Transcript Firm Size, Finance and Growth - Faculty Directory | Berkeley-Haas

Firm Size, Finance and Growth
Thorsten Beck
Asli Demirguc-Kunt
Luc Laeven
Ross Levine
Motivation

Does finance have distributional effects?

Income distribution / poverty (BDL, 2005)

Small (poor) firms do not access financial system,
so finance benefits large (rich) firms more


(Greenwood / Jovanovic, 1990)
Financial development lowers fixed costs
(transaction & information), so helps small (poor)

(Banerjee / Newman, 1993; Galor / Zeira, 1993)

(How) does finance affect growth?

Policy: (a) Political economy and (b) SMEs
This paper’s goals …

Does financial development boost
the growth of small firms more than
large firms?



Distributional effects
Mechanisms through which finance
affects growth
Policy
Methodological Strategy

Do “small-firm” industries grow faster in countries
with well-developed financial systems?
 Coase (1937)




Firms optimally internalize some activities, but size
enhances coordination costs
Industry’s “natural” firm size depends on that
industry’s production technologies
Step 1: Compute each industry’s natural firm size:
Share of employment in “small firms.”
Step 2: Test whether industries that are naturally
composed of small firms, grow faster in countries
with well-developed financial systems.
More on the methodology …

We use the U.S. as the benchmark to
compute each industry’s natural firm size

Industry Firm Size = F{Industry & Country}
Assume USA has comparatively few distortions
 Then, role of country traits is small.



Obtain proxy for industry’s natural firm size
(Similar to RZ, who compute industry’s
natural tendency to use external finance.)
Related literature

Guiso, et al: Small firms benefit more from
regional financial development in Italy



Nice.
But, we focus across countries
Beck, et al (2005): reported financial
obstacles to growth is stronger in small
firms in under-developed financial systems


Nice.
But, based on survey responses
Data
1.
2.
3.
Industry growth
Small firm share
Financial development
Industry growth


Average annual growth rate of real
value added of industry k in country
i over the period 1980-1990.
We show the results hold over
different sample periods.
Small firm share


Industry k’s share of employment in firms
with less than 20 employees in the U.S.
(1992 Census, earliest date possible)
Robustness



Different firm size cut-offs (5: 500)
(1997 Census … correlation of 92%)
Concerns about U.S.:
 Control for other factors that may invalidate
the US as a benchmark.
U.S. markets do not have to be perfect. They
have to give a reasonable ranking.
 Different benchmark countries

Table 1: Firm size across U.S. industries
(A few, select observations)
ISIC
3411
314
353
311
342
390
331
Average
Industry name
Manufacture of pulp, paper and paperboard
Tobacco manufactures
Petroleum refineries
Food manufacturing
Printing, publishing and allied industries
Other Manufacturing Industries
Manufacture of wood and wood and cork products, except
furniture
S20
0.14
0.30
0.36
3.82
16.32
16.95
21.37
5.85
Financial development


Private credit
Others



Liquid liabilities
Stock market development
Legal & accounting systems
Methodology
Growthi ,k    i Countryi    k Industry k   Sharei ,k 
i
k
 ( Small Firm Share k * FDi )   i ,k ,
Growth = average annual growth of real value added of industry k in
country i, averaged over 1980-90
Share = Initial share of industry i in 1980 in total manufacturing
FD = Claims of financial institutions on private sector relative to GDP
in country i.
Small firm share = benchmark share of small firms in industry k
OLS and IV, also cluster at industry or country level
Sample: 36 industries across 44 countries
Table 3: Financial development, small firm
share and growth
Share in value added
Private Credit * Small firm share
OLS
-1.012***
(0.253)
0.409**
(0.172)
OLS
-1.095***
(0.253)
0.445**
(0.173)
0.144***
(0.039)
IV
-1.086***
(0.253)
0.567**
(0.220)
0.101***
(0.037)
1242
0.26
1242
0.28
1242
0.27
Private Credit * External dependence
Observations
R-squared
Financial development, small firm share and
growth - economic significance

Small Firm Share:



Private Credit:




25th percentile: Spinning
75th percentile: Furniture (lots of small firms)
25th percentile: India
75th percentile: Canada
Furniture grows 1.4% faster than
spinning in Canada than in India
Average growth rate = 3.4%
But, …

Small firm share in the U.S. may be


correlated with other industry-specific
traits that
interact with country-level
characteristics to explain industry
growth
Robustness: industry traits …

Is small firms share a proxy for …


External dependence?
Intangible assets?




Claessens and Laeven show this with property rights
protection
But, we interact it with both property rights and
private credit
Good, or bad, growth prospects?
Technology factors that  firm size in U.S.?

Control for median firm size of the large, listed
firms by industry in the U.S. (in a few slides)
Robustness: country traits …

Is financial development a proxy for …


Economic development?
Schooling?


Human capital may affect natural firm size
Size of the market?
Openness to trade
 Size of the economy

Conclusions

Finance has distributional effects




Small firm industries grow faster (than big firm
industries) with better financial development
{BDL: the poor enjoy faster income growth (than the
rich) with better financial development.}
Mechanism linking finance and growth:
Alleviates constraint on small firm growth
Policy: Political economy & SMEs