OF STUDENTS' FATHERS (1970)(in percent)

    Primary Secondary Higher RatioSocio-Economic StatusaL (1) (2) (3) (3)/(1)

    Highest