Regional Economist - July 2013

7/27/2019 Regional Economist - July 2013

1/24

HOUSE

FOR SALE

9.33% 256.36 2.78 56.39 5.23 6.3

2.78 40.39 7.23 2.35% 158.36 2.52

A Quarterly Reviewo Business andEconomic Conditions

V. 21, N. 3

July 2013

The Federal reserve Bank oF sT. louis

CeNtral to ameriCas eCoNomy

Mortgage DeaultsWhy Are ForeclosureRates So Much Lowerin Europe?

Metro ProfleHealth Care, Food SeDrive Economyo Louisville, Ky.

Financial Markets

An Engine orEconomic Growth


2/24

3 pre s i d e n t s m e s s a g e

10 The Racial Earnings Gap:

Changes since 1960

By Maria Canon

and Elise Marifan

Te earnings gap between

black and white men narrowed

between 1960 and 2000 or those

born in the South but widened

or those born in the North.

Since then, the gap has worsened

in both regions o the country.

12 Mortgage Deaults:

U.S. vs. Europe

By Juan Carlos Hatchondo,

Leonardo Martinez

and Juan M. SnchezDuring the last global recession,

house prices ell in some European

countries almost as much as

in some U.S. states. However,

mortgage deaults occurred at a

much lower rate in Europe. Te

authors say the dierence might be

explained by two regulations that

apply in Europe but are used on

a limited or much less restrictive

basis in the U.S.

14 Mortgage Applicants

Turn to Credit Unions

By Yang Liuand Rajdeep Sengupta

An examination o mortgage

data or the Eighth Distr ict shows

that ar ewer applications or

mortgages were made postcrisis

(annual average o 2010-2011)

than in 2004, beore the crisis

began. Tose who did seek

mortgages aer the crisis turned

increasingly to credit unions, as

opposed to banks and thris.

16 e co n o m y a t a g la n c e

17 n a t i o n a l o ve rv i e w

Mixed Signals,

but Moving Forward

By Kevin L. Kliesen

Te housing market is improving,

ostering growth in related seg-

ments o the economy. However,

continued high unemployment

and lackluster growth in real earn-

ings are causes or concern. Still,

GDP is growing, albeit slowly.

18 met ro pro ile

Louisville Transitions

to Service Economy

By Charles S. Gascon

and Sean P. Grover

Louisville is the ocus in the rst

installment o this new data-

driven eature in Te Regiona

Economist. Unlike many oth

older cities, Louisville has

smoothly transitioned rom t

industrial economy to the ser

economy, thanks in no small

part to its strong health-care

ood-service industries.

21 d i s t r i ct o ve rv i e w

Mapping the Big Servic

Manuacturing Industr

By Rubn Hernndez-Murillo

and Elise Marifan

Urban areas still host most ma

acturing jobs, despite the act

that most manuacturing jobs

over the past decades were in t

areas. At the same time, urban

areas have become increasingmore service-oriented, as serv

industries thrive near large co

centrations o people.

23 re a d e r e xch a n g e

c o n T e n T s

HOUSE

FOR SALE

9.33% 256.36 2.78 56.39 5.23 6.35% 248.

. .36 2.78 40.39 7.23 2.35% 158.36 2.52

AQuarterlyReviewofBusinessandEconomicCondi tions

Vol.21,No.3

July2013

The Federal reserve Bank oF sT.louis

CeNtral to ameriCas eCoNomy

MortgageDeaultsWhyAreForeclosureRatesSoMuch LowerinEurope?

MetroProfleHealthCare,Food ServiceDriveEconomyoLouisville,Ky.

Financial MarketsAn Engine for

Economic Growth

Financial Markets:An Engine or Economic GrowthBy Yongseok Shin

Do developed nancial markets lead to economic growth orresult rom it? While some economists argue or the latter,the author maintains that nancial marketsdespite theirshortcomings o lateare an essential ingredient or an

economy to grow in the long run.

4

The Regional

EconomistJULY 2013 | vol. 21, no. 3

The Regional Economist ub

qu b r pub a

r Bk

s. lu. i , -

u

, u

e r d. v

s. lu r s.

dt r

c J. w

s Py a

cu c. cu

dpty dt r

d c. wk

dt Pb a

K B

et

subu B

M et

a sbk

at dt

J w

T et F r dtt u

ak, mu, u

i i, Kuk

t, m. t

e d l r k,

lu, m s. lu.

P t y mmtt sby Bypyy

t 314-444-7425 by -m t

[email protected]..

Y wt t m t t

bw. sbm

tt t t t t t

t pt t t wb t /

pb t The Regional Economist

t wt tt tw.

W t t t t tt

ty t.

s-py bpt

bt b y t t wt

u.s. . T bb, t

www.t./pbt.

Y wt t The Regional

Economist, Pb a of,

F r Bk st. l,

P.o. Bx 442, st. l, Mo 63166-0442.

online EXTRA

Exploring the Linkbetween Drug Useand Job Statusin the U.S.

By Alejandro Badel

and Brian Greaney

Does more unemployment

increase drug abuse? Does

drug abuse prolong jobless-

ness? See what data rom th

National Survey on Drug U

and Health say about these

questions in this online-onl

article. Read it at www.

stlouised.org/publications/

2 th rn ecn |July 2013


3/24

James Bullard, p ceo

r Bk s. lu

wo dierent price indexes are popularor measuring ination: the consumerprice index (CPI) rom the Bureau o Labor

Statistics and the personal consumption

expenditures price index (PCE) rom the

Bureau o Economic Analysis. Each o these

is constructed or dierent groups o goods

and services, most notably a headline (or

overall) measure and a core (which excludesood and energy prices) measure. Which

one gives us the actual rate o ination that

consumers ace?

On the headline vs. core issue, I preer to

ocus on headline ination, measured as the

percentage change in the price index rom a

year ago to smooth out the uctuations in the

data. As I have discussed previously, head-

line measures attempt to reect the prices

that households pay or a wide variety o

goods, not a subset o those goods.1 Headline

ination is, thereore, designed to be the bestmeasure o ination that we have.

Between the two headline indexes, the CPI

tends to show more ination than the PCE.

From January 1995 to May 2013, the average

rate o ination was 2.4 percent when mea-

sured by headline CPI and 2.0 percent when

measured by headline PCE. Hence, aer

setting both indexes equal to 100 in 1995, the

CPI vs. PCE Infation:Choosing a Standard Measure

P r e s i d e n T s M e s s a g e

Comparing Price Indexes: CPI vs. PCE

SOURCES: Data obtained rom FRED (Federal Resere Economic Data) and authors calculations.

1995 1997 1999 2001 2003 2005 2007 2009 2011 2013

160

150

140

130

120

110

100

90

Headl ine CPI Headl ine PCE 2% annual ination

INDEX

(JANUARY

1995=

100)

CPI was more than 7 percent higher than the

PCE in May 2013.2 (See the chart.)

An accurate measure o ination is impor-

tant or both the U.S. ederal government and

the Federal Reserves Federal Open Market

Committee (FOMC), but they ocus on di-

erent measures. For example, the ederal

government uses the CPI to make ination

adjustments to certain kinds o benets, suchas Social Security.3 In contrast, the FOMC

ocuses on PCE ination in its quarterly eco-

nomic projections and also states its longer-

run ination goal in terms o headline PCE.

Te FOMC ocused on CPI ination prior to

2000 but, aer extensive analysis, changed

to PCE ination or three main reasons: Te

expenditure weights in the PCE can change

as people substitute away rom some goods

and services toward others, the PCE includes

more comprehensive coverage o goods and

services, and historical PCE data can berevised (more than or seasonal actors only).4

Given that the two indexes show dierent

ination trends in the longer run, having a

single preerred measure that is used by both

the ederal government and the FOMC might

be appropriate. What would it mean i it

were determined that headline PCE ination

is the better measure o prices consumers ace E N D N O T E S

1 See Measuring Ination: Te Core Is Rotten,

Federal Reserve Bank o St. Louis Review, July/

August 2011, Vol. 93, No. 4, pp. 223-33. See http://

research.stlouised.org/publications/review/11/07/

bullard.pd.2 In 2002, the Bureau o Labor Statistics began

releasing a chain-weighted version o the CPI,

which behaves similarly to the PCE over long

periods o time.3 CPI-W, the index or urban wage earners and

clerical workers, is used to adjust these benets or

ination, whereas CPI-U (headline) is shown in

the chart. Te two show similar trends rom 1995

to the present.4 For more discussion, see theMonetary Policy

Report to the Congress, on Feb. 17, 2000, at

www.ederalreserve.gov/boarddocs/hh/2000/

February/FullReport.pd.

(indicating, thus, that the CPI overstates the

true ination rate)? Continuing to use the

CPI would imply over-adjusting or ination

and, in eect, giving real increases in benets

over time. In this scenario, benets should be

adjusted or ination using the PCE instead.

Conversely, i it were determined that head-

line CPI ination is the better measure (and,

thereore, that the PCE understates the trueination rate), then the FOMC should target

CPI ination rather than PCE ination.

Te FOMC careully considered both

indexes when evaluating which metric to

target and concluded that PCE ination is the

better measure. In my view, headline PCE

should become the standard and, thereore,

should be consistently used to estimate

and adjust or ination. Although adopt-

ing a standard measure would likely not be

a simple matter, it would provide clarity to

the public about which one more accuratelyreects consumer price ination.

FRED is a registered trademark o the Federal Reserve Bank o St. Louis.

th rn ecn | www.stlouised.org 3


4/24

SOLD

9.33% 256.36 2.78 56.39 5.23 6.35

4

F i n a n c i a l s Y s T e M

Financial MarketsAn Engine or Economic GrowthBy Yongseok Shin

In the aermath o the 2008 nancial crisis, it is natural to

wonder about the roles that the highly developed nancial sector

plays in our economy. Some might wonder whether this sector

causes more harm than it does good. In this article, I examine

data rom countries with varying degrees o economic

development and argue that developed nancial markets

are an essential ingredient o long-run economic growth.



5/24

2 9.61 2.86% 189.45 5.11 8.75 4.01%

2

Beore I begin, let me clariy two things.

First, it is not my contention that allnancial market activities have a posi-

tive impact on economic growth. o the

contrary, excesses and abuses in nancial

markets can be detrimental to economic

growth in the long run. Second, developed

nancial markets provide useul services

that do not directly contribute to economic

growth. For example, most insurance

policies are designed to enhance economic

welare through better allocation o risk,

not through the promotion o economic

growth. More broadly, the purpose o this

article is not to list all the pros and cons o

nancial market development. Rather, I

show the importance o nancial markets to

economic growth. Knowing the important

contributions o well-unctioning nancial

markets will help us gure out (1) which

nancial market activities to promote and

(2) where to direct our regulatory and

supervisory eorts.

The Schumpeterian Hypothesis

Te nexus o nance and economicgrowth was rst emphasized by Joseph

Schumpeter in 1911. In Schumpeters

theory, widely known as the theory o cre-

ative destruction, innovation and entrepre-

neurship are the driving orces o economic

growth. He viewed nance as an essential

element o this process. Innovation and

entrepreneurship will thrive when the

economy can successully mobilize produc-

tive savings, allocate resources efciently,

reduce problems o inormation asymmetry

and improve risk management, all o whichare services provided by a developed nan-

cial sector.

Te surest way to test such a hypoth-

esis would be to perorm a randomized,

controlled experiment, in which we would

improve nancial markets in a randomly

chosen group o countries and shut down

nancial markets in the others. Since it is

not possible (or desirable) to conduct such

experiments on national economies, econo-

mists have tried to iner the importance onance or economic growth rom observa-

tions on countries with varying degrees o

nancial and economic development.

Te rst attempts at empirical evaluations

o Schumpeters hypothesis came in the late

1960s and the early 1970s; these attempts

documented close relationships between

nancial development and economic

development across countries.1 However,

critics reuted this evidence, rightly, since

correlation does not imply causation. Many

prominent economists argued that nance

simply ollows economic development.2

More recently, researchers have responded

to this criticism. I highlight three dierent

approaches in this article.

Empirical Patterns across Countries

First, in a 1993 paper, Robert King and

Ross Levine addressed the correlation-not-

causation issue by showing that countries



6/24

with higher levels o nancial development

in 1960 experienced higher rates o economic

growth in the ollowing three decades. King

and Levine measured a countrys nancial

development in terms o the levels o credit

(e.g., bank loans and bonds issued) and stock

market capitalization, a metric that is still

widely used. Based on their ndings, they

rejected the idea that nance merely ollows

economic growth. But their results did not

proveor at least two reasonsthat nance

causes economic growth.

First, even though a countrys nancial

development in 1960 is a predetermined

variable relative to the economic growth in

the next three decades, both nancial and

economic development may still be mere

consequences o a common omitted ac-

tor. Second, because nancial markets are

orward-looking, nancial development in

1960 may be the consequence o anticipatedeconomic growth o the next ew decades.

In this reverse causality view, nancial

development may be a mere leading indica-

tor o economic growth rather than a cause.

Industry-Level Evidence

Researchers then tried to come up with

ways o testing Schumpeters hypothesis that

could surmount the above criticisms and

clearly determine causality. In an inuential

paper in 1998, Raghuram Rajan and Luigi

Zingales worked with detailed rm-level datathat had not been used in the literature until

then to test Schumpeters hypothesis. Teir

theory is that, i Schumpeter were correct,

industries that are more dependent on exter-

nal nancing would grow aster in countries

with more-developed nancial markets.

Using a database o publicly traded rms in

the United States (Compustat), they ranked

industries in terms o external dependence,

which is a measure o how dependent an

industry is on external nancing. Roughly

speaking, it is the raction o a rms invest-ment in a given year that is nanced with

debt and equity, rather than the years cash

ow.3 Tere is a large variation in external

dependence across industries, with phar-

maceuticals having the highest (1.49) and

tobacco the lowest (0.45).4

Rajan and Zingales ound that industries

that are more dependent on external nanc-

ing grew aster than those industries that

are less dependent on external nancing in

FIGURE 1

Relationship between Financial Development and Economic Development

FIGURE 2

Relationship between Financial Development and Manuacturing-ServicesRelative Productivity

SOURCE: Buera, Kaboski and Shin.


NOTE: In the right panel, the 18 countries are: Australia (AUS), Austria (AUT), Belgium (BEL), Czech Republic (CZE), Denmark (DNK), Finland (FIN), France

(FRA), Germany (GER), Hungary (HUN), Ireland (IRL), Italy (ITA), Japan (JPN), the Netherlands (NLD), Portugal (PRT), Spain (ESP), Sweden (SWE), the

United Kingdom (GBR) and the United States (USA).

0 0.7 1.4 2.1 2.8 0 0.7 1.4 2.1 2.8

1.0

0.8

0.6

0.4

0.2

0

1.0

0.8

0.6

0.4

0.2

0

GDP/WORKERRELATIVETOTHEU.S.

TOTALFACTORPRODUCTIVITY(TFP)REL

ATIVETOTHEU.S.

EXTERNAL-FINANCE-TO-GDP RATIO EXTERNAL-FINANCE-TO-GDP RATIO

Each dot is a country. Each dot is a country.

0 0.5 1.0 1.5 2.0 2.5 0 0.5 1.0 1.5 2.0 2.5

2.5

2.0

1.5

1.0

0.5

0

0.5

1.0

2.5

2.0

1.5

1.0

0.5

0

0.5

1.0

MANUFACTURING-

SERVICE

SRELATIVEPRICE

MANUFACTURING-

SERVICESRELATIVETFP

EXTERNAL-F INANCE-TO-GDP RAT IO EXTERNAL-F INANCE-TO-GDP RATIO

Each dot is a country.

AUSAUT

BEL

CZE DNK

ESP

FIN

FRAGER

HUN

IRL

ITA

JPN

NLD

PRT

SWE

GBR USA



7/24

countries with developed nancial markets,5

but it is the other way around in countries

with underdeveloped nancial markets.

Tey concluded that their result is consistent

with the view o nance as a lubricant, just

as Schumpeter hypothesized.

While their test result is not a proo

o nance as a causal actor o economic

growth, many economists count it as the

most convincing evidence. Te reason is

that it is much harder, albeit not impossible,

to come up with a plausible omitted-variable

argument or reverse-causality argument

on the relative perormance o industries

across countries.

Building an Economic Laboratory:

A Model with Two Sectors

One weakness o the above empiri-

cal approaches is that the ndings do not

shed much light on the exact mechanismthrough which nance aects economic

growth. o answer this question, the third

and nal approach that I discuss here takes

a dierent tack. Indeed, it turns the previ-

ous approaches on their head. It starts

by building an economic model whereby

nancial markets do have an impact on the

long-run economic growth. Te question

is not whether nance is a causal actor or

economic development (which is true by

assumption) but how big an impact nancial

development has on economic development.We can also determine the exact channels

through which nance aects economic

development.

For a representative and concrete example

o this modeling approach, I rely heavily

on a study that I conducted with Francisco

Buera and Joseph Kaboski in 2011, in which

we built a model with multiple industrial

sectors and with rictions in nancial

markets that interere with efcient alloca-

tion o resources. Te modeling o multiple

industrial sectors was partly motivated bythe ndings o Rajan and Zingales.

We started by establishing important

empirical acts on cross-country dier-

ences in economic development. First,

countries levels o nancial development

are closely correlated with their levels o

economic development measured by output

per worker. Second, poor countries low

levels o output per worker are primarily

explained by their low levels o total actor

productivity (FP). FP measures the level

o the technology that combines capital

and labor to produce output. A country

with a high FP produces more with a

given amount o capital and labor than a

country with a low FP. Finally, the FP

gap between rich and poor countries varies

systematically across industrial sectors o

the economy. For instance, less-developed

countries are particularly unproductive in

producing manuactured goods, includ-

ing equipment and machinery. Tese acts

synthesize the ndings o the two empirical

studies discussed above and shi the ocus

onto an economys FP rather than income

or output levels.

Te le panel o Figure 1 shows the rela-

tionship between a countrys nancial devel-

opment, measured by the ratio o private

credit to gross domestic product (ollowing

the metric o King and Levine), and its levelo economic development, measured by

output per worker.6 Each dot is a country,

and the tted straight line shows the average

relationship between the two variables. Te

output per worker is relative to the output

per U.S. worker. Te gure conrms that

more-developed economies also have more-

developed nancial markets.

Te right panel o Figure 1 shows the

relationship between a countrys nancial

development and its level o aggregate FP.

Te gure is a reection o the act thatthe dierence across countries in terms o

economic development, measured in terms

o output per worker, is primarily explained

by the dierence in their FP levels.

o have a clear analysis, we consider

the simplest multisector economy: an

economy with two sectorsmanuacturing

and services. We ocus on the scale di-

erences between manuacturing produc-

tion and services production. On average,

, i u i u

Bu J Kbk 2011,

bu u u

k

u.

manuacturing operates at larger scales,

which translates into more dependence on

external nancing.7

Sector-level FP data are not available or

most countries. We take advantage o the

standard economic theory which implies

that the relative price between the output o

two sectors is the reciprocal o their relative

productivity. In the le panel o Figure 2,

we show the positive correlation between

a countrys relative price o manuactured

goods to services and its level o nancial

development. Tis can be interpreted as

lower relative FP o manuacturing to

services in countries that are less nancially

developed.

In the right panel o Figure 2, we only

look at countries with sector-level FP

data and show their relative manuacturing-

services FP against their level o nancial

development. We veriy that, or thesecountries, the relative sector-level FP

data are consistent with the sector-level

relative prices.

Te primary goal o our 2011 study was

to present a rich quantitative ramework

and analyze the role o nancial rictions in

explaining the above empirical regularities

in economic development.

In our theory, a rms productivity

changes over time, generating the need to

reallocate capital rom previously produc-

tive rms to currently productive ones.Financial rictions hinder this reallocation

process by limiting the amount o credit

required or the expansion o newly produc-

tive rms. Te degree o nancial ric-

tions is dierent across countries because

countries dier in terms o the eectiveness

with which credit contracts are enorced. In

countries with ineective contract enorce-

ability, creditors are likely to have trouble

recovering their loans. Knowing this, they



8/24

will reduce the size o loans and demand

larger collateral.

We discovered that nancial rictions

explain a substantial part o the above devel-

opment regularities. Essential ly, nancial

rictions distort the allocation o capital

across rms and also their entry and exit

decisions, lowering aggregate and sector-

level FP. While the use o internal unds

or sel-nancing can alleviate the resulting

misallocation, it is inherently more difcult

to do so in sectors with larger scale and

larger nancing needs. Tus, sectors with

larger scale (i.e., manuacturing) are aected

disproportionately more by nancial ric-

tions. Tis explains the empirical ndings

o Rajan and Zingales.

Te variation in nancial development

across countries can explain a actor-o-

two dierence in output per worker across

economies, which is equivalent to almost80 percent o the dierence in output per

worker between Mexico and the U.S. Con-

sistent with the consensus view in the litera-

ture, the dierences in output per worker in

our model are mostly accounted or by the

low FP in economies with underdeveloped

nancial markets.

In our model economy, the impact o

nancial rictions is particularly large in the

large-scale, manuacturing sector. While

the sector-level FP declines by less than

30 percent in services, it declines by morethan 50 percent in manuacturing, a result

broadly in line with the available sector-

level productivity data shown in the right

panel o Figure 2. Te dierential impacts

o nancial rictions on sector-level pro-

ductivity are reected on the higher relative

prices o manuactured goods to services in

nancially underdeveloped economies.

Our analysis provides a clear decomposi-

tion o the main margins distorted by nan-

cial rictions. First, or a given set o rms

in operation, nancial rictions distort theallocation o capital among them (misallo-

cation o capital). Second, or a given num-

ber o rms in operation, nancial rictions

distort rms entry decisions, with produc-

tive-but-undercapitalized rms delaying

their entry and unproductive-but-cash-rich

rms remaining in business (misallocation

o entrepreneurial talent). Tird, nancial

rictions distort the number o rms operat-

ing in each sector. In our model economy,

whereas the misallocation o capital is

responsible or 90 percent o the eect o

nancial rictions on the service-sector

FP, it is the misallocation o entrepreneur-

ial talent that accounts or more than

50 percent o the eect on the manuactur-

ing-sector FP.

Te dierential impacts o nancial ric-

tions across sectors in our model economy

produce an interesting testable implication

on the rm size distribution o each sector.

Financial rictions, together with the result-

ing higher relative price o manuactured

goods, lead to too ew rms and too large

rms in manuacturing, and too many rms

and too small rms in services. o evalu-

ate this implication, we perorm a detailed

case study o Mexico and the U.S., and nd

empirical support or it.

Figure 3 plots the average plant size in

Mexico (dened as the number o employ-ees, vertical axis) against the average plant

size in the U.S. (horizontal axis) or 86

manuacturing industries and 12 service

industries. Te overall average plant size is

substantially smaller in Mexico than in the

U.S., almost by a actor o three. However,

many industries (those lying above the

45-degree dashed line) have an average plan

that is larger in Mexico than in the U.S.

Indeed, the data have a slope (solid line)

that is signicantly steeper than the

45-degree line. Tat means that the indus-tries that are large scale in the U.S. have

an even larger scale in Mexico, while those

that are small scale in the U.S. have an even

smaller scale in Mexico. With the exception

o administration/management services,

those above the 45-degree line are manuac-

turing industries.

In summary, we developed a theory

linking nancial development to output per

worker, aggregate FP and sector-level rela-

tive productivity. Financial rictions distort

the allocation o capital and entrepreneurialtalent and have sizable adverse eects on

macroeconomic outcomes. Based on these

ndings, we concluded that nancial devel-

opment, so long as it removes or alleviates

such rictions, promotes economic growth

in the long run.

Legal Origins and Financial Development

Empirical and theoretical analyses o

nance and economic development across

FIGURE 3

Average Establishment Size o Industriesin the U.S. and Mexico

0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0

8.0

7.0

6.0

5.0

4.0

3.0

2.0

1.0

0

MEXICO VS. U.S. SCALES: DISAGGREGATED INDUSTRIES

LOG WORKERS PER ESTABLISHMENT IN THE U.S.

Each data point is an industry.

**** *

**** **

*** **

* * ** **

** *******

* ** **** *** **

** ******** * **

**** *** ** **

** *****

** ***

****

**

**

Manufacturing

Services

Auto

SteelComputers

AV Equipment

Dairy ProductsClay Products

Retail Bakery and Tortilla

Food and Accommodations

Transportation

Admin. and Mgmt.

LOGWORKERSPERESTABLISHMENTINMEXICO




9/24

E N D N O T E S

1 Te Schumpeterian hypothesis had been much

debated beore then, but the relevant data required

or an empirical analysis were not available beore

the late 1960s.2 Joan Robinson argued, By and large, it seems to

be the case that where enterprise leads, nance

ollows. See p. 86 o her book in the reerences.3 A rms external dependence is dened as capital

expenditures (investment) minus cash ow rom

operations, divided by capital expenditures. Tis

reveals what raction o a rms investment is

nanced with internal unds (cash ow) and

external unds. An industrys external depen-

dence is then dened as the median value o the

rm-level external dependence o all the rms

in that industry. Rajan and Zingales urther

assume that an industrys external dependence is

a technological eature o the industry and, hence,

the external dependence o an industry computed

rom the U.S. data is common across all countries4 Te external dependence in the data primarily

depends on two actors. First, industry-leve l

technologies are dierent in the lag between

investment and revenue generation. It is longer

in pharmaceuticals, in which it takes years o

research and development to produce marketable

new drugs. obacco rms, on the other hand,

have a stable revenue stream that can more than

pay or new investments. Second, in all industri es

young rms have higher external dependence than

mature rms, which can use the proceeds rom

their past investment to pay or current invest-

ment. It turns out that most pharmaceutical rms

are young, and most tobacco rms are old.5 Rajan and Zingales measured a countrys nancial

development rst in terms o the metric o King

and Levine and then in terms o the degree o dis-

closure prescribed by each countrys accounting

standards.6 Gross domestic product (GDP) is computed in

international prices to account or the act that the

same goods and services are oen cheaper in poor

countries than in rich countries. Economists call

this procedure purchasing-power parity (or PPP

adjustment. Te data are or 1996 and come rom

Penn World ables Version 6.1.7 In the U.S., the average number o employees or a

manuacturing est ablishment is 47, while it is 17 or

a service establishment. Across all the Organisa-

tion or Economic Co-operation and Development

member countries, the average manuacturing rm

hires 28 employees and the average service rm 8.8 See La Porta et al.

R E F E R E N C E S

Buera, Francisco J.; Kaboski, Joseph P.; and Shin,

Yongseok. Finance and Development: A ale o

wo Sectors. TeAmerican Ec onomic Revie w,

2011, Vol. 101, No. 5, pp. 1,964 -2,002.

King, Robert G.; and Levine, Ross. Finance and Growth:

Schumpeter Might Be Right. Te Quarterly Jour-

nal o Economics, 1993, Vol. 108, No. 3, pp. 717-37.

La Porta, Raael; Lopez-de-Silanes, Florencio;

Shleier, Andrei; and Vishny, Robert W. Law and

Finance. Journal o Political Economy, 1998,

Vol. 106, No. 6, pp. 1,113-55.

Rajan, Raghuram G.; and Zingales, Luigi. Financia

Dependence and Growth. TeAmerican Eco -

nomic Review, 1998, Vol. 88, No. 3, pp. 559-86.

Robinson, Joan. Te General isation o the General

Teory, in Te Rate o Interest and Other Essays.

London: Macmillan, 1952.

countries naturally raise the ollowing

questions. Why are some countries more

nancially developed than others? Why

dont less developed countries adopt or

import more-advanced nancial markets?

Recent research on this topic nds answers

in countries institutions, especially their

legal ramework and rule o law.

For most countries, their overarch-

ing legal ramework was either shaped

long beore the emergence o the modern

nance-growth nexus or imposed on them

through colonial rule. Legal scholars have

categorized the laws that pertain to eco-

nomic and nancial contracts into our

traditions: (English) common law, French

civil law, German civil law and Scandina-

vian civi l law. Te scholars have ound that

common-law countries generally have the

strongest, and French-civil-law countries

the weakest, legal protections or investors,with German- and Scandinavian-civil-

law countries in the middle. Te strength

o investor protection explains, in turn, a

signicant raction o the dierences in

nancial development across countries.8

Tis nding also explains why it may

be difcult or countries to improve their

nancial markets, at least in the short term.

Financial markets are governed by rules that

are embedded into the institutional ounda-

tions o an economy, and such rules are

persistent and sluggish by nature. A reormo nancial markets, thus, likely presup-

poses an all-reaching, large-scale reorm o

the whole economy.

Policy Implications

Our analysis shows that, when the nan-

cial markets are not unctioning properly,

there is room or a government to intervene

and improve upon the allocation o capital

across rms. Indeed, this is one o the most

cited justications or industrial policy.

Tere are two important caveats. First, torepeat the popular rerain against industrial

policy, governments cannot pick winners

that is, it is not clear whether governments,

even with the best o intentions, can better

identiy who deserves more capital than

can the market. Economic history shows

that the odds are not in governments

avor. Second, it is hard to change policies

that avor particular groups once those

policies are instituted. A rm may well

deserve the governments directed credit

initially, but the rm will become over time

either unproductive or sufciently capital-

ized on its own. I the government cannot

wean such undeserving beneciaries rom

directed credit, the governments eorts

only worsen the misallocation o capital in

the long run.

Te studies reviewed in this article sug-

gest that governments aiming or nancial

development should ocus on reorming

bureaucratic and judicial procedures o

the enorcement o economic contracts.

With transparent and eective contract

enorcement in place, nancial development

will ollow.

Concluding Remarks

Tis article is not intended to be a whole-

sale deense o the nancial sector. Rather,

my goal is to remind us o the essential

services that a developed nancial sector

provides or technological innovation and

economic growthmobilizing savings,

evaluating projects, managing risk, moni-

toring managers and acilitating transac-

tions, just as Schumpeter envisioned.

We need to keep these essential services

in mind as we rethink our regulatory and

supervisory approaches in the wake o the

nancial crisis.

Yongseok Shin is an economist at the FederalReserve Bank o St. Louis. For more on hiswork, see http://research.stlouised.org/econ/shin/.



10/24

Changes in the Racial

Earnings Gap since 1960By Maria Canon and Elise Marifan

l a B o r i s s u e s

Income inequality between races hasbeen a widely used indicator o economicprosperity and opportunity (or the lack

thereo) within the diverse population o the

U.S. Te Civil Rights Act o 1964 prohibited

discrimination in public places, provided or

the integration o schools and other publicacilities, and made employment discrimina-

tion illegal, thus improving the quality o

education and providing more job opportu-

nities or Arican-Americans. Nevertheless,

disparities remain. Labor economists have

investigated various sources o earnings

inequality in America since the act was

passed; some economists have considered

how the disparities in earnings change

within and across regions o the country.

Much o the research covers the 1960-2000

period; much less is known about racialinequality in earnings over the years since.

O particular interest might be the impact o

the Great Recession on such inequality.

Tis article aims to provide insight into the

recent trends in earnings inequality between

black men and white men. We replicated the

analysis in a 2006 study by Jacob Vigdor o

the 1960-2000 period using census data and

then examined disparities in annual earnings

since then, using yearly American Commu-

nity Survey data rom 2000 to 2011.

Figures 1 (1960-2000) and 2 (2000-2011)present key results. Te dotted line shows

the percentage dierential in earnings or

Northern-born black males relative to North-

ern-born white males, holding constant other

variables.1 (For example, in 1960 Northern-

born black males earned on average 40 percent

less than their Northern-born white counter-

parts.) Te solid line plots the same com-

parison between Southern-born blacks and

whites. (Be aware that Northern-born and

Southern-born does not necessarily mean

that the men continued to live in the North

or South, respectively.)

In Figure 1, we see that inequality declined

among both the Northern-born and

Southern-born rom 1960 to 1970. Racial

earnings inequality among the Northern-born increased markedly rom 1970 to 1990

and remained relatively stable rom 1990

to 2000. On the other hand, among the

Southern-born, racial earnings inequality

declined only slightly rom 1970 to 1980 and

increased slightly rom 1980 to 2000. In 2000,

black-white earnings inequality among the

Northern-born was considerably greater than

the level in 1960, while inequality among the

Southern-born was reduced.Figure 2 shows the results or 2000-2011.

Te values or the Northern-born indicate

that the economic situation o blacks (as mea-

sured by annual earnings) declined consider-

ably relative to that o whites; in other words,

earnings inequality continued to increase or

those born outside the South.

Similarly, the percent dierential in

Southern-born blacks annual earnings rela-

tive to Southern-born whites worsened over

the 2000-2011 period. Tose blacks born in

the South did not show evidence o converg-

ing aster with those blacks born in the North

during this decade. In addition, the increase

in slope magnitude rom 2007 to 2010 indi-

cate that during the Great Recession and in

the year ollowing, racial earnings inequalityamong the Southern-born increased even

more than in previous years. Lastly, it is

important to note that being a Southern-born

black male corresponds with a greater wage

dierential relative to white counterparts

than does being a Northern-born black male.

For example, the results indicate that in 2011,

the annual earnings o Southern-born black

males were approximately 72 percent less

than those o Southern-born white males,

whereas Northern-born black males 2011

earnings were 61 percent less than those oNorthern-born white males.

What driving orces can explain

these trends?

Vigdor examined three hypotheses to

understand why it appears that the South

demonstrated more rapid progress than the

North in reducing the earnings gap between

blacks and whites rom 1960 to 2000.2 While

each hypothesis seems to have had an eect

at some point throughout the 40-year period,

the results o his analysis suggest that much

o the improvement in the racial wagegap in the South was merely a reection o

changing regional demographicswhat

he calls selective migrationand not o

actual improvement in relative earnings or

Southern-born blacks. Te improvements

were the result o blacks and whites o dier-

ing abilities moving rom South to North and

vice versa.

Vigdors results indicate that selective

migration accounted or 40 percent o the

mu

1960-2000 ; u

k bu

qu

. o u

b

g r

u qu.



11/24

Souths relative improvement rom 1960 to

2000 and all o the improvement rom 1980

to 2000. More specically, the black-white

wage gap improved among Southern resi-dents but not or those born in the South.

Other studies have attempted to explain

the trends in earnings inequality between the

races through 2000. A 2010 paper by Dan

Black, Natalia Kolesnikova and Lowell aylor

considered the average annual weeks worked

rom 1970 to 2000 and ound that this num-

ber declined or black men in each o the

14 cities examined, sometimes by as much as

25 percent, while the declines or white men

E N D N O T E S

1 Vigdor uses the term North to reer to Census

Bureau regions other than the South. Tereore,

the North in his study comprises the Northeast,

Midwest and West regions.2 His second hypothesis is that changes in regional

labor markets, ollowing the Civil Rights legisla-

tion and the manuacturing decline in the North,

improved Southern blacks economic prosperity.

His third hypothesis is that the Souths progress in

reducing the black-white earnings gap rom 1960 to

2000 could be a consequence o greater educational

attainment among blacks, ollowing desegregation

and the reduction o racial disparity in education.3 Te unemployment rate data are or black men age

20 and over, seasonally adjusted, and white men

age 20 and over, seasonally adjusted, rom the U.S.

Bureau o Labor Statistics/Haver Analytics.

R E F E R E N C E S

Black, Dan A.; Kolesnikova, Natalia A.; and aylor,

Lowell J. Te Economic Progress o Arican

Americans in Urban Areas: A ale o 14 Cities.

Federal Reserve Bank o St. Louis Review, Septem

ber/October 2010, Vol. 92, No. 5, pp. 353-79.

Kaplan, Greg; and Schulhoer-Wohl, Sam. Under-

standing the Long-Run Decline in Interstate

Migration. Federal Reserve Bank o Minneapolis

October 2012 (revised February 2013), Working

Paper 697.

Ruggles, Steven; Alexander, J. rent; Genadek, Katie;

Goeken, Ronald; Schroeder, Matthew B.; a nd Sobek

Matthew. Integrated Public Use Microdata Series

(IPUMS): Version 5.0 [Machine-readable database].

Minneapolis: University o Minnesota, 2010.

Vigdor, Jacob L. Te New Promised Land: Black-

White Convergence in the American South, 1960-

2000. National Bureau o Economic Research,

March 2006, Working Paper 12143.

FIGURE 1

Relation among Race, Geography andEarnings: 1960-2000

NOTES: The percentages on the ertical axis indicate, or example, that North-

ern-born black men made in 1960 about 40 percent less than Northern-born

white men. Samples are deried rom the Integrated Public Use Microdata

Series (IPUMS) census data on white and black males age 21-60 born in the

48 contiguous states. Indiiduals with zero earnings are assumed to hae

potential earnings below the median or their region/race cell and age.

Samples are weighted using IPUMS weights where appropriate.

1960 1970 1980 1990 20000

10

20

30

40

50

60

70

Black-White Gap among Northern-Born Men

Black-White Gap among Southern-Born Men

BLACKSEARNINGSRELATIVETOWHITES(%)

FIGURE 2

Relation among Race, Geography andEarnings: 2000-2011

NOTES: The percentages on the ertical axis indicate, or example, that North-

ern-born black men made in 2000 about 28 percent less than Northern-born

white men. Being Northern-born or Southern-born doesnt necessarily mean

that the men still lie in the North or South, respectiely. Samples are deried

rom the Integrated Public Use Microdata Series (IPUMS) American Community

Surey data on white and black males age 21-60 born in the 48 contiguous

states. Indiiduals with zero earnings are assumed to hae potential earnings

below the median or their region/race cell and age. Samples are weighted

using IPUMS weights where appropriate.

00 02 04 06 0801 03 05 07 09 10 110

10

20

30

40

50

60

70

80

Black-White Gap among Northern-Born Men

Black-White Gap among Southern-Born MenBLACKSEARNINGSRELATIVETOWHITES(%)

were relatively smaller. At the same time,

black mens weekly hours o work remained

stable. ogether, these two points suggest

that the earnings decline was likely related

to labor orce attachment (maniested as

a drop in the average number o weeks

worked in a year) rather than to declines in

the number o hours that black men were

working. Te authors ound substantial

declines in the proportion o black men

employed, increases in the proportion o

black men unemployed and even larger

increases in the proportion o black men

not in the labor orce. In other words,

black mens labor orce trends rom 1970 to

2000 help explain why their average annual

weeks worked declined and why their total

annual earnings declined relative to those

o white men.

Yet can these studies by Vigdor and by

Black et al. explain the behavior in racialearnings gaps over the 2000-2011 period?

While Vigdor argued that selective migra-

tion explained the Souths improvement

in racial earnings inequality relative to the

Norths, recent research by Greg Kaplan and

Sam Schulhoer-Wohl suggests that interstate

migration has been decreasing. In other

words, the selective migration story observed

in previous decades would no longer apply.

An alternative explanation or the

increased racial inequality could be that

Arican-American men were dispropor-tionately hit by the Great Recession. Teir

unemployment rate increased rom 8.5

percent to 15.4 percent between December

2007 and December 2011. In compari-

son, the unemployment rate o white men

rose rom 3.9 percent to 7.1 percent over

the same period.3 Given these unemploy-

ment rates, it is no surprise that or both

Southern-born and Northern-born blacks,

earnings declined relative to whites during

the Great Recession. Furthermore, the act

that the labor orce attachment or Arican-Americans has decreased even more since

the Great Recession might help explain the

increase in the earnings gap since 2009.

Maria Canon is an economist and Eli seMarifan is a research associate, both at theFederal Reserve Bank o St. Louis. For moreon Canons work, see http://research.stlouised.org/econ/canon/.



12/24

It is well-known that house prices declinedsharply and mortgage deaults increasedabruptly rom 2006 to 2010 in the U.S. In

Europe, where mortgage regulations are

signicantly dierent, the behavior o house

prices and mortgage deaults displays some-

what dierent dynamics. Comparing theexperiences in these two regions sheds light

on the impact o alternative regulations.

Figures 1 and 2 show the evolution o

house prices and mortgage deaults in the

U.S. and Europe, respectively. o acilitate

the comparison, both series are normalized

to 100 in 2007. In the U.S., house prices

declined about 20 percent during this period,

and deaults increased by about 300 percent.1

In Europe, house prices declined much less,

slightly more than 5 percent, while mortgage

deaults increased little, about 26 percentrom trough to peak, 2007-2010.2 Given that

changes in prices and deaults are dierent, it

is hard to compare the experiences in Europe

and the U.S. directly. Here, this problem is

dealt with by comparing changes in deaults

or periods in which the changes in prices

were similar in the two regions.

Te rst panel o the table compares

changes in mortgage deaults in Europe

and the U.S. or periods when the change

in prices was similar. From 2008 to 2009,

prices declined almost 7 percent on aver-age in Europe. As a response, deault rates

increased, but only by 11 percent. In the U.S.,

rom 2007 to 2008, house prices declined

on average by almost 8 percent. Te corre-

sponding increase in mortgage deaults was

much larger: more than 93 percent.

Why would there be such a dierence in

the response o mortgage deaults to almost-

equal changes in house prices? A 2011 report

by the International Monetary Fund (IMF)

points to two regulations used in Europe to

prevent mortgage deaults, one implemented

to a limited extent in only some states o the

U.S. and the other implemented on a much

less restrictive basis across the U.S.

Te rst regulation gives homeowners in

Europe more responsibilities aer deaultthan most U.S. homeowners ace. In Europe,

mortgages are recourse loans, meaning

that, aer deault, borrowers are respon-

sible or the dierence between the value o

the outstanding debt and the value o the

house. Consider this hypothetical case: I

Jaime bought a house in Spain or 500,000

in 2007 and deaulted in 2010 when he still

owed 450,000 but the house was worth only

400,000 then, under recourse laws he is

responsible or 50,000.

Tis policy increases the cost o deault,which makes it less appealing to the home-

owner. In most o the states in the U.S.,

mortgages are, in practice, nonrecourse.

Even when recourse is allowed, the de-

ciency judgment (the dierence between the

loan and house value) could be discharged

in bankruptcy.

Te second policy in Europe limits the

amount that households can borrow using

their house as collateral. Some European

countries have limits on loan-to-value (LV)

ratios o 80, 85 or 90 percent. For example,i the LV limit is 80 percent, an owner o a

house worth 500,000 cannot borrow (using

the house as collateral) more than 400,000.

As a result o this policy, households have

more home equity. More equity means that

ewer mortgages end up underwater when

house prices drop. As a result, the deault

rate is lower in Europe. In the U.S., LV

policies are much less restrictive.

Te impact o the recourse and LV

policies is illustrated in the rest o the table.

Te second panel o the table compares the

dynamics o house prices and deaults in state

with recourse laws to those in states without

recourse laws.3 We compare dierent periods

to evaluate the change in mortgage deaults

given similar changes in house prices. From2007 to 2010, house prices declined by about

9 percent in recourse states, while the deault

rate increased by about 217 percent. A very

similar change in pricesabout 10 percent

is observed or nonrecourse states between

2007 and 2008; or that group, deaults

rose about 186 percent, similar to what was

observed in recourse states. Te lesson here

is that recourse as designed and implemented

in the U.S. has little eect on the deault rate

on mortgages.4

As mentioned above, to understand whyrecourse does not have as much eect on

deault rates in the U.S. as it does in Europe,

one has to look at the interaction o recourse

laws in the U.S. with Chapter 7 bankruptcy.

In a 2009 paper, economists Wenli Li and

Michelle White estimated the probability o

bankruptcy or homeowners with mort-

gages and ound that the probability o ling

bankruptcy was about 25 times greater i the

mortgage creditor had begun oreclosure

within the previous three months than i the

mortgage creditor had not done so.5

Te third panel o the table illustrates that

recourse in Europe does play an impor-

tant role in preventing deaults. Te panel

compares a group o U.S. states with a group

o European countries; both groups have

recourse policies but no LV policies.6 Te

main dierence between these two regions is

how recourse regulations are actually imple-

mented, in particular, the act that Chapter 7

bankruptcy restricts the role o recourse in

Europe May ProvideLessons on Preventing

Mortgage DeaultsBy Juan Carlos Hatchondo, Leonardo Martinez

and Juan M. Snchez

h o u s i n g i s s u e s



13/24

the U.S. because a U.S. household can usually

discharge that obligation in bankruptcy.

Over roughly the same time period, house

prices in each group declined about the same

amount, but the increase in deault rates was

very dierent: about 14 percent in Europe

and about 217 percent in the U.S. Tis

suggests that recourse, when designed and

implemented as in Europe, plays an impor-

tant role in preventing deaults.

Limiting the amount o debt taken by

homeowners seems important, too. Te last

panel o the table compares European coun-

tries with and without LV limits. Over the

same period, each group experienced roughly

the same decline in house prices (about 10

percent). However, the deault rate increased

only slightly in countries with an LV limit,

while it increased by more than 14 percent incountries without such a limit.

E N D N O T E S

1 For prices and deaults or the U.S., we used

data provided by Zillow Real Estate Research.

Prices are rom the Zillow Home Value Index

or all homes, and deaults are oreclosures per

10,000 homes.2 Tese data are an average o prices and deaults

or seven European countries with available data

rom 2005 to 2010. Prices were obtained rom the

International House Price Database provided by

the Globalization and Monetary Policy Institute o

the Federal Reserve Bank o Dallas. Deaults are

actually arrears on mortgage or rent payment

provided by Eurostat. A more comparable concep

in the U.S. is mortgage delinquencies. Growth

rates o mortgage delinquencies and oreclosures

in the U.S. were similar during this period.3 States are grouped according to their recourse

policies, using the recourse classication rom

the 2011 paper by Andra C. Ghent and Marianna

Kudlyak. Te states without recourse policies

or which we also have price and deault data are

Arizona, Caliornia, Minnesota, Oregon, Wash-

ington and Wisconsin. Te states with recourse

policies that we used are Alabama, Arkansas, D.C

Maryland, Massachusetts and Missouri. Tese ar

the recourse states that take the shortest time to

resolve a oreclosure.4 See Clauretie. Tis view, however, is challenged b

Ghent and Kudlyak, using household-level data on

mortgage characteristics.5 In a related 2011 paper, Kurt Mitman models

dierences in bankruptcy a nd nonrecourse laws

across U.S. states.6 Te countries that are considered in Europe are

Denmark, France, Ireland, Italy, the Netherlands,

Spain and the United Kingdom. Te data on

loan-to-value ratio limits a re obtained rom the

IMF report mentioned above. Count ries with

maximum LV on new loans smaller than

100 percent are considered as countries with LV

limits. In our sample, only Denmark and Italy

belong to this group.7

See Hatchondo, Martinez and Snchez (2011).

R E F E R E N C E S

Clauret ie, errence. Te Impact o Interstate Fore-

closure Cost Dierences and the Value o Mort-

gages on Deault Rates. Journal o the American

Real Estate and Urban Economics Association ,

September 1987, Vol. 15, No. 3, pp. 152-67.

Ghent, Andra C.; Kudlyak, Marianna. Recourse an

Residential Mortgage Deault: Teory and Evi-

dence rom U.S. States. Te Review o Financial

Studies, 2011, Vol. 24, No. 9, pp. 3,139-86 .

Hatchondo, Juan Ca rlos; Martinez, Leonardo;

and Snchez, Juan M. Mortgage Deaults.

Working Paper 2011-019, Federal Reserve Bank o

St. Louis, 2011. See http://research.s tlouised.org/

wp/more/2011-019.International Monetary Fund, Global Financial

Stabilit y Report. April 2011.

Li, Wenli; White, Michelle J. Mortgage Deault,

Foreclosure and Bankr uptcy. Working Paper

15472, National Bureau o Economic Research,

2009.

Mitman, Kurt. Macroeconomic Eects o Bank-

ruptcy & Foreclosure Policies. Worki ng Paper

11-015, Penn Institute or Economic Research,

Department o Economics, University o Pennsyl-

vania, 2011.

At the Federal Reserve Bank o St. Louis,

a lie-cycle model in which households make

housing and nancial decisions is being

built.7 Te model reproduces many eatures

o U.S. mortgage and housing markets. Tat

articial economy can be used to simulate

the eect o implementing limits on LV and

recourse in the U.S. economy. Hopeully, the

results will shed light on the pros and cons o

implementing these policies.

Juan Carlos Hatchondo is an assistant proessorat Indiana University and an economist at theFederal Reserve Bank o Richmond. LeonardoMartinez i s an economist at the IMF Instituteor Capacity Development. Juan M. Snchez isan economist at the Federal Reserve Bank oSt. Louis. For more on Snchezs work, see

http://research.stlouised.org/econ/sanchez/.

PANEL 1 Europe U.S.

Period 2008-2009 2007-2008

Decline in prices 6.8% 7.7%

Increase in deaults 11.0% 93.2%

PANEL 2 U.S. recourse states U.S. nonrecourse states

Period 2007-2010 2007-2008



PANEL 3 U.S. recourse states Europe, non-LTV-limit countries

Period 2007-2010 2007-2009



PANEL 4 Europe, LTV-limit countries Europe, non-LTV-limit countries

Period 2007-2009 2007-2009



FIGURE 1 FIGURE 2

The Role o Recourse and LTV Limits in Preventing Mortgage Deaults

SOURCES: Zillow Real Estate Research, Federal Resere Bank o Dallas, Eurostat, Ghent and Kudlyak,

and Global Financial Stability Report by the International Monetary Fund.

2005 2006 2007 2008

YEAR

2009 2010 2011

105

100

95

90

85

80

75

250

200

150

100

50

Defaults (right axis)Prices

PR

ICES

(2007=

100)

DEF

AULTS

(2007=

100)

Prices vs. Deaults U.S. Prices vs. Deaults Europe

2005 2006 2007 2008

YEAR

2009 2010 2011

105

100

95

90

85

80

75

250

200

150

100

50

Defaults (right axis)

Prices

PR

ICES

(2007=

100)

DEF

AULTS

(2007=

100)

SOURCES: Prices rom International House Price Database proided by the

Federal Resere Bank o Dallas. Deaults rom Eurostat.

SOURCE: Zillow Real Estate Research.



14/24

Mortgage ApplicantsTurn to Credit Unionsater the Crisis

By Yang Liu and Rajdeep Sengupta

h o u s i n g i n T h e e i g h T h d i s T r i c T

he origins o the recent nancial crisishave oen been traced to the excessesin the U.S. mortgage market. Most accounts

o the crisis tend to ocus on a signicant

decline in underwriting standards or

mortgages since 2000. Aer the crisis, the

pendulum appears to have swung in theother direction. Anecdotal evidence sug-

gests that borrowers are nding it difcult to

obtain housing loans. Some observers have

remarked that this difculty may be one o

the causes o the slump in the U.S. market

or housing.

Using a data set o loan applications and

originations, we analyzed these trends or

the Federal Reserves Eighth District, based

in St. Louis.1 Our data came rom the Home

Mortgage Disclosure Act (HMDA) les or

2004, 2009 and 2010.2

Te HMDA data or2004 were used as an indicator o the pre-

crisis mortgage market conditions, whereas

HMDA data or 2009-2010 were used to

indicate postcrisis mortgage conditions. We

restricted our observations to rst-lien, one-

to our-amily home mortgage loans.

As expected, the data show that the nan-

cial crisis adversely aected the demand or

mortgage loans in the District. Figure 1

displays a panel o scatter plots showing

pre- and postcrisis mortgage applications in

each county o the District. Te horizon-tal axis o each plot measures the level o

2004 mortgage loan applications, while the

vertical axis measures the annual average o

2009-2010 mortgage loan applications. Each

dot in the chart represents one o the 339

counties. Te plot also shows the 45-degree

line where the level o 2004 applications

equals the annual average o 2009-2010

applications. Simply put, a dot below the

45 degree line indicates that postcrisis

applications or that county were ewer than

precrisis applications; a dot above indicates

the opposite.

For the District, there were 290,091 ewer

mortgage applications annually during

2009-2010 than in 2004 (a reduction o 33.3

percent). Figure 1A shows that 327 out othe 339 counties in the District were located

below the 45-degree linea widespread drop

in mortgage applications across the District.

Te drop was greater or new purchases

(Figure 1B) when compared with re-

nances (Figure 1C). Annual applications

or purchases ell by 47.3 percent (139,707

applications) aer the crisis; 319 counties

experienced a decline in purchase applica-

tions. In contrast, applications or renances

ell by 25.3 percent (138,634 applications);

307 counties experienced a decline in

renance applications. Clearly, the drop

in numbers was roughly the same or bothpurchases and renances, but purchases

constituted a smaller proportion o applica-

tions near the peak o the boom in 2004.

Interestingly, HMDA data also allowed

us to sort the applications by the agency

that supervises each lending institution to

which the application is made. Since di-

erent agencies supervise dierent types o

lending institutions, we could use this vari-

able to examine the dierences in pre- and

postcrisis applications by lending institu-

tions. We sorted loan data by three dierent

types o nancial institutions: banks and

thris, credit unions and HUD-supervised

mortgagees. Tis last category denotes

loans made by institutions that are not

supervised by any o the major agencies.3

Banks and thris in the District experi-

enced a moderate decrease in annual mort-

gage applications o 14 percent (or 71,738

applications) aer the crisis (Figure 1D).

Consumers led ewer mortgage application

loans to banks and thris in 252 counties.

HUD-supervised mortgagees suered the

largest loss in mortgage loan applications

on an annual basis (Figure 1F). Tey

received 229,219 ewer loan applications,

or a decline o 65.5 percent. In all but one

o the Districts counties, consumers ledewer mortgage loan applications to HUD-

supervised mortgagees. It is important

to point out that the reduction o 229,219

applications in this sector accounted or

79 percent o the annual loan application

decline in the District.

In contrast, credit unions enjoyed a

surprising boom in home mortgage appli-

cations (Figure 1E). On an annual basis,

mortgage applications rose by 10,813an

increase o 122 percent. O the 275 counties

in the District that recorded loan applica-tions led with credit unions, 222 coun-

ties recorded an increase in applications.

Furthermore, annual applications increased

by more than 100 percent in 123 District

counties.

Although urther research is needed

to account or this rapid and anomalous

increase, some anecdotal evidence may

explain this rapid growth in the popular-

ity o credit unions. First, there has been

au u

u

u.



15/24

E N D N O T E S

1 Te Eighth Federal Reserve District includes all

o Arkansas and portions o Illinois, Indiana,

Kentucky, Mississippi, Missouri and ennessee.2 In what ollows, we use the annual average or the

2009-2010 HMDA data. However, the choice o

years or pre- and postcrisis indicators is ad hoc.3 Te major supervisory agencies include the Fed-

eral Reserve System, Federal Deposit Insurance

Corp., Ofce o the Comptroller o the Currency,

Ofce o Tri Supervision and National Credit

Union Administration.4 See Prevost.5 Credit unions are nonprot depository institu-

tions that are democratically controlled by their

members. Membership in a credit union is usually

limited by law and is organized around a common

bond or eld o membership.6 See Morrison.7 Te term origination here implies the actual dis-

bursement o unds upon approval o the mortgage

applicati on. All originations require approval o

the mortgage application. However, not all

approved applications lead to originations since

the borrower can still reject the terms o the loan.8 A line o t (shown in Figure 2) is a line that is

drawn through the data on a scatter plot to

describe the trend o the data. Tis is dierent

rom the 45-degree line in Figure 1.9 A word o caution is in order here: While the plots

include condence intervals or the lines o t,

stricter criteria may not reveal statistically signi-

cant dierences between the lines o t in some

o the plots. Nevertheless, this remains a simple

and useul way to distinguish between pre- and

postcrisis origination rates.10 Te majority o the dots and crosses overlap in the

lower le o each gure.

R E F E R E N C E S

Prevost, Lisa. Te Credit Union Alternat ive,

Te New York imes, Dec. 13, 2012. See www.

nytimes.com/2012/12/16/realestate/mortgages-

credit-unions-grow-in-popularity.html?_r=0.

Morrison, David. wo Credit Unions Oeri ng

No Money Down Mortgages, Credit Union

imes, March 14, 2013. S ee www.cutimes .

com/2013/03/14/two-credit-unions-oering-

no-money-down-mortgages.

record growth in the membership o credit

unionsmuch o this has been attributed to

consumer disillusionment with big banks.4

Moreover, a large share o the growth in

mortgage business is concentrated among

the largest credit unionswhich typically

have lower limits on membership.5 Second,

at least two o these large credit unions have

reportedly been oering members mort-

gages without requiring any down payment

or mortgage insurance.6

o nd out how loan-approval patterns

in 2009-2010 diered rom those in 2004,

we examined the mortgage loan origination

rate during the two periods. Figure 2

displays a panel o scatter plots showing

In Figure 1, the horizontal axis o each panel shows the leel o 2004 mortgage loan applications, while the ertical axis

shows the annual aerage o 2009-2010 mortgage loan applications. Each dot in the chart represents one o the 339 coun-

ties in the District. The plot also shows the 45-degree line where the leel o 2004 applications equals the annual aerage

o 2009-2010 applications. Simply put, a dot below the 45-degree line indicates that postcrisis applications or that countywere ewer than precrisis applications; a dot aboe indicates the opposite.

Figure 2 displays a panel o scatter plots showing pre- and postcrisis mortgage origination is--is applications or each

county o the District. The horizontal axis shows the number o applications in the county, while the ertical axis measures

the number o originations. The dots in red show the 2004 leels or each county, while the blue dots show the annual aerage

or 2009-2010 in the same counties. The red and blue lines are the corresponding lines o ft or each period. A higher line

indicates a higher origination rate or a gien leel o applications.10

A. All Loans B. Purchase C. Renance

D. Banks and Thrifts E. Credit UnionsE. Credit Unions F. HUD-Supervised Mortgagees

Mortgage Loan Applications in the Eighth District Pre- and Postcrisis

FIGURE 1

POSTCRISIS

MORTGA

GE

LOAN

APPLICATIONS

PRECRISIS MORTGAGE LOAN APPLICATIONS

COUNTIES IN THE EIGHTH DISTRICT

96,000

64,000

32,000

0

39,000

52,000

26,000

13,000

0

30,000

20,000

10,000

0

2,700

1,800

900

0

66,000

44,000

22,000

0

42,000

28,000

14,000

0

0 32,000 64,000 96,000

0 13,000 26,000 39,000 52,000

0 10 ,000 20,000 30,000

0 900 1,800 2,700 0 14,000 28,000 42,000

0 22 ,000 44,000 66,000

A. All Loans B. Purchase C. Renance

D. Banks and Thrifts E. Credit UnionsE. Credit Unions F. HUD-Supervised Mortgagees

Mortgage Loan Applications vs. Originations in the Eighth District

FIGURE 2

MORTGAGE

LOAN

ORIGINATIO

NS

MORTGAGE LOAN APPLICATIONS

2004 EIGHTH DISTRICT COUNTIES 95% CONFIDENCE INTERVAL

2009-10 EIGHTH DISTRICT COUNTIES 2009-10 TREND LINE

2004 TREND LINE

60,000

40,000

20,000

0

39,000

26,000

13,000

0

21,000

14,000

7,000

0

2,100

1,400

700

0

36,000

24,000

12,000

0

24,000

16,000

8,000

0

0 32,000 64,000 96,000

0 18,000 36,000 54,000

0 10 ,000 20,000 30,000

0 900 1,800 2,700 0 14,000 28,000 42 ,000

0 22 ,000 44,000 66,000



16/24

Eleen more charts are aailable on the web ersion o this issue. Among the areas they coer are agriculture, commercial

banking, housing permits, income and jobs. Much o the data are specifc to the Eighth District. To see these charts, go to

www.stlouised.org/economyataglance.

U . S . A G R I C U L T U R A L T R A D E FA R M I N G C A S H R E C E I P T S

08 09 10 11 12 13

90

75

60

45

30

15

0

NOTE: Data are aggregated over the past 12 months.

Exports

Imports

MayTrade Balance

BILLIONS

OF

DOLLARS

08 09 1310 11 12

240

220

200

180

160

140

120

100

NOTE: Data are aggregated over the past 12 months.

March

Crops

Livestock

BILLIONS

OF

DOLLARS

C I V I L I A N U N E M P L O Y M E N T R AT E I N T E R E S T R AT E S

08 09 10 11 12 13

11

10

9

8

7

6

5

4

PERCENT

June

08 09 10 11 12 13

5

4

3

2

1

0

10-Year Treasury

Fed Funds Target

June

1-Year Treasury

PERCENT

NOTE: On Dec. 16, 2008, the FOMC set a target range for

the federal funds rate of 0 to 0.25 percent. The observations

plotted since then are the midpoint of the range (0.125 percent).

I N F L A T I O N - I N D E X E D T R E A S U R Y Y I E L D S P R E A D S R AT ES ON FE DE RA L F UN DS FU TU RE S O N S EL EC TE D D AT ES

3.0

2.5

2.0

1.5

1.0

0.5

0.0

0.5

NOTE: Weekly data.

5-Year

10-Year

20-Year

PERCENT

July 5

09 10 11 12 13 June 13 July 13 Aug. 13 Sept. 13 Oct. 13 Nov. 13

0.16

0.15

0.14

0.13

0.12

0.11

0.10

0.09

CONTRACT MONTHS

PERCENT

01/30/13

03/20/13 06/19/13

05/01/13

R E A L G D P G R O W T H C O N S U M E R P R I C E I N D E X ( C P I )

08 09 10 11 12 13

Q1

8

6

4

2

0

2

4

6

8

10

NOTE: Each bar is a one-quarter growth rate (annualized);

the red line is the 10-year growth rate.

PERCENT

08 09 10 11 12 13

6

3

0

3PERCENT

CHANGE

FROM

A

YE

AR

EARLIER

May

CPIAll Items

All Items, Less Food and Energy

e c o n o M Y a T a g l a n c e

pre- and postcrisis mortgage origination

vis--vis applications or each county o the

District. Te horizontal axis o each plot

shows the number o applications in the

county, while the vertical axis measures the

number o originations.7 Te dots in red

show the 2004 levels or each county, while

the blue dots show the annual average or

2009-2010 in the same counties.

We plotted the corresponding line o t

or each period.8 A higher line indicates a

higher origination rate or a given level o

applications.9 At rst glance, thereore, it

is surprising that the postcrisis line o t in

almost all plots o Figure 2 appears higher

than the precrisis trend lines. A possible

explanation o this eature o the data is

that although there are ewer applications

postcrisis, their quality is signicantly bet-

ter. Tis may be partly due to the act that

real-estate salesmen are only willing to dobusiness with preapproved buyers.

Figure 2 reveals two important patterns.

First, the dierences in origination rates or

renances (Figure 2C) appear to be greater

than those or purchases (Figure 2B). Re-

nancing aer a sharp decline in home prices

can be tricky because existing homeowners

would likely have to cover or the shortall in

home equity i they wanted to take advantage

o lower mortgage rates. While this reduces

the set o applicants, it can also ensure an

improvement in the applicant pool, therebyresulting in higher origination rates.

Second, among all lending institutions,

only credit unions loan origination rates

show a marginal decline (Figure 2E), primar-

ily due to smaller origination growth relative

to a larger increase in applications. In light

o the anecdotal evidence given above, a pos-

sible explanation is that a signicant increase

in annual mortgage applications made credit

unions more selective.

Rajdeep Sengupta is an economist ormerlywith the Federal Reserve Bank o St. Louis.Yang Liu is a senior research associate atthe Bank.



17/24

n a T i o n a l o v e r v i e W

Despite pockets o strength, the U.S.economy continues to struggle to buildconsistent momentum. Real GDP growth

rebounded in the rst quarter o 2013 aer

ending 2012 on a relatively weak note. Real

GDP grew at a 0.4 percent annual rate in

the ourth quarter but then sped up to a

modest 1.8 percent annual rate in the rst

quarter. Te momentum swing in the rst

quarter, though, was not expected to carry

into the second quarter. According to theMaySurvey o Proessional Forecasters, real

GDP growth was expected to slow to about

1.75 percent in the second quarter beore

rebounding to an average o about 2.5 per-

cent over the second hal o this year.

Housings Strength Spreads

Breaking down the GDP data indicates

that housing continues to be a source o

strength. Trough the rst ve months o

2013, new and existing home sales, as well as

housing permits, posted double-digit annu-alized growth rates compared with the same

period in 2012. Moreover, house prices rose

sharply, boosting the condence o home

builders. By contrast, commercial construc-

tion exhibited much less vigor.

Brisk gains on the housing ront are

beginning to boost other segments o the

economy. For example, the housing boom

appears to be triggering an upswing in

household spending. Trough the rst

our months o 2013, sales o household

urnishings and durable equipment likeappliances increased at about a 3.5 percent

annual ratemuch stronger than the

1.9 percent growth in total personal

consumption expenditures. Elsewhere,

automotive manuacturers have boosted

production o light trucks, which are used

extensively in the construction industry.

Te rebound in consumer spending,

at rst glance, is perhaps not too surpris-

ing, given other key developments. First,

Mixed Signals,but Moving Forward

By Kevin L. Kliesen

consumer condence and household wealth

rose sharply over the rst hal o 2013. Sec-

ond, gains in private-sector jobs averaged a

little more than 200,000 per month over the

rst six months o 2013.

However, other actors were workingin the opposite direction. Tese include

the payroll tax increase in January, higher

gasoline prices over the rst hal o the year,

tepid growth o real average hourly earnings

over the past ew years and the relatively

high levels o long-term unemployment.

Tese actors may help explain some o the

unexpected soness in total consumption

spending that occurred in April and May.

Despite healthy prot margins and a rela-

tively low cost o capital, real business xed

investment increased at just a 0.4 percentannual rate in the rst quarter and was up

only 3.7 percent rom our quarters earlier.

Similar to Aprils weak consumption data,

production o business equipment ell by

0.5 percent in April. However, there are

signs that business investment is picking up,

as new orders to manuacturers or capital

goods increased strongly in April and May.

Tus, consistent with most orecasts, the

data point to modest growth in the second

quarter. Moreover, with abundant levels o

cash on corporate balance sheets, it appearedthat many rms still harbored a considerable

amount o uncertainty about the near-term

outlook. Financial market conditions,

though, remain healthy, according to the

St. Louis Feds Financial Stress Index.

Few Worries on the Infation Front

Reecting a notable slowing in ood

price gains and sizable drop in consumer

energy costs, headline ination has been

exceptionally modest thus ar in 2013.

Trough the rst ve months o the year, the

consumer price index (the headline version,

which actors in ood and energy) increased

at only a 0.7 percent annual rateabout

1 percentage point slower than or the sameve-month period in 2012. Core ination

(excluding ood and energy) also slowed

relative to last year, but by not as much as

the headline ination rate. Over the rst

ve months o 2013, the core consumer

price index (CPI) advanced at a 1.8 percent

annual rate, 0.5 percentage points slower

than last years gain over the same period.

Blue Chip orecasters dont expect these

exceptionally low levels o ination to persist:

Te headline CPI is projected to increase at

about a 2 percent annual rate over the secondhal o this year. But, signals rom the bond

market suggest that longer-term ination

concerns appear relatively muted. In early

June, yields on ination-sensitive 30-year

reasury securities remained well below their

peak o 4.9 percent (in early April 2010) dur-

ing this business expansion.

On balance, stable ination expectations

and a lessening o some o the uncertainties

and headwinds that have hampered hiring

and business investment the past year or more

should lead to aster growth and low inationgoing orward. Indeed, this is the takeaway

rom the latest economic projections o the Fed-

eral Open Market Committee. (See chart.)

Kevin L. Kliesen is an economist at the FederalReserve Bank o St. Louis. Lowell R. Ricketts, asenior research associate at the Bank, providedresearch assistance. For more on Kliesens work,see http://research.stlouised.org/econ/kliesen/.

SOURCE: Board o Goernors o the Federal Resere System.

NOTES: Projections are the midpoints o the central tendencies. The actual and projected unemployment rates are

or the ourth quarter. The growth o gross domestic product (GDP) and personal consumption expenditures (PCE)

is the percentage change rom the ourth quarter o the preious year to the ourth quarter o the indicated year.

FOMC June 2013 Economic Projections or 2013-2015

10

8

6

4

2

0

2012 (Actual)

2013

20142015

PERCENT

REAL GDP UNEMPLOYMENT RATE PCE INFLATION

1.7

2.5

7.87.3

6.76.0

3.3 3. 3

1.61.0

1.7 1.8



18/24

In the postrevolutionary U.S., Louis-ville was an important western outpost.Situated at the Falls o the Ohio, Louisville

became a key port or the western rontier.

Similar to its inland-port contemporaries,

such as Cincinnati and St. Louis, Louisville

had an industrial river economy in the

beginning; growth was driven by heavy

manuacturing, shipping and trade. Lou-

isville also gained attention or its bourbon

whiskey, Louisville Slugger baseball bats and

Kentucky Derby, cultural hallmarks that livestrong today.

Postwar Louisville saw a movement away

rom heavy manuacturing and away rom

river trade, as production processes and labor

needs changed across the country. Coupled

with deurbanization and population loss,

Louisvilles economic transition was typical

o that o industrial cities. Atypical was

Louisvilles ease o adapting to a modern

postindustrial service economy.

Much o the MSAs growth over the period

can be attributed to Kentucky-based counties

surrounding the city. Spencer, Shelby and

Oldham counties grew the astest, with rates

o 31.7, 25.5 and 24.7 percent, respectively,

between 2002 and 2012. Tese three coun-

ties largely outpaced population growth in

Kentucky and the nation. As a result o this

growth, Spencer, Shelby and Oldham coun-

ties gained about 1.7 percent o Louisvilles

population; today, about 10 percent o the

population is located in these three counties.

Economic Drivers

Humana, a managed-health-care company

on the Fortune 100 list, has headquarters in

downtown Louisville. By revenue, Humana

is the largest publicly traded company based

in town. With 11,000 local employees, it is

the second-largest Louisville company by

local employee count.

Six o Louisvilles 10 largest employers

As it stands today, the city is a particularly

strong hub or the health-care and ood-

service industries. Logistics and distribution,

as well as recently expanding manuacturing,

are other industries o note.

Tis economic transition also helped

dampen urban population loss experienced

in similar cities and even helped garner

healthy population growth in recent years.

Over the past 10 years (2002-2012), Lou-

isvilles population increased by 9.7 percent,

noticeably aster than Kentuckys growtho 7.1 percent and just above the national

rate o growth, 9.1 percent. Within the

metro area, Jeerson County, Ky., holds a

substantial majority o Louisvilles popula-

tion: about 60 percent o the total. In the

past decade, there has been some shi in the

population: Jeerson County grew 7.3 per-

cent, while Clark County, Ind., the second-

largest county

Regional Economist - July 2013

Documents

Transcript of Regional Economist - July 2013