-
THORSTEN BECKASLI DEMIRGUC-KUNTLUC LAEVENROSS LEVINE
Finance, Firm Size, and Growth
Although research shows that financial development accelerates
aggregateeconomic growth, economists have not resolved conflicting
theoretical pre-dictions and ongoing policy disputes about the
cross-firm distributional ef-fects of financial development. Using
cross-industry, cross-country data, theresults are consistent with
the view that financial development exerts a dispro-portionately
positive effect on small firms. These results have implicationsfor
understanding the political economy of financial sector reform.
JEL codes: G2, L11, L25, O1Keywords: firm size, financial
development, economic growth.
ALTHOUGH RESEARCH SHOWS that financial development ac-celerates
economic growth (Levine 2006), economists have not resolved
conflictingtheoretical predictions about the distributional effects
of financial development. Some
We thank Pok-Sang Lam (the editor) and two anonymous referees
for comments that improved thepaper. We also thank Maria Carkovic,
Stijn Claessens, Bill Easterly, Alan Gelb, Krishna Kumar,
MichaelLemmon, Karl Lins, Alan Winters, and seminar participants at
the World Bank, University of Minnesota,New York University,
University of North Carolina, the University of Stockholm, Tufts
University, and theUniversity of Utah for helpful comments on
earlier drafts of the paper, and Ying Lin for excellent
researchassistance. We also thank Lori Bowan at the U.S. Census
Bureau for help with the U.S. Economic Censusdata on firm size
distribution. This paper was partly written while the third author
was at the World Bank.This papers findings, interpretations, and
conclusions are entirely those of the authors and do not
representthe views of the International Monetary Fund, the World
Bank, their Executive Directors, or the countriesthey
represent.
THORSTEN BECK is a Senior Economist, Development Research Group,
World Bank(E-mail: [email protected]). ASLI DEMIRGUC-KUNT is a
Senior Research Manager,Development Research Group, World Bank
(E-mail: [email protected]). LUCLAEVEN is a Senior
Economist, Research Department, International Monetary Fund,
Centrefor Economic Policy Research (CEPR), European Corporate
Governance Institute (ECGI)(E-mail: [email protected]). ROSS LEVINE
is the James and Merryl Tisch Professor of Eco-nomics, Department
of Economics, Brown University, National Bureau of Economic
Research(NBER) (E-mail: Ross [email protected]).Received October 26,
2006; and accepted in revised form February 19, 2008.Journal of
Money, Credit and Banking, Vol. 40, No. 7 (October 2008)C 2008
International Monetary Fund with Exclusive License to Printby The
Ohio State University
-
1380 : MONEY, CREDIT AND BANKING
theories imply that financial development disproportionately
helps small firms. Ifsmall firms find it more difficult to access
financial services due to greater informa-tion and transaction
costs, then financial development that ameliorates these
frictionswill exert an especially positive impact on small firms
(Cestone and White 2003,Galor and Zeira 1993). In contrast, if
fixed costs prevent small firms from accessingfinancial services,
then improvements in financial services will disproportionatelyhelp
large firms (Greenwood and Jovanovic 1990, Haber, Razo, and Maurer
2003).In this paper, we seek to provide empirical evidence to help
resolve this debate.
Besides assessing theoretical disputes, policy considerations
motivate our study ofthe distributional effects of financial
development. For example, if financial devel-opment helps small
firms more than large ones, then even if financial developmenthelps
all firms, large firms might oppose reforms that diminish their
comparativepower.1 Rather than analyzing political lobbying by
firms, we examine the more ba-sic question of whether financial
development has distributional effects. In addition,governments and
development agencies spend billions of dollars per year
subsidizingsmall firms, with the expressed goals of stimulating
growth, reducing poverty, andencouraging entrepreneurship.
Research, however, suggests that (i) subsidizing smallfirms does
not have these beneficial effects (Beck, Demirguc-Kunt, and Levine
2005),while (ii) improving the financial system accelerates growth
and alleviates poverty(Levine 2006, Beck, Demirguc-Kunt, and Levine
2007). In this paper, we test whetherfinancial development exerts a
disproportionately positive effect on small firms.
We examine whether industries that have a larger share of small
firms for tech-nological reasons grow faster in economies with
well-developed financial systems.As formulated by Coase (1937),
firms should internalize some activities, but size en-hances
complexity and coordination costs. Thus, an industrys technological
firmsize depends on that industrys particular production processes,
including capitalintensities and scale economies. After computing
an estimate of each industrys tech-nological share of small firms,
we use a sample of 44 countries and 36 industries inthe
manufacturing sector to examine the growth rates of different
industries acrosscountries with different levels of financial
development. If small-firm industriesindustries naturally composed
of small firms for technological reasonsgrow fasterthan large-firm
industries in economies with more developed financial systems,this
suggests that financial development boosts the growth of small-firm
industriesmore than large-firm industries. In contrast, we might
find that financial develop-ment disproportionately boosts the
growth of large-firm industries or that financialdevelopment
fosters balanced growth.2
1. A large literature examines the political economy of
financial policies, e.g., Kroszner and Stratmann(1998), Kroszner
and Strahan (1999), Rajan and Zingales (2003), Pagano and Volpin
(2005), and Perottiand von Thadden (2006).
2. Besides the argument that financial development
disproportionately helps large firms because smallfirms are cut off
from financial development, Petersen and Rajan (1994, 1995) show
that local bankingmonopolies foster close ties between banks and
small firms that ease credit constraints. Therefore,
financialdevelopment that intensifies competition and loosens these
ties might hurt small firms. On a global scale,Gozzi, Levine, and
Schmukler (2008) show that when financial development lowers
barriers to firmsaccessing international capital markets, it has
predominantly helped large firms.
-
THORSTEN BECK ET AL. : 1381
More specifically, we use a difference-in-differences approach
to examine whetherfinancial development enhances economic growth by
easing constraints on industriesthat are technologically more
dependent on small firms. We first measure an in-dustrys
technological composition of small firms relative to large firms as
theshare of employment in firms with less than 20 employees in the
United Statesin 1992. Assuming that financial markets are
relatively frictionless in the UnitedStates, we therefore identify
each industrys technological share of small firmsin a relatively
frictionless financial system. Then, we extensively test the
validityof this benchmark measure of technological Small Firm Share
by (i) using datafrom the United States in 1958 to compute Small
Firm Share, (ii) measuring SmallFirm Share at different stages of
the U.S. business cycle, (iii) computing techno-logical Small Firm
Share from different countries, and (iv) defining small
firmsdifferently.
The results indicate that small-firm industries grow
disproportionately faster ineconomies with well-developed financial
systems. This does not imply that financialdevelopment slows the
growth of large firms. Rather, financial development exerts
aparticularly positive growth effect on small-firm industries.
Furthermore, our analysessuggest that large-firm industries are not
the same as industries that rely heavily on ex-ternal finance.
Rajan and Zingales (1998) show that industries that are
technologicallymore dependent on external finance grow
disproportionately faster in economies withbetter developed
financial systems. When controlling for cross-industry
differencesin external dependence, we continue to find that
financial development dispropor-tionately accelerates the growth of
industries that are composed of small firms fortechnological
reasons.
We also show that the level of financial development affects
industrial composi-tion. In countries with greater financial
development, small-firm industries represent agreater proportion of
total manufacturing value added than in countries with lower
lev-els of financial development. Thus, financial development
disproportionately boostsboth the growth rate of small-firm
industries and the level of value added contributedby small-firm
industries to total value added.
The results also provide information regarding which particular
characteristics ofsmall-firm industries account for their greater
sensitivity to financial development.One possibility is that small
firms are more informationally opaque than large firms,so that
financial improvements that lower the marginal costs of acquiring
informationdisproportionately facilitate the flow of capital to
small firms. Another possibilityis that small firms rely more on
intangible assets, so that financial innovations thatreduce the
need for collateral ease credit constraints on small firms more
than largeones. A different possibility is that the results are
spurious and arise only becausesmall-firm industries enjoyed
greater growth opportunities than large-firm industriesover the
sample period. From this perspective, financially more developed
economieswere simply better at exploiting these growth
opportunities that happened to be con-centrated in small-firm
industries. If these potential characteristics of small-firm
in-dustries are driving the results, then our findings should
vanish when we control forthem.
-
1382 : MONEY, CREDIT AND BANKING
The results indicate that financial development still exerts a
disproportionatelypositive impact on small-firm industries even
when controlling for cross-industrydifferences in informational
opacity, asset intangibility, industry concentration, andgrowth
prospects. This suggests that financial development affects
small-firm indus-tries beyond opacity, collateral, and growth
prospects. Although we do not have directmeasures of firms access
to financial services, these findings are consistent with theview
that financial development makes it affordable for more small firms
to purchasefinancial services. Accordingly, the results suggest
that financial development influ-ences the extensive margin by
allowing new small firms to access financial servicesas well as
facilitating the intensive margin by improving financial services
for thosealready using the financial system.3
Our paper complements recent empirical work on finance and firm
size. Threeinfluential papers examine individual countries or
regions. Guiso, Sapienza, andZingales (2004) find that financial
development helps small firms more than largefirms in Italy.
Cetorelli and Strahan (2006) find that uncompetitive local
bankingmarkets in the United States represent a barrier to the
entry of new firms becausethe new firms have difficulty accessing
credit. Kumar, Rajan, and Zingales (2001)assess the impact of
different country and industry characteristics on industry
sizedistribution across 15 European countries. Our work builds on
this research. Ratherthan focusing on one country or region, or one
characteristic of financial developmentsuch as competition, we
examine a broad cross section of countries and test whetheroverall
financial development influences small-firm industries differently
from large-firm industries. Thus, we do not examine whether
financial reforms influence thedistribution of firms in a country
because (i) there are very limited cross-country dataon the
distribution of firm sizes and (ii) theory stresses the link
between financialmarket imperfections and small firms, not
necessarily the link between finance andentire distribution of firm
sizes in an economy. We instead examine whether industriesthat are
naturally composed of small firms for technological reasons perform
betterin countries with well-developed financial systems. Our
research also complementsthat work by Beck, Demirguc-Kunt, and
Maksimovic (2005), who use survey datato assess the relationship
between the financing obstacles that firms report they faceand firm
growth. They find that the negative impact of reported obstacles on
firmgrowth is stronger for small firms than large firms and
stronger in countries withunderdeveloped financial systems.4 Their
study has the advantage of using cross-country, firm-level data,
but it has the disadvantage of relying on survey responses
3. Although Beck, Demirguc-Kunt, and Maksimovic (Forthcoming)
show that small firms finance ahigher percentage of investment with
external finance in countries with stronger property rights
protection,we do not have direct evidence on fixed costs or on
whether a higher proportion of small firms accessesfinancial
services in more financially developed economies. Thus, we can only
draw the cautious conclusionthat the results are consistent with
the view that financial development lowers the fixed costs of
accessingfinancial services with disproportionately positive
ramifications on small firms.
4. Beck, Demirguc-Kunt, and Maksimovic (2006) find that
financial development reduces constraintson firms choosing their
optimal sizes.
-
THORSTEN BECK ET AL. : 1383
regarding the obstacles that firms encounter. We use a different
methodology thatassesses whether industries that are naturally
composed of small firms grow fasterin countries with
better-developed financial systems. Our research provides
comple-mentary information on whether financial development fosters
aggregate growth bydisproportionately facilitating the growth of
small-firm industries.
1. DATA
To assess whether (i) financial development boosts the growth of
small-firm indus-tries more than large-firm industries and whether
(ii) financial development booststhe level of output accounted for
by small-firm industries, we construct a new cross-country,
cross-industry database. We compile data on (i) the relative size
and growthrates of each industry across countries, (ii) each
industrys technological firm size,and (iii) country-level
indicators of financial development. This section describesthese
key variables. Furthermore, in robustness tests presented below, we
construct,define, and use additional information on industry and
country traits. The data cover44 countries and 36 industries in the
manufacturing sector. Tables 1 and 2 presentdescriptive and summary
statistics.
1.1 Industry Growth Rates and SharesGrowthi,k equals the average
annual growth rate of real value added of industry k
in country i over the period 1980 to 1990. The data are from the
Industrial StatisticsYearbook database. When we extend the
measurement period to 2000, the sample isreduced by one-third
because of missing observations for several countries and
indus-tries. Nevertheless, we demonstrate the robustness of these
findings to (i) expandingthe estimation period from 198090 to
19802000 and (ii) examining Growthi,k overthe period 19902000.
Industry Sharei,k is the share of industry k in total
manufacturing value added ofcountry i. Thus, besides testing
whether financial development has differential ef-fects on the
growth rate of large- and small-firm industries by examining
Growthi,k,we also examine whether financial development shapes the
level of industrial out-put patterns by examining Industry
Sharei,k. Specifically, we test whether a coun-trys level of
financial development shapes the cross-sectional distribution of
in-dustries by increasing the proportion of value added accounted
for by small-firmindustries.
Although we examine Industry Sharei,k, we focus on Growthi,k for
two reasons.First, building on Rajan and Zingales (1998), a large
literature examines the rela-tionship between financial development
and industry growth. This provides a naturalframework for our
analyses and facilitates comparisons, so that we identify an
inde-pendent relationship between financial development and the
growth rates of small-firm
-
1384 : MONEY, CREDIT AND BANKING
TABLE 1FIRM SIZE DISTRIBUTION IN THE UNITED STATES IN 1992
ISIC Industry name S5 S10 S20 S100
311 Food manufacturing 0.56 1.68 3.82 13.77313 Beverage
industries 0.60 1.76 4.04 14.75314 Tobacco manufactures 0.09 0.20
0.30 1.49321 Manufacture of textiles 0.40 1.17 2.81 13.43322
Manufacture of wearing apparel, except footwear 1.30 3.60 8.18
31.74323 Manufacture of leather and products of leather 1.94 4.78
10.45 36.89324 Manufacture of footwear 0.31 0.81 1.61 7.40331
Manufacture of wood and wood and cork products 4.20 11.20 21.37
47.31332 Manufacture of furniture and fixtures 1.57 4.19 9.09
28.74341 Manufacture of paper and paper products 3.03 16.16342
Printing, publishing, and allied industries 3.64 9.16 16.32
35.80352 Manufacture of other chemical products 0.87 2.68 5.80
17.67353 Petroleum refineries 0.05 0.18 0.36 1.90354 Manufacture of
miscellaneous products of petroleum and coal 1.26 3.93 9.26
29.80355 Manufacture of rubber products 0.38 1.21 3.15 13.23356
Manufacture of plastic products not elsewhere classified 0.69 2.24
6.09 27.19361 Manufacture of pottery, china, and earthenware 2.30
4.91 8.80 26.52362 Manufacture of glass and glass products 1.15
2.82 5.05 13.92369 Manufacture of other nonmetallic mineral
products 1.87 5.88 14.17 40.78371 Iron and steel basic industries
0.20 0.59 1.62 8.05372 Nonferrous metal basic industries 0.50 1.78
4.76 18.65381 Manufacture of fabricated metal products 1.28 4.07
9.98 33.87382 Manufacture of machinery except electrical 2.15 6.37
13.68 34.60383 Manufacture of electrical machinery apparatus, and
appliances 0.50 1.48 3.44 14.18384 Manufacture of transport
equipment 0.18 0.54 1.21 4.20385 Manufacture of professional and
scientific equipment 0.68 1.87 4.01 12.88390 Other Manufacturing
Industries 3.54 8.72 16.95 43.483211 Spinning, weaving, and
finishing textiles 0.26 0.73 1.91 9.143411 Manufacture of pulp,
paper, and paperboard 0.14 1.293511 Manufacture of basic industrial
chemicals except fertilizers 0.29 0.89 1.75 6.513513 Manufacture of
synthetic resins, plastic materials, and fibers 0.11 0.31 0.66
3.173522 Manufacture of drugs and medicines 0.26 0.86 2.10 8.093825
Manufacture of office, computing, and accounting machinery 0.48
1.32 2.85 10.433832 Manufacture of radio, television, and
communication equipment 0.57 1.40 3.09 11.673841 Ship building and
repairing 1.73 3.58 6.56 16.353843 Manufacture of motor vehicles
0.32 1.00 2.28 8.04Average 1.07 2.88 5.85 18.42
NOTES: This table shows employment shares by firm size bin in
the United States by ISIC Revision 2 industries. Sx is the
industrys shareof employment by firms with less than x employees,
and is calculated using data from the U.S. Census on all U.S. firms
for the year 1992.Employment shares are expressed in percentages of
total number of employees.
industries relative to large-firm industries above and beyond
the effects established bypast work. Second, focusing on growth
links helps link our paper to an extensive bodyof theoretical and
empirical work on the finance-growth relationship. As reviewedby
Levine (2006), many theoretical models predict that a higher level
of financialdevelopment will induce a faster rate of economic
growth, not just an increase in thelevel of economic development.
Thus, a higher level of financial development mightexert a
disproportionately positive effect on the growth rate of particular
types ofindustries, such as industries naturally composed of small
firms facing high informa-tional asymmetries. This further
motivates our focus on Growthi,k. Moreover, all ofthe results are
confirmed with Industry Sharei,k.
-
THORSTEN BECK ET AL. : 1385
1.2 Measure of Small Firm ShareWe construct measures of each
industrys natural or technological share of small
firms based on an extensive body of research on the theory of
the firm. As discussed, forexample, by Coase (1937) and Sutton
(1991), differences in productive technologiesinfluence an
industrys technological firm size. To get a proxy measure of each
indus-trys natural or technological share of small firms,
therefore, we need a benchmark
TABLE 2SUMMARY STATISTICS
Variable Mean Median St. dev. Minimum Maximum
Panel A. Country-industry variablesGrowth in real value added
0.034 0.029 0.099 0.447 1.000Industry share in value added 0.016
0.009 0.021 0.000 0.224Panel B. Industry variablesSmall Firm Share
(empl
-
1386 : MONEY, CREDIT AND BANKING
TABLE 2CONTINUED
Small Firm External Sales RatingShare dependence growth
Intangibility splits R2
Panel E. Cross-industry correlation of industry
characteristicsExternal dependence 0.16
(0.38)Sales growth 0.16 0.76
(0.39) (0.00)Intangibility 0.41 0.12 0.34
(0.02) (0.51) (0.05)Rating splits 0.24 0.09 0.10 0.19
(0.21) (0.66) (0.61) (0.33)R2 0.03 0.14 0.20 0.21 0.06
(0.87) (0.44) (0.27) (0.25) (0.75)Concentration 0.57 0.06 0.18
0.18 0.13 0.28
(0.00) (0.75) (0.32) (0.32) (0.50) (0.13)NOTES: This table
reports summary statistics and correlations for the main variables
in our analysis. Country-industry variables: Growthin real value
added is average growth in real value added over the period 198090
by country and ISIC industry. Industry share in valueadded is the
industrys share in total value added of the countrys manufacturing
sector in 1980. Industry variables: Small firm share is
theindustrys share of employment by firms with less than 20
employees, and is calculated using data from the U.S. Census on all
U.S. firmsfor the year 1992. Small firm share (empl
-
THORSTEN BECK ET AL. : 1387
only has relatively frictionless financial markets; it has
relatively frictionless marketsin general. Again, the United States
is a reasonable initial benchmark. The UnitedStates has the full
spectrum of human capital skills. Furthermore, comparative
studiesof U.S. and European labor markets suggest that the United
States has many fewerpolicy distortions. Moreover, the U.S.
internal market is huge andgiven its sizeitis comparatively open to
international trade. Many studies also point to the UnitedStates as
having a superior contracting environment and well-developed
institutions(Barth, Caprio, and Levine 2006).
The empirical methodology does not require that the United
States has perfectfinancial markets, labor markets, contracting
systems, or institutions. Rather, we re-quire that policy
distortions and market imperfections in the United States do
notdistort the ranking of industries in terms of the technological
share of small firmswithin each industry. Thus, we begin with the
following benchmark measure of eachindustrys technological share of
small firms.
Small Firm Sharek equals industry ks share of employment in
firms with lessthan 20 employees in the United States and is
obtained from the 1992 Census. Wemeasure Small Firm Share in 1992
because the U.S. Census did not start collectingcomprehensive firm
size distribution data at the firm level until 1992. For a less
refinedcategorization of firms by employment size, the data extend
back to 1958. We confirmthe findings with the 1958 data. In our
baseline regressions, we use Small Firm Shareas the measure of each
industrys natural or technological share of small firms.
Table 1 lists the Small Firm Share for each industry in the
sample. The SmallFirm Share has a mean of 6% but varies widely from
0.1% in manufacturing of pulp,paper, and paperboard to 21% in wood
manufacturing.5 We omit three industries withfewer than 10 firms
for each size bucket (Tobacco (ISIC 314), Petroleum refineries(ISIC
353), and Pulp and paper (ISIC 3411)) because the low number of
observationsmay impede an accurate estimate of the natural Small
Firm Share. Nevertheless, thepapers findings hold when including
these three industries.
Below, we present a large battery of sensitivity analyses of the
benchmark mea-sure of Small Firm Share. We use different measures
of Small Firm Share, differentbenchmark years from the United
States, different benchmark countries, and differ-ent cutoffs for
the definition of a small firm. We also control for numerous
indus-try traits, including asset tangibility and opacity, sales
growth, and dependence onexternal finance. We further condition on
country characteristics, including the level ofeconomic
development, labor market frictions, market size, and barriers to
firm entry.
1.3 Indicator of Financial DevelopmentIdeally, one would like
indicators of the degree to which the financial system
ameliorates information and transactions frictions and
facilitates the mobilization
5. Note that the share of small firms among all firms in the
United States is substantially higher thanfor the subset of listed
enterprises. From the Compustat database, the share of small listed
manufacturingfirms is only 0.009%. Thus, it is not useful to only
consider listed firms to assess which firm characteristicsdrive the
results.
-
1388 : MONEY, CREDIT AND BANKING
and efficient allocation of capital. Specifically, we would like
indicators that capturethe effectiveness with which financial
systems research firms and identify profitableprojects, exert
corporate control, facilitate risk management, mobilize savings,
andease transactions. Unfortunately, no such measures are available
across countries.Consequently, we rely on a traditional measure of
financial development that existingwork shows are robustly related
to economic growth.
Private Crediti equals the value of credits by financial
intermediaries to the privatesector divided by GDP for country i.
It captures the amount of credit channeled throughfinancial
intermediaries to the private sector. Levine, Loayza, and Beck
(2000) showthat Private Credit is a good predictor of economic
growth. In our baseline regression,we measure Private Credit in the
initial year of our estimation period, 1980 (or thefirst year in
which data are available), to control for reverse causation. Since
usinginitial values instead of average values implies an
informational loss, we also confirmthe robustness of the results
when using Private Credit averaged over the full period198089 and
employing instrumental variables to control for endogeneity. Data
forPrivate Credit are from Beck, Demirguc-Kunt, and Levine (2000).
There is widevariation in Private Credit, ranging from 7% in
Bangladesh to 117% in Japan. Below,we define and use several
alternative indicators of financial development, includinga measure
of stock market development.
2. METHODOLOGY
To examine whether industries that are naturally composed of
small firms growfaster than large-firm industries in countries with
higher levels of financial devel-opment, we interact an industry
characteristiceach industrys technological SmallFirm Sharewith a
country-characteristicthe level of financial development.
Indescribing the econometrics, we only discuss the interaction
between financial de-velopment and Small Firm Share. In the actual
implementation, we control for manyinteractions between country and
industry characteristics.
Consider the following regression:
Growthi,k =
ii Countryi +
kk Industryk + Industry Sharei,k
+ (Small Firm Sharek FDi) + i,k,where Growthi,k is the average
annual growth rate of value added, in industry kand country i, over
the period 1980 to 1990. Countryi and Industryk are countryand
industry dummies, respectively, and Industry Sharei,k is the share
of industry kin manufacturing in country i in 1980. Small Firm
Sharek is the benchmark shareof small firms in industry k, which in
our baseline specification equals the shareof employment in firms
with less than 20 employees in the United States in 1992.FDi is an
indicator of financial development for country i, which equals
PrivateCredit in our baseline regression. We include the
interaction between the share ofsmall firms in an industry and
financial development. We do not include financial
-
THORSTEN BECK ET AL. : 1389
development on its own, since we focus on within-country,
within-industry growthrates. The dummy variables for industries and
countries control for country- andindustry-specific characteristics
that might determine industry growth patterns. Wethus isolate the
effect that the interaction of Small Firm Share and Private
Credithas on industry growth relative to country and industry
means. By including theinitial share of an industry we control for
a convergence effect: industries with alarge share might grow more
slowly, suggesting a negative sign on . We includethe share in
manufacturing rather than the level since we focus on
within-country,within-industry growth rates. We exclude the United
States (the benchmark country)from the regressions.
The focus of our analyses is on the interaction between
financial development andSmall Firm Share; i.e., we focus on the
sign and significance of . If is positive andsignificant, this
suggests financial development exerts a disproportionately
positiveeffect on small-firm industries relative to large-firm
industries. This would suggestthat financial development tends to
ease growth constraints on small firms more thanon large firms.
We conduct the regression analyses under alternative assumptions
to assess thevalidity of the results. In the baseline regressions,
we use ordinary least squares (OLS),which assumes that the error
term is uncorrelated across both industries and countries.We then
relax these restrictions and allow first for correlation across
observations fromthe same industry and second for correlations
across observations from the samecountry. We thus present standard
errors based on clustering at both the industry- andcountry-level
in Table 3.6 For simplicity, we do not report standard errors based
onindustry- and country-clustering in the rest of the paper.
However, all of the papersfindings are robust to clustering at the
industry and country level and these results areavailable on
request.
3. RESULTS, EXTENSIONS, AND SENSITIVITY TESTS
3.1 Main ResultsThe results presented in Table 3 indicate that
small-firm industries (industries
with technologically larger shares of small firms) grow faster
in economies withbetter-developed financial intermediaries. The
interaction of Private Credit with SmallFirm Share enters
positively and significantly at the 5% level in column (1). We
alsofind that the coefficient on Industry Share enters negatively
and significantly, suggest-ing some convergence in industrial
composition. The results indicate that industrieswhose organization
is based more on small firms than on large firms grow faster
incountries with better-developed financial intermediaries.
6. Two-way clustering, at the country-industry level, is
infeasible in this papers econometric specifi-cation where (i)
there are only country-industry observations, (ii) country and
industry fixed effects areestimated, and (iii) the number of
clusters is small, with 42 countries and 33 industries. Under these
condi-tions, the asymptotic justification for the robustness of
clustered standard errors does not hold (Cameron,Gelbach, and
Miller 2006).
-
1390 : MONEY, CREDIT AND BANKINGTA
BLE
3FI
NA
NCI
AL
DEV
ELO
PMEN
T,SM
ALL
FIR
MSH
ARE
,A
ND
GR
OWTH
(2)(3)
(4)(5)
(6)(1)
Exte
rnal
Clus
terin
gat
Clus
terin
gat
Excl
udin
gin
dustr
ies
Oth
erin
dustr
yan
dB
asic
depe
nden
cein
dustr
yle
vel
cou
ntr
yle
vel
with
low
shar
eco
un
try
char
acte
ristic
s
Indu
stry
shar
ein
val
uead
ded
1.0
56
1.1
48
1.1
48
1.1
48
0.8
59
0.8
76
(0.28
4)(0.
282)
(0.33
4)(0.
442)
(0.19
5)(0.
256)
Priv
ate
Cred
it
Smal
lFirm
Shar
e0.
436
0.56
7
0.56
7
0.56
70.
698
0.
538
(0.18
7)(0.
194)
(0.18
4)(0.
306)
(0.19
6)(0.
228)
Priv
ate
Cred
it
Exte
rnal
depe
nden
ce0.
166
0.
166
0.
166
0.
141
0.
085
(0.04
3)(0.
040)
(0.05
9)(0.
041)
(0.03
8)Pr
ivat
eCr
edit
Sa
lesg
row
th0.
415
(0.40
8)Pr
ivat
eCr
edit
R
atin
gsp
lits
0.06
0(0.
073)
Priv
ate
Cred
it
R20.
003
(0.17
7)Pr
ivat
eCr
edit
Co
ncen
tratio
n0.
000
(0.00
1)Pr
oper
tyrig
hts
Inta
ngib
ility
0.00
5(0.
002)
Entry
regu
latio
n
Smal
lfirm
shar
e0
.761
(0.
309)
Ope
nnes
sSm
allF
irmSh
are
0.00
0(0.
002)
Perc
apita
GD
P
Smal
lFirm
Shar
e0
.068
(0.05
3)O
bser
vatio
ns1,
147
1,14
71,
147
1,14
758
691
8R2
0.27
0.29
0.29
0.29
0.56
0.33
Diff
eren
tiali
nre
algr
owth
rate
(%)
1.10
1.43
1.43
1.43
1.76
1.36
NO
TES:
Dep
ende
ntvar
iabl
eis
aver
age
grow
thin
real
val
uead
ded
over
the
perio
d19
80
90by
cou
ntr
yan
dIS
ICin
dustr
y.In
dustr
ysh
are
inval
uead
ded
isth
ein
dustr
ys
shar
ein
tota
lval
uead
ded
of
the
cou
ntr
ys
man
ufa
ctur
ing
sect
or.
Priv
ate
Cred
itis
clai
ms
byfin
anci
alin
stitu
tions
on
the
priv
ate
sect
ordi
vid
edby
GD
Pin
1980
.Sm
allfi
rmsh
are
isth
ein
dustr
ys
shar
eo
fem
ploy
men
tby
firm
sw
ithle
ssth
an20
empl
oyee
s,an
dis
calc
ulat
edu
sing
data
from
the
U.S
.Cen
sus
on
allU
.S.
firm
sfo
rth
eye
ar19
92.E
xter
nald
epen
denc
eis
am
easu
reo
fth
ein
dustr
ys
depe
nden
ceo
nex
tern
alfin
ance
,fro
mR
ajana
nd
Zing
ales
(1998
).Th
ein
dustr
ym
easu
res
are
base
do
nU
.S.d
ata.
The
stan
dard
erro
rsin
regr
essio
n(3)
are
adjus
tedfo
rclu
sterin
gat
the
indu
stry-
level
.The
stan
dard
erro
rsin
regr
essio
n(4)
are
adjus
tedfo
rclu
sterin
gat
the
cou
ntr
y-le
vel
.Th
ere
gres
sion
inco
lum
n(5)
excl
udes
indu
strie
sbe
low
the
med
ian
initi
alin
dustr
ysh
are
inval
uead
ded
for
each
cou
ntr
y.A
llre
gres
sions
excl
ude
indu
strie
sw
ithle
ssth
an10
firm
sin
each
size
buck
et;
thes
ear
e:To
bacc
o(IS
IC31
4),Pe
trole
umre
finer
ies
(ISIC
353),
and
Pulp
and
pape
r(IS
IC34
11).
Sale
sgr
owth
isan
indu
stry
mea
sure
of
sale
sgr
owth
from
Fism
anan
dLo
ve
(2007
)an
dis
calc
ulat
edas
real
ann
ual
grow
thin
net
sale
so
fU.S
.firm
sover
the
perio
d19
80
90.I
ntan
gibi
lity
isa
mea
sure
oft
hein
dustr
ys
depe
nden
ceo
nin
tang
ible
asse
tsfro
mCl
aess
ens
and
Laev
en(20
03)a
nd
isca
lcul
ated
asth
era
tioo
fint
angi
ble
asse
tsto
fixed
asse
tso
fU
.S.
firm
sover
the
perio
d19
80
90.R
atin
gsp
lits
isth
ein
dustr
y-av
erag
era
tioo
fbo
ndiss
ues
with
split
ratin
gsbe
twee
nS&
Pan
dM
oody
s
from
Mor
gan
(2000
).A
high
ersc
ore
indi
cate
sm
ore
indu
stry-
opaq
uene
ss.R
2is
indu
stry-
aver
age
R2fro
mD
urne
v,M
orck
,an
dYe
un
g(20
04).
Ahi
gher
sco
rein
dica
tes
mo
rest
ock
retu
rnsy
nchr
onic
ityan
dth
usle
ssin
form
ativ
epr
icin
g.Co
ncen
tratio
nis
the
four
-fir
mco
nce
ntr
atio
nra
tiofo
rU.S.
firm
sfro
mth
e19
92U
.S.C
ensu
s.Th
ein
dustr
ym
easu
res
are
base
do
nU
.S.d
ata.
Perc
apita
GD
Pis
the
loga
rithm
oft
heco
un
try
sre
alG
DP
perc
apita
in19
80.O
penn
ess
isth
esu
mo
fex
ports
and
impo
rtsre
lativ
eto
GD
Pin
1980
.Ent
ryre
gula
tion
isth
eco
sto
fen
try
regu
latio
nsas
ash
are
ofp
erca
pita
GD
Pin
1999
from
Djan
kov
etal
.(200
2).A
high
ersc
ore
deno
tesm
ore
cost
lyen
try
regu
latio
ns.A
llre
gres
sions
are
estim
ated
usin
gO
LSan
dinc
lude
cou
ntr
yan
dind
ustry
dum
mie
s,bu
tthe
sear
en
otr
epor
ted.
Rob
ust
stan
dard
erro
rsar
ein
pare
nthe
ses.
ind
icat
essig
nific
ance
at5%
level
; i
ndic
ates
signi
fican
ceat
1%le
vel
.
-
THORSTEN BECK ET AL. : 1391
Given the influential findings of Rajan and Zingales (1998), we
were concernedthat there might be a large, negative correlation
between industries that are naturallyheavy users of external
finance and industries that are naturally composed of smallfirms.
If this were the case, then it would be difficult to distinguish
between the findingthat externally dependent industries grow faster
in economies with well-developedfinancial systems and our result
that small-firm industries grow faster in economieswith
well-developed financial systems. While there is a negative
correlation betweenSmall Firm Share and External Dependence, it is
very small (0.16) and insignificantas shown in Table 2, Panel E.
This suggests that the industry characteristics explainingfirm size
distribution are not the same as the characteristics explaining
technologicaldependence on external finance, and that the firm size
channel we have identified isdifferent from the external financial
dependence channel.
The column (2) regression of Table 3 demonstrates the robust
link between financialdevelopment, Small Firm Share, and industry
growth when controlling for externaldependence. As shown in column
(2), the interaction between each industrys levelof external
dependence and financial development (Private Credit External
De-pendence) enters positively and significantly. This indicates
that industries that arenaturally heavy users of external finance
grow faster in economies with higher levelsof financial
development.
Moreover, column (2) shows that the interaction between each
industrys techno-logical Small Firm Share and financial development
(Private Credit Small FirmShare) enters positively and
significantly when controlling for external dependence.Thus, we
find that industries with technologically larger shares of small
firms growmore quickly in countries with higher levels of financial
development even whencontrolling for cross-industry differences in
external dependence. In unreported re-gressions, we also tested
whether the interaction between Private Credit and SmallFirm Share
varies across industries with different degrees of external
dependence.The triple interaction term does not enter significantly
and the interaction of PrivateCredit with Small Firm Share
continues to enter significantly and positively. This re-sult
suggests that small firms consistently face high financing
constraints, irrespectiveof whether they are in an industry with a
naturally high or low demand for externalfinance.
The relationship between financial development, an industrys
Small Firm Share,and industry growth is not only statistically, but
also economically large. To illustratethe effect, we compare the
growth of an industry with a relatively large share ofsmall firms
and an industry with a relatively low share of small firms across
twocountries with different levels of financial development. The
last row in Table 3 (andsubsequent tables) shows the growth
difference between industries at the 25th and75th percentiles of
the Small Firm Share and countries at the 25th and 75th
percentilesof Private Credit. Take the example of column (2). The
estimation suggests that thefurniture industry (75th percentile of
Small Firm Share) should grow 1.4% per annumfaster than the
spinning industry (25th percentile of Small Firm Share) in Canada
(75thpercentile of Private Credit) than in India (25th percentile
of Private Credit). Sincethe average growth rate in our sample is
3.4%, this is a relatively large effect.
-
1392 : MONEY, CREDIT AND BANKING
To assess the robustness of the results, we relax assumptions
concerning the dis-tribution of the error term in the estimation
equation. First, industry-specific shocksacross all countries would
invalidate the standard OLS assumption of independenterrors. Thus,
in column (3), we cluster at the industry level; i.e., we allow
error termsto be correlated within industries but not across
industries. As shown, this does notchange the results. Second,
country-specific shocks across all industries within acountry would
also invalidate the standard OLS assumption of independent
errors.Thus, column (4) presents a regression with clustering at
the country-level; i.e., weallow errors to be correlated within
countries but not across countries. While thesignificance of the
Small Firm Share-Private Credit interaction term decreases,
thecoefficient remains significant at the 7% level.7
We were also concerned that including industries that provide
very little valueadded in countries could bias the results.
Consequently, we excluded industries belowthe median share of value
added for each country. These results are presented incolumn (5) of
Table 3. With this subsample, financial development continues to
exerta particularly large impact on small-firm industries.
3.2 Controlling for Different Country and Industry
CharacteristicsIn this subsection, we control for additional
country and industry traits. If finan-
cial development simply proxies for other country
characteristics that interact withindustry firm size to shape
cross-industry growth rates, we might draw inappropriateinferences
about the independent impact of the financial system on
cross-industrygrowth rates unless we control for these other
country characteristics. Similarly, byomitting key industry traits
from the analyses, we might inappropriately interpret theresults as
relating to the natural firm size of industries rather than to
other industrytraits correlated with firm size. Thus, we control
for numerous country and industrytraits to gauge the robustness of
the findings.
Based on a large and growing literature, we control for an array
of country traits.First, we control for the interaction of Small
Firm Share with GDP per capita sincefinancial development might
simply reflect overall development, as measured by GDPper capita,
and not something particular about the financial system. If this is
the caseand overall development exerts a particularly beneficial
effect on small firms, thenwe will draw inappropriate inferences
about the impact of financial development onthe growth of
small-firm industries if we do not control for GDP per capita.8
Second,industries that depend on relatively large firms may grow
faster in economies with
7. The relationship between financial development and industry
growth is robust to controlling forreverse causality by using legal
origin and other historic and geographic characteristics of each
countryas instrumental variables, and when correcting the standard
errors for clustering at the industry or countrylevels.
8. We also included a proxy for educational attainment and its
interaction with Small Firm Share.A more educated population might
be more conducive to the growth of industries composed of
smaller(or larger) firms since technical, entrepreneurial, and
managerial skills influence industrial organizationand growth.
Adding this additional term did not change the results on the
interaction between financialdevelopment and Small Firm Share and
did not enter independently significantly.
-
THORSTEN BECK ET AL. : 1393
larger markets that allow them to exploit economies of scale
more fully (Braun andRaddatz 2008). To test this, we include a
proxy for market size: openness to interna-tional trade, which is
measured as exports plus imports divided by GDP. Furthermore,using
the size of the economy (GDP) as a proxy for market size rather
than the tradeyields the same results. Third, financial market
frictions might be highly correlatedwith regulatory impediments to
labor mobility and new firm formation. If this is thecase, we might
inappropriately interpret the results as applying to finance when
theyreally apply to other frictions. For instance, Klapper, Laeven,
and Rajan (2006) findthat new firms are disproportionately hurt by
regulatory impediments to labor mo-bility and high entry barriers.
We therefore control for an interaction of Small FirmShare with
Entry Regulation, which is the cost of registering an enterprise
relativeto GDP in 1999 (Djankov et al. 2002). Table 2 provides
summary statistics on thesecountry indicators.
Since a parallel literature examines how different industry
traits affect cross-industry growth rates across countries, we also
use this research to inform our robust-ness tests. First, if (i)
financial development has a disproportionately positive effecton
industries with good growth opportunities (Fisman and Love 2007)
and (ii) small-firm industries just happened to enjoy good growth
opportunities over the sampleperiod, then we might erroneously
infer that financial development exerts an espe-cially positive
impact on small firms. We therefore control for Sales Growth,
whichis calculated as real annual growth in net sales of U.S. firms
over the period 1980 to1989 using data from Compustat. Second, if
(i) small firms rely heavily on intangibleassets and (ii) strong
private property rights are closely associated with financial
de-velopment, then our findings may simply be confirming the
results in Claessens andLaeven (2003). We therefore control for the
interaction of Property Rights with thepercentage of intangible
assets in each industry, computed as the ratio of intangibleassets
to fixed assets of U.S. firms over the period 1980 to 1989 using
data fromCompustat. Third, differences in informational asymmetries
might account for finan-cial developments disproportionate
influence on small-firm industries. To test this,we use two
measures of the informational opacity of industries. First, Rating
Splitsmeasures disagreement between the two major bond rating
agenciesMoodys andS&Pabout the risk of U.S. firms, based on the
bond ratings of almost 8,000 firmsduring the period 198393 (Morgan
2002). Greater disagreement suggests greateropacity. The second
measure of informational opacity comes from Durnev, Morck,and Yeung
(2004), who compute the degree to which individual stock prices
movewith average stock prices in an industry based on an R2 measure
of synchronic-ity, with higher R2greater synchronicityas an
indication that investors have amore difficult time discerning
firm-specific differences. Fourth, the Small Firm Sharemight simply
proxy for the degree of industry concentration and we therefore
con-trol for the interaction of Private Credit with the four-firm
concentration ratio basedon U.S. Census data. Table 2 provides
summary statistics and partial correlationsamong these industry
characteristics. Small Firm Share is significantly correlatedonly
with the share of intangible assets (positively) and with industry
concentration(negatively).
-
1394 : MONEY, CREDIT AND BANKING
After controlling for all of these country and industry
characteristics, we con-tinue to find that financial development
disproportionately boosts the growth rate ofsmall-firm industries.
The interaction of Private Credit with Small Firm Share
enterspositively and significantly at the 1% level in column (6) of
Table 3. The interac-tion between Private Credit and External
dependence also enters significantly.9 Onlytwo other interaction
terms enter significantly. The interaction between Small FirmShare
and Entry Regulation enters negatively and significantly. This
suggests thatregulations that impede entry are particularly harmful
to industries that are naturallycomposed of small firms for
technological reasons. Furthermore, the interaction be-tween
Intangibility and Property Rights enters positively and
significantly, indicatingthat industries that are naturally
characterized by a high proportion of intangible as-sets grow
relatively faster in countries with comparatively well-functioning
propertyrights systems.
These results indicate that (i) Small Firm Share does not only
reflect other industrycharacteristics and (ii) Private Credit does
not simply reflect other national traits.Rather, we find an
independent relationship between financial development and
therelative growth rates of industries that are naturally composed
of smaller firms fortechnological reasons. The robustness of Small
Firm Share indirectly suggests thatfinancial development operates
at the extensive margin by allowing new small firmsto access
growth-enhancing financial services.
3.3 Alternative Definitions of a Small FirmTable 4 indicates
that the results are robust to using alternative definitions of
a
small firm below 20 employees. We use four different cutoffs to
define a small firm:5, 10, and 100 employees, respectively.10 Table
1 lists Small Firm Share for thedifferent definitions of a small
firm, where Sx in the table indicates the industrysshare of
employment by firms with less than x workers. There is a high
correlationamong the different measures of Small Firm Share, and
the average correlation is 91%(Table 2, Panel D). Not surprisingly,
the correlation decreases with higher thresholdmeasures of firm
size. The correlation between S5 and S10 is 99%, but 87% betweenS5
and S100. Nevertheless, using different cutoffs provides additional
robustnesstests and more fully characterizes the relationship
between cross-industry firm size,
9. As shown, the size of the coefficient on the interaction
between Small Firm Share and PrivateCredit does not change much
when including all of the industry and country control variables.
Rather thanincluding all of the control variables simultaneously as
reported in regression (6), we also included themone at a time.
When we only include the two interaction terms of Private Credit
with External dependenceand Entry barriers with Small Firm Share,
we drive the coefficient on the interaction between Private
Creditand Small Firm Share from 0.54 in regression 6 to 0.27, but
it remains significant at the 5% level. Theabsolute value of the
coefficient on the interaction term between Entry barriers and
Small Firm Share alsofalls, from 0.76 in regression (6) to 0.63.
While this suggests a relation between entry barriers andfinancial
development, the regression results demonstrate that both financial
impediments on firms andnonfinancial barriers to firm entry have
independent, negative effects on small-firm industries.
10. Two industries (Manufacture of paper and paper products, and
Manufacture of pulp, paper, andpaperboard) drop from the sample due
to missing U.S. Census data when using 5 or 10 employees as
thecutoff.
-
THORSTEN BECK ET AL. : 1395
TABL
E4
ALT
ERN
ATI
VE
MEA
SURE
SO
FFI
RM
SIZE
DIS
TRIB
UTI
ON
(1)(2)
(3)(4)
(5)(6)
(7)(8)
Indu
stry
shar
ein
val
uead
ded
1.1
59
1.1
64
1.1
48
1.1
28
1.1
06
1.1
73
1.1
06
1.1
27
(0.29
2)(0.
293)
(0.28
2)(0.
281)
(0.27
9)(0.
303)
(0.27
4)(0.
278)
Priv
ate
Cred
it
Exte
rnal
finan
cial
depe
nden
ce0.
174
0.
174
0.
166
0.
164
0.
161
0.
182
0.
152
0.
159
(0.
044)
(0.04
4)(0.
043)
(0.04
3)(0.
043)
(0.04
7)(0.
042)
(0.04
3)Pr
ivat
eCr
edit
Sm
allfi
rmsh
are
(empl