The Quarterly Review of Economics and Finance 65 (2017) 1624

Contents lists available at ScienceDirect

The Quarterly Review of Economics and Finance

journal homepage: www.elsevier.com/locate/qref

Bank lending to small business in India: Analyzing productivity

and efciency
Harri Ramcharran
Finance and International Business, Department of Finance College of Business, University of Akron, Akron, OH 44325, USA

a r t i c l e i n f o a b s t r a c t

Article history: This study extends the literature of the performance of small and medium size industries (SME) by
Received 19 October 2015 empirically estimating the efciency of bank loans to the SME sector of India using data from 1979 to
Received in revised form 8 May 2016 2013. We use a production function methodology, specically estimating a non-homogeneous production
Accepted 3 June 2016
function which, unlike other specications, provides time varying econometric estimates of productivity,
Available online 27 June 2016
this helps to detect the impact of policy changes on efciency overtime. The theoretical rationale for this
approach is the money as an input in the production function argument. The results have important
JEL classication:
policy implications regarding capital allocation and risk management since bank loans are the main
C22
C52
source of nancing SME. The results indicate increasing productivity (output elasticity) of bank credit
E50 from 0.76 to 1.23; the productivity of labor is negative, but increases from 1.57 to 0.628. The sectors
G21 efciency improved from returns to scale of 0.89 to 0.607 mainly due to the increasing productivity of
bank credit. The decrease in the number of sick units with large outstanding debt apparently contributes
Keywords: to this improvement. Current problems facing the banking sector could increase borrowing cost, decrease
Bank loan
credit availability and thus the productivity of SME. Policies to improve the skills of workers and to reduce
Small scale industry
the number of sick units are important for improving efciency.
Production efciency
Returns to scale 2016 Board of Trustees of the University of Illinois. Published by Elsevier Inc. All rights reserved.

1. Introduction Several researchers have addressed the efciency aspect of SME,

The growing signicance of small and medium size enter-
prises (SME) for employment and income generation has led
various governments and international organizations to support
their development and expansion. The World Bank (2010) provides
data to illustrate the growth of micro, small, and medium size enter-
prises (MSME) worldwide. There is a growing literature (Ayyagari,
Beck, & Demirguc-Kunt, 2007; Schiffer & Weder, 2011; Tybout,
2000) on different aspects of SME operation and performance.
Several studies focus on the pattern of nancing, for example,
Ayyagari, Demirguc-Kunt, and Maksimovic (2012), Calice, Chando,
and Sekioua (2012), Beck, Demirguc-Kunt, and Maksimovic (2008),
Beck, Demirgu-Kunt, Laeven, and Levine (2008), Beck, Demirguc-
Kunt, and Martinez Peria (2011), Tybout (2000) and Clarke, Cull,
Peria, and Sanchez (2005). Tybout (2000) notes that in studies of
microenterprises, scale economies based on estimated production
functions are consistently missing; this research contributes signif-
icantly to an under-investigated area of the literature.
Tel.: +1 330 972 6882; fax: +1 330 972 5970.
E-mail address: ramchar@uakron.edu
they contend that the performance of SMEs has to be efciency
driven in order to achieve economic growth, employment creation,
and poverty reduction. Mead and Liedholm (1998) articulate the
necessity of containing production costs and of achieving economic
efciency to foster the growth of SME, while Dewatripont and
Maskin (1995) emphasize credit efciency as a goal for lending to
SME. Tybout (2000) lists several factors that may prevent scale ef-
ciency in SME; some are (i) surplus unskilled labor and a lack of long
term nancing, (ii) poor infrastructure, including communication
facilities and transportation network, and (iii) volatility in business
environment which discourages mass production techniques. Sev-
eral authors (Ayyagari et al., 2012; Beck et al., 2011) document the
growing importance of bank nancing for SME around the world.
Beck et al. (2011) also emphasize the necessity of research of the
supply side of bank nancing. The efcient use of bank loans has
not been investigated despite various reports of their signicant
role in SME nancing; the ndings of such research have important
policy implications related to efcient capital allocation, manag-
ing risk at the rm level and policies to improve productivity and
performance.
This research, using a methodology derived from recent
advances (Fried, Knox Lovell, & Schmidt, 2008) in production eco-
nomics, estimates the efciency of bank loans to the SME sector of

http://dx.doi.org/10.1016/j.qref.2016.06.003
1062-9769/ 2016 Board of Trustees of the University of Illinois. Published by Elsevier Inc. All rights reserved.
 H. Ramcharran / The Quarterly Review of Economics and Finance 65 (2017) 1624 17

India using data from 19781979 to 20122013. We use a non- Table 1

Descriptive statistics.
homogeneous production function (NHPF), developed by Vinod
(1972), analyzed by Bairam (1997) and Intrilligator (1978), in which Q L CR
bank loans to the SME sector and the level of employment are used Mean 1,611,624. 17.33486 35,131.49
as inputs. . . The rationale for using labor is the labor intensity of Median 1,181,000. 15.83000 17,938.00
the production process and also the government priority to protect Maximum 5,329,790. 62.63000 213,539.0
labor employment through rigid labor laws, e.g. making retrench- Minimum 342,000.0 3.970000 904.0000
Std. dev. 1,284,048. 13.48860 44,537.94
ment difcult during economic downturn. Two streams of studies
Skewness 1.464307 1.887274 2.207891
provide the rationale for the methodology. First, the issue of the ef- Kurtosis 4.224861 6.978293 8.683047
ciency of nancial institutions (commercial banks, saving and loans,
JarqueBera 14.69572 43.85797 75.53604
credit unions, and insurances rms) is well-documented in sev- Probability 0.000644 0.000000 0.000000
eral studies (Berger & Humphery, 1997; Berger, Hunter, & Timme,
Sum 56,406,830 606.7200 1,229,602.
1993). Recent studies have used different estimation techniques
Sum sq. dev. 5.61E+13 6186.036 6.74E+10
including parametric frontier analysis with different specications
of cost, prot/revenue and production functions. Second, several Observations 35 35 35

theoretical and empirical models, incorporating the role of real

money balances (bank loans, nancial assets) as a factor of pro-
duction, have been advanced in various studies, e.g. Laumas and
Mohabbat (1980), Finnerty (1980), Sinai and Stokes (1981), Khan
and Ahmad (1984), and Hasan and Mahmud (1993).
The SME sector of India provides an interesting case study for
examining the efciency of bank lending. This sector has grown
signicantly in terms of increasing value added, the level of employ-
ment, the number of units and loans received from the government.
From 1993 to 2005, production growth from this sector averages
about 8.38% annually; this is higher than the 5.8% from the over-
all (economy wide) industrial sector. The sector relies heavily on
external nancing mainly from the central government. Lending to
this sector is mandated by the Reserve Bank of India Priority Sector
Lending Program that requires every commercial bank to allocate
40% of these loans to what is dened as priority sectors (SME and
Agriculture). Preferential treatment is also provided to this sector
in the form of low interest rate on bank credit.1
The NHPF specication has several advantages, unlike linear
homogenous production functions (Cobb-Douglas and Constant
Elasticity of Substitution) which assume constant productivity
estimate at all levels of output, it provides parameter estimates
of efciency that vary with output and factor (input) propor-
tions. Time-varying parameters enable an examination of the
pattern of productivity/efciency changes over the period of study,
as well as making inferences about policies implemented to
impact the performance of this sector, for example, the economic
reform/liberalization policies during the 1990s.2 The efciency
parameters estimated and analyzed are: (i) the output elasticity of
labor, (ii) the output elasticity of bank loans, and (iii) returns to scale
(RTS). The model specication and estimation technique are fully
described in another section of the paper. In addition to regression
analysis, we perform several diagnostic tests to ensure the struc-
tural stability of the model and the reliability of the results; the tests
1
Kamesam (2003) lists some of the various measures implemented by the GOI
for the development of SME, they include: (i) product reservations, (ii) scal conces-
sions, (iii) preferential allocations of credit and interest subsidy in a credit rationing
framework, (iv) extension of business and technical services, (v) preferential pro-
curement by the government, and (vi) the provision of lower interest rates as well
as requirement for a minimum credit allocation from the commercial banks.
2
An important aspect of the reform program includes policies targeting the per-
formance of the SME sector: (a) industrial policy that remove barriers to industrial
licensing to enable the creation of new investment and expand existing capacity, (b)
labor market reforms that gradually remove some of the restrictive labor laws by
creating a safety net (National Renewal Fund), to deal with problems of displaced
workers, and (c) nancial sector (banking) reform to enable the nancial sector to
mobilize savings and allocate resources toward more efcient use in restructuring
the real economy. All these three polices have resulted in a vast increase in the num-
ber of units of enterprise, the level of employment and funds (bank loans) available
to this sector (see Ahluwalia, 2002, 1994).
are for (i) the stationarity of the data, and (ii) the cointegration of
the variables.
The results have practical policy implications for both govern-
ment and business. Signs of weakness in the economy and banking
sector could result in a credit crunch and increasing borrowing cost
which could impact the performance of this sector. Despite some
impressive performance data, a major problem with this sector is
the large number of sick units, dened as units with outstand-
ing accounts/debt remained overdue for a period of more than 2.5
years. Recent data indicate that about 10 percent of the units are
classied as sick. Previous studies (Bhavani, 1991; Ramaswammy,
1994) of Indias SME sector use the stochastic production fron-
tier technique to estimate RTS in different industries; they have
not examined issues related to the productivity of bank credit
(the countrys scarce resource) and of labor (the abundant factor).
Moreover, performance resulting from recent changes in economic
policies targeting this sector should be addressed. As a key part
of Indias manufacturing sector, its performance requires a funda-
mental understanding of the productive/efciency characteristics
regarding factor productivity and RTS since the Government con-
tinues to commit resources to expand it.
The rest of the paper is structured as follows: (i) an overview of
SSI in India, (ii) nancing small businesses in developing countries,
relevant literature, (iii) literature on money as an input in the pro-
duction function, (iv) model specication and data, (v) discussion
of results, and (vi) conclusion.
2. An overview of the SME sector of India
The Government of India (GOI) has always recognized that
SME will contribute to the economic progress of the country with
its labor intensive (capital scarce) factor endowment. The argu-
ments for its expansion are: (i) small rms would create additional
employment opportunities, (ii) small rms are capable of produc-
ing large quantities of consumer goods and increasing demand and
income would generate new investment in heavy basic industries,
(iii) SME could produce a wide range of products with technology
varying from traditional to state of art.3 In India, the denition of
SME is in terms of cumulative investment in plant and machin-
ery; Table 1A shows the changing sizes in terms of xed asset
3
Kumar (www.preservearticles.com) also notes the following important roles
played by SME: (i) mobilization of resources and entrepreneurial skill, (ii) promot-
ing equitable distribution of income, (iii) enabling regional dispersal of industries,
(iv) providing opportunities for the development of technology, (v) indigenization
of production, (vi) export promotion, (vii) supports the growth of large industries
through linkages, and (viii) better industrial relations and social inclusion.
18 H. Ramcharran / The Quarterly Review of Economics and Finance 65 (2017) 1624

Table 2
Results of ADF test of unit roots.

Variable Test in Included in test Coefcient t (tau) value Prob. Decisiona

LQ 1st difference Constant 0.9639 5.367 0.0000 Reject Ho

Constant and trend 0.966 5.292 0.0000 Reject Ho
None 0.783 4.5156 0.0000 Reject Ho

LL 1st difference Constant 1.066 5.9729 0.0000 Reject Ho

Constant and trend 1.087 5.963 0.0000 Reject Ho
None 0.719 4.2333 0.0000 Reject Ho

LCR 1st difference Constant 0.609 2.699 0.1012 Reject Ho

Constant and trend 0.597 2.4311 0.0212 Do not reject Ho
None 0.056 0.5182 0.6079 Do not reject Ho
a
Ho: unit root exists. Decision is based on the Augmented DickyFuller test statistic, MacKinnon (1996).

Fig. 1. Number of units of SSI (million). Fig. 3. Employment (million).

(GDP), and (v) employment of 19.3 million persons. Government

Fig. 2. Output from SSI (crore ).
over past years.4 The sector mainly relies on bank nance for
its operation, thus adequate ow of credits has been an overrid-
ing policy objective of the government. The growth of this sector
has been spectacular; from 1979 to 2013, (i) the number of units
of SME increased from 0.73 million to 46.75 million, see Fig. 1,
(ii) production increased from 58,200 crore (582,000 million)
to 1,809,976 crore (18,099,760 million) at constant price, see
Fig. 2, (iii) employment increases from 6.38 million to 106.14 mil-
lion, see Fig. 3, (iv) export from this sector increased from 11
billion crore (11,000 million) or $1303 million to 6973.18 bil-
lion (6.97318 million) or $128,162 million.5 Recent data from
Reserve Bank of India (2006) attest to the economic importance of
this sector, it accounts for: (i) 95% of industrial units in the coun-
try, (ii) 39.92% of value added in the manufacturing sector, (iii)
34.29% of national exports, (iv) 6.86% of Gross Domestic Product
policy to increase the size of this sector involves the passage of
the Micro, Small, and Medium Enterprises Development (MSMED)
Act of 2006; with the passage of this Act, the aggregate number of
units increased from 12.34 million in 2006 to 46.75 million in 2013.
Table 2A highlights some of the provisions. The economic liberaliza-
tion process that began in 1991 aimed at deregulating the economy
to achieve greater competition in the domestic markets and open-
ness to international trade and investments also included specic
policies that target the development of the SME sector (Ahluwalia,
1994). Foremost was the removal of industrial licensing to create
new investments; a controversial aspect of the delicensing policy
is the restriction of large rms from producing selecting manufac-
tured products reserved for SME.
The sector mainly relies on bank nance for its operation, thus
adequate ow of credits has been an overriding policy objec-
tive of the government. Beginning 1969, within nationalization
of the banking industry, banking policy mandated that 40% of
banks credit should be allocated to priority sectors which include
SME and agriculture. Preferential treatment includes low interest
rates. The Small Industries Development Bank (SIDB) was estab-
lished in 1990 as the premier nancial institutions for promoting
and developing of this sector. Bank loans (advances) to the SME
sector increase steadily from 22.32 billion in 1979 to 6872 bil-
lion in 2013 (see Fig. 4) Delinquent loans (loans outstanding for
more than 2.5 years) constitute a major problem affecting this
sector; however, the number of units has been decreasing since
2001. Fig. 5 shows the pattern of sick units with outstanding
loans.
Table 3A provides a summary of recent targets for priority sector
lending by commercial banks.6 Banks have been advised to achieve

4
The denition of SME is by the value of xed assets; in other countries it is
6
dened by the level of employment or by sales (World Bank, 2010). See Master Circulation Lending to Priority Sector, Reserve Bank of India, July
5
The data are from Handbook of Statistics of India Economy (RBI, 20102011). 2012 and Chakrabarty (2012).
 H. Ramcharran / The Quarterly Review of Economics and Finance 65 (2017) 1624 19

Table 3
Johansen cointegration, test trend assumption: no deterministic trend, series: LL LQ LCR.

Lags interval (in rst differences): 1 to 1

Unrestricted cointegration rank test (trace)

Hypothesized no. of CE(s) Eigenvalue Trace statistic 0.05 critical value Prob.**
*
None 0.419483 20.56092 24.27596 0.1371
At most 1 0.070088 2.614336 12.32090 0.8931
At most 2 0.006536 0.216396 4.129906 0.6987
Trace test indicates no cointegration at the 0.05 level

Unrestricted cointegration rank test (maximum eigenvalue)

Hypothesized no. of CE(s) Eigenvalue Max-eigen statistic 0.05 critical value Prob.**

None* 0.419483 17.94658 17.79730 0.0475

At most 1 0.070088 2.397940 11.22480 0.8748
At most 2 0.006536 0.216396 4.129906 0.6987

Max-eigenvalue test indicates 1 cointegrating eqn(s) at the 0.05 level.

*
Rejection of the hypothesis at the 0.05 level.
**
MacKinnonHaugMichelis (1999) p-values.

3. Financing small business, review of the literature

A growing literature on the nancing of SME includes a semi-

nal paper by Ayyagari et al. (2012) who examine nancing pattern
of 99 countries based on institutional and economic factors; they
analyze the pattern of nancing based on the following stylized
facts about rms: (i) concentration of ownership, (ii) capital struc-
ture choice, (iii) bank versus market sourcing, (iv) access to foreign
capital, (v) cross border mergers, (vi) productivity, (vii) industry
structure and entrepreneurship, (viii) the role of small rms, and
(ix) informality. The important ndings are (i) debt nancing (bank
loans) is the major source of external nancing, (ii) foreign bank

Fig. 4. Credit of SSI (billion ). Table 4

Estimates of elasticity and RTS.

Period EL ECR RTS

a 20% annual growth to SMEs; the allocation of 60 percent of the
19781979 1.5701 0.7605 0.8096
advances is to be achieved in three stages: 50 percent in the year
19791980 1.5460 0.7636 0.7824
20102011, 55% in the year 20112012, and 60 percent in the year 19801981 1.5282 0.7856 0.7427
20122013. Over 70% of loans to the SME sector are from public 19811982 1.4944 0.7996 0.6948
sector banks, next of importance are private sector banks, while 19821983 1.4609 0.8136 0.6473
foreign banks provide less than 5%. Table 4A provides data on bank 19831984 1.4068 0.8423 0.5645
19841985 1.3819 0.8507 0.5311
loans from 2008 to 2011. Kamesam (2003) identies other prob-
19851986 1.3159 0.8607 0.4551
lems faced by this sector, some include the following: (i) a limit for 19861987 1.2949 0.8702 0.4247
collateral free loans, many SME entrepreneurs are facing difculties 19871988 1.2625 0.8791 0.3833
in providing collateral security (ii) high borrowing cost for loans, 19881989 1.2292 0.8901 0.3390
19891990 1.1981 0.9016 0.2965
and (iii) considerable delay in the settlement of dues/payment of
19901991 1.1698 0.9127 0.2570
bills, and (iv) marketing, small SSI have to sell output individually. 19911992 1.1449 0.9222 0.2228
A number of measures have been taken to address these problems 19921993 1.1111 0.9314 0.1797
to promote sustainable growth of this sector. 19931994 1.0903 0.9408 0.1494
19941995 1.0752 0.9506 0.1246
19951996 1.0552 0.9989 0.0563
19961997 1.0458 1.0071 0.0387
19971998 1.0282 1.0160 0.0122
19981999 1.0052 1.0235 0.0184
19992000 0.9714 1.0316 0.0603
20002001 0.9438 1.0374 0.0936
20012002 0.9249 1.0442 0.1193
20022003 0.8938 1.0502 0.1565
20032004 0.8728 1.0561 0.1833
20042005 0.8558 1.0626 0.2068
20052006 0.8467 1.0713 0.2246
20062007 0.8278 1.0793 0.2514
20072008 0.8341 1.0869 0.2528
20082009 0.8176 1.0943 0.2767
20092010 0.7901 1.0988 0.3087
20102011 0.7569 1.1061 0.3493
20112012 0.7174 1.2270 0.5096
20122013 0.6283 1.2360 0.6076
Fig. 5. Number of sick units .
20 H. Ramcharran / The Quarterly Review of Economics and Finance 65 (2017) 1624
entry has potential of increasing lending, (iii) labor productivity is
low because of a mix of nancial and organization factors (poor
access to nance and poor management), and (iv) informal rms
(unregistered) account for up to half of all economic activity in
developing countries; also informal nancing channels play any
important loan in facilitating access to nance. Calice et al. (2012)
use some of these facts to analyze African countries. With the pres-
ence of foreign banks in developing countries because of nancial
liberalization, Clarke et al. (2005) nd that foreign banks play a sig-
nicant role in lending to rms in some Latin America countries.
Beck, Demirguc-Kunt, and Maksimovic (2005), Beck, Demirguc-
Kunt, and Maksimovic (2008), Beck, Demirgu-Kunt, Laeven, et al.
(2008) and De la Torre, Martinez Peria, and Schmukler (2010)
nd that the sources (and pricing) of lending to SMEs in devel-
oping countries include large, small, private, government-owned,
and foreign banks. This is beyond relationship lending, the type
of nancing based on soft information generated by loan of-
cers through direct and personalized contacts with owners of
SMEs. Beck, Demirguc-Kunt, and Maksimovic (2008) and Beck,
Demirgu-Kunt, Laeven, et al. (2008) also nd that size matters,
small rms in countries with poor institutions use less external
nance. A recent study by Beck et al. (2011) indicates the follow-
ing variables signicant in bank nancing of SME, (i) ownership
types, (ii) foreign banks, (iii) domestic banks, (iv) different lend-
ing technology, (v) organization structure. Tybout (2000) lists other
cultural/institutional factors affecting the nancing of small rms,
some are (i) policies tends to favor large rms because they are
low risk and cheaper to service, (ii) private sector credit is rela-
tively scarce, (iii) information networking are poorly developed,
(iv) binding interest rate contracts are very common, (v) some small
producers operate partly or wholly outside the realm of govern-
ment regulation and rely heavily on the informal credit market.
Several studies investigate the impact of bank loans on eco-
nomic growth. King and Levine (1993a), King and Levine (1993b)
and Levine (1997) articulate a positive relationship between nan-
cial development and economic growth. Rajan and Zingales (1998)
provide evidence that rms/industries with external nancing
grow faster. Fishman and Love (2007) provide modied technique
to investigate this issue. Levine and Zervos (1998) examine the
impact of nancing development on economic growth using the
amount of bank loans available to rms as one of the indicators of
nancial development.
4. Money as an input in the production function: review of
the literature.
The production approach methodology to examine the impact
of real balances (loans and credit) on output was stimulated by
Friedmans (1969) theory of the optimal amount of money stating
that money should be treated as a productive input similar to cap-
ital and labor in explaining the behavior of rms. Finnerty (1980)
provides several arguments why real money balances should be
included in the rms production function, he contends (p. 671), we
have also shown that, in general, the rms expansion path will not
be identical with or without cash balances as a factor input, which
suggests that including cash balances in the production function
may be the more appropriate formulation. Laumas and Mohabbat
(1980) note that if real balances are used as an input, then it must
have productivity characteristics (e.g. marginal and average pro-
ductivity) similar to physical inputs. They use a three factor (capital,
labor, and real money balances) aggregate Cobb-Douglas (CD) pro-
duction function to estimate factor output elasticity for the French
economy; real balances are dened as currency outside banks plus
demand deposit, the results show the signicance of the output
elasticity of real balances. Sinai and Stokes (1981) suggest that the
CD specication may have errors; instead, they test a three fac-
tor Constant Elasticity of Substitution (CES) model with different
measures (M1 , M2 , and nancial assets held by non-nancial cor-
porations) of real balances. Khan and Ahmad (1984) applied a three
input (capital, labor, and real money balances) CD production tech-
nique to large scale manufacturing sector of Pakistan. They use a
multi equation framework and nd the coefcient of real money
balance statistically signicant. Hasan and Mahmud (1993) use a
translog cost function with the following independent variables,
capital, skilled labor, unskilled labor, and real balances (measured
by cash and marketable securities). They estimate the following
elasticities (i) own price, (ii) cross price, and (iii) substitution; all
the elasticities are statistically signicant.
Falls and Natke (1988) delineate the demand approach to ana-
lyze scale efciency; it emphasizes the rms transaction demand
for liquid assets (money). Their model includes liquid assets as the
dependent variables and the following independent variables (i)
a scale factor, measured by sales, (ii) ownership, indicated by for-
eign or domestic and (iii) different industry classications. Falls and
Natke (2010) extend their model to examine the difference in scale
elasticity in money holding between small rms and large US cor-
porations; scale is measured by sales or assets. The results show
small rms having higher scale elasticity.
5. Methodology
The literature on estimating production efciency (Bairam,
1997; Chung, 1994; Fried et al., 2008; Intrilligator, 1978), lists
several techniques. Two common functional forms used in early
studies are: (i) the Cobb-Douglas (C-D) and (ii) the Constant
Elasticity of Substitution (CES). Their specications contain sev-
eral problems; one of which is the restriction on the value of
the parameter measuring RTS. Both models assume that RTS is
xed (constant) at all levels of output7 ; later studies recognize
the possibility of variable RTS with changing input (capital/labor)
ratio. Revankar (1971) develops a variable elasticity of substitu-
tion (VES) specication in which factor productivity varies with
the input ratios. Christensen, Jorgenson, and Lau (1973) articulate
the importance of developing empirical tests of production theory
employing the non-constant scale elasticity (non-homogeneity);
they develop the transcendental logarithmic (translog) production
function which has been applied by others, including Kim (1992).8
The NHPF model that we apply is a special case of a log-quadratic
production function used by Vinod (1972) to estimate factor inten-
sity and returns to scale. The virtues of this model are: (i) no
restriction is imposed on the data, (ii) the specication is exible,
and (iii) it estimates the properties of production, for example, out-
put elasticity and return to scale with different factor proportion,
these estimates vary over the period of study and (v) the model is
linear in its parameters and can be estimated by OLS.
This specication enables the estimation of productivity param-
eters that vary overtime with the level of output and factor
proportion, this helps to analyze the variation in efciency and to
relate it to change in specic policies. The data on the SME of India
7
This results in the long run average total cost curve (LRATC) being either constantly
rising (decreasing return to scale), horizontal (constant returns to scale), or constantly
falling (increasing returns to scale).
8
A main problem with the translog production function is the possibility of multi-
collinearity arising from using the squared value of each input and the product of
two inputs as independent variables. For a two-input case, Q = f (X1 , X2 ), the translog
specication is
2 2
ln Q = a0 + a1 ln X1 + a2 ln X2 + a3 (ln X1 ) + a4 (ln X2 ) + a5 (ln X1 ln X2 )
If a3 , a4 , and a5 are not statistically signicant, the CD specication is appropriate
(see Intrilligator (1978, Chap. 8).
 H. Ramcharran / The Quarterly Review of Economics and Finance 65 (2017) 1624 21

show signicant variation in output, amount of bank loans, num- Table 1A

Denition of SSI Since 1950.
ber of units of SME, and the level of employment overtime, thus
a exible model specication is appropriate to estimate produc- Year Investment limits Additional conditions
tion efciency. Additionally, there have been various public policies 1950 Upto Rs. 5.0 lakh in xed assets Less than 50/100 persons with
enacted by the government to enhance the performance of this sec- or without power
tor. Ramcharran (2001, 2011, 2012) has applied this methodology 1960 Upto Rs. 5.0 lakh in xed assets No condition
to analyze production efciency in different industries. 1966 Upto Rs. 7.5 lakh in plant and No condition
machinery
The model specication is:
1975 Upto Rs. 10 lakh in plant and No condition
Q = ea0 CRa1 +a3 ln L a2
L . (1) machinery
1980 Upto Rs. 20 lakh in plant and No condition
The NHPF model includes multiplicative input combination to machinery
1985 Upto Rs. 35 lakh in plant and No condition
assess their joint contribution to productivity. Since Eq. (1) is mul-
machinery
tiplicative, it can be written in double logarithmic format as 1991 Upto Rs. 60 lakh in plant and No condition
machinery
ln Qt = a0 + a1 ln CRt + a2 ln Lt + a3 (ln CR ln L)t , (2) 1997 Upto Rs. 300 lakh in plant and No condition
machinery
where Q is the production at constant prices (millions of rupees),
1999 Upto Rs. 100 lakh in plant and No condition
CR is the amount of bank credit (millions of real rupees), L is the machinery
number of employees (millions). 2001 Upto Rs. 100 lakh in plant and No condition
The main restriction on the model is that if a3 is not statisti- machinery
cally signicant from zero (say at the 5% level), we may reject the Source: Annual Report (20022003), Government of India, Ministry of Small Scale
non-homogeneous formulation (Eq. (2)) and use for example a CD industries
formulation.
The elasticity output of the amount of credit (ECR ) and of labor
literature (Enders, 1995; Gujarati & Porter, 2009), however, the unit
(EL ) is respectively:
root test is very prominent. The results, shown in Table 2, indicate
ECR = ln Q/ ln CR = a1 + a3 ln L, (3) that for the variables L and Q the null hypothesis of the existence of
unit root (non-stationarity of the data) is rejected at the rst differ-
EL = ln Q/ ln L = a2 + a3 ln CR. (4) ence level and in the three cases that allow for (i) an intercept, (ii)
The scale elasticity is expressed as RTS an intercept and deterministic (linear) trend, and (iii) none. For the
variable CR, the null hypothesis is rejected only for the rst case.
RTS = (ECR + EL ), or a1 + a2 + a3 ln(CR L). (5)

Note that the productivity measure of each input is related to
the productivity of the other input and the level of output; these
are consistent with some of the fundamental assumptions of the
NHPF.
6. Data
The sources of the data are: (i) Handbook of Statistics on the
Indian Economy, 20122013 (Reserve Bank of India) and (ii) Min-
istry of Micro, Small, and Medium Enterprises, Annual Report (2002,
2006, and 2013), Government of India. The variables used are (i) L,
employment in millions, (ii) CR, the amount of loans in millions
of real rupees, and (iii) Q, is output at constant price (millions of
rupees).
6.1. Descriptive statistics
Descriptive statistics on the distributional properties of the vari-
ables are presented (actual values) in Table 1. Test for normality
based on the coefcient of skewness indicates non-normality, all
three variables are positively skewed; the coefcient of kurtosis
also indicates non-normality, it is greater than three (leptokurtic)
for all three variables. The values of the JarqueBera statistics also
indicate the rejection of the null hypothesis of normality (skew-
ness = 0, excess kurtosis = 0) for the variables at = 0.10; these
values are higher than the critical values of the Chi-square distri-
bution with 2 degrees of freedom (Brooks, 2002). The low values of
 also indicate rejection of the null hypothesis.
6.2. Test of unit root
To avoid the problems of spurious regression in empirical
studies using time series data, we test for the stationarity of the
data, using the ADF (augmented DickeyFuller) test which corrects
for uncorrelated error terms. There are several tests discussed in the
6.3. Test of cointegration
The importance of a long run stable relationship among the
variables used in time series econometric models is widely doc-
umented in the literature (Enders, 2010; Johansen, 1988; Maddala
& Kim, 1998). The results of a model derived from cointegrating
variables are stable over the period analyzed and are valid for statis-
tical inferences. Granger (1986) avers that the test of co-integration
can be thought of as a pre-test to avoid the problems of spurious
regression. We examine two versions of the unrestricted rank test
using (i) trace statistics test, and (ii) Max-Eigen statistic under
the assumption of no determining trend. The results, presented on
Table 2A
Provisions of MSMED Act 2006.
A.1. The Government of India has enacted the Micro, Small and Medium
Enterprises Development (MSMED) Act, 2006 in terms of which the
denition of micro, small, and medium enterprises is as under:
(a) Enterprises engaged in the manufacture or production, processing or
preservation of goods as specied below:
(i) A micro enterprise is an enterprise where investment in plant and
machinery does not exceed Rs. 25 lakh; (ii) a small enterprise is an
enterprise where the investment in plant and machinery is more than Rs. 25
lakh but does not exceed Rs. 5 crore; and (iii) a medium enterprise is an
enterprise where the investment in plant and machinery is more than Rs. 5
crore but does not exceed Rs. 10 crore.
(b) Enterprises engaged in providing or rendering of services and whose
investment in equipment (original cost excluding land and building and
furniture, tting and other items not directly related to the service rendered
or as may be notied under the MSMED Act, 2006, are specied below.
(i) A micro enterprise is an enterprise where the investment in equipment
does not exceed Rs. 10 lakh, (ii) A small enterprise is an enterprise where the
investment in equipment is more than Rs. 10 lakh but does not exceed Rs. 2
crore, and (iii) A medium enterprise is an enterprise where the investment
in equipment is more than Rs. 2 crore but does not exceed Rs. 5 crore.
Source: Annual Report (20052006), Government of India Ministry of Small Scale
Industry.
Note: 1 crore = 10 million, 1 lakh = 100,000.
22 H. Ramcharran / The Quarterly Review of Economics and Finance 65 (2017) 1624

Table 3A
Targets for priority sector lending by domestic commercial banks.

A. The domestic commercial banks are expected to enlarge credit to priority sector and ensure that priority sector advances (which include the micro and small
enterprises (MSE) sector) constitute 40 per cent of Adjusted Net Bank Credit (ANBC) or credit equivalent amount of Off Balance Sheet Exposure, whichever is higher.
B. In terms of the recommendations of the Prime Ministers Task Force on MSMEs, banks are advised to achieve a 20 per cent year on year growth in credit to micro
and small enterprises and a 10 per cent annual growth in the number of micro enterprise accounts
C. In order to ensure that sufcient credit is available to micro enterprises within the MSE sector, banks should ensure that:
(a) 40 percent of the total advance to MSE sector should got to micro (manufacturing) enterprises having investment in plant and machinery up to Rs. 5 lakh and
micro (service) enterprises having investment in equipment up to Rs. 2 lakh;
(b) 20 per cent of the total advances to MSE sector should go into micro (manufacturing) enterprises with investment in plant and machinery above Rs. 5 lakh and
up to Rs. 25 lakh and micro (service) enterprises with investment in equipment about Rs. 2 lakh and up to Rs. 10 lakh. Thus, 60 per cent of MSE advances should go
to the micro enterprises.
(c) While banks are advised to achieve the 60% target as above, in terms of the recommendations of the Prime Ministers Task Force, the allocations of 60% of the
MSE advances to the micro enterprises is to be achieved in stages viz. 50% in the year 20102011, 55% in the year 20112012 and 60% in the year 20122013.

Source: Micro and small enterprises (MSE) information kit: Central Bank of India Priority Sector Department.

Table 4A
Outstanding credit to the SME sector by scheduled commercial banks (in Rs. billion).

Year Public sector banks Private sector banks Foreign banks All SCBs

Amount % Amount % Amount % Amount %

2008 1511.374 70.77 469.118 20.96 154.892 7.25 2135.386 100

2009 1914.083 72.73 466.563 18.21 180.634 7.025 2561.280 100
2010 2763.189 76.27 648.247 17.89 211.470 5.83 3622.907 100
2011 3694.300 70.2 881.160 18.41 209.810 4.38 4785.270 100

Source: BIS central bankers speeches: Address by Dr K C Charkrabarty, Deputy Governor of the Reserve Bank of India, at the SME Banking Conclave 2012, Mumbai, 4 Feburary
2012.

Table 3, indicate cointegrating relationship among all three vari-

ables L, CR, and Q expressed logarithmic transformation (ln L, ln CR,
and ln Q) based on the rank test.
Despite some evidence of non-normality in the distribution of
the data, the existence of stationarity and cointegration of the vari-
ables ensure the reliability of the estimated results.

7. Discussion of regression results

The regression results of Eq. (2) (with t values in parentheses)

are:

Fig. 6. Productivity estimates, EL , ECR , and RTS.

2
= 0.874, DW = 0.523, F = 80.8, Prob(F-statistic) = 0.00,
R2 = 0.885, .R
* denotes signicance at = 0.05 level and less.
All three coefcients, based on the t-values, are statistically sig-
nicant at the 1% level; the coefcient of CR is positive and of L
negative. Importantly, the signicance of the coefcient a3 justi-
es the relevance of the NHPF technique. An adjusted .R 2 of 0.87
indicates a high explanatory power of the model; the F statistic is
also signicant. The DW statistic indicates no evidence of positive
autocorrelation at = 0.10. Since R2 < DW, there is no reason to sus-
pect that the estimated results are spurious (Granger & Newbold,
1974), this is supportive of the results of the unit root test and the
cointegration test. The productivity estimates based on the above
results are presented in Table 4. Fig. 6 shows the graphical pattern
of productivity estimates.
The results indicate improved efciency in the operation of SME
over the period studied an increase in RTS from 0.809 to 0.607
mainly due to the increasing productivity of bank loans. From
19781979 to 19971998 RTS is negative; however in the post
19981999 period RTS is positive and operation is at decreasing
RTS (RTS < 1). The improvement is apparently due to the effects of
the special preferential treatment to this sector under economic
liberalization reforms during the 1990s. The other studies, Bhavani
(1991), Ramaswamy (1994), and Little, Mazumdar, and Page (1987)
nd RTS less than unity using different production specication and
data from an earlier period. The output elasticity of labor (EL ) is neg-
ative throughout the period but increases from 1.5 to 0.728; the
primary cause is the low level of skills. However, the increase could
be attributed to the easement of some of the rigid labor laws dur-
ing this period; Ahluwalia (1994) and Aiyar (2011) have noted that
Indias restrictive labor laws (which make it difcult to retrench
workers in small companies) and prevent exibility in input choice
(production technique). Other studies (Bloom, Mahajan, McKenzie,
& Roberts, 2010; Bloom & Van Reenen, 2007) also report very low
productivity of labor in rms in developing countries.
The important aspect of our ndings is the positive and increas-
ing productivity of bank loans, ECR increases from 0.76 to 1.23.
The possible contributory factors are: (i) increasing allocation to
the SME sector (see Fig. 2), and (ii) the decrease in the number of
sick units beginning 2000 (see Fig. 3). Several studies (Berger &
DeYoung, 1996; Mester, 1996, 1997) have documented the impact
of problem loans on the efciency of nancial institutions.9 Other
studies (McKenzie & Woodruff, 2005; Mel, McKenzie, & Woodruff,
2008) also report high return on capital investment in small rms
in developing countries. Since RTS is the sum of ECR and EL ,
this improved overall efciency is attributed to the positive and
increasing value of ECR . The negative EL has important implications
for policy making since one of the goals of SME development is the
9
We could not control for the problems of bad loans on efciency, for example, by
including it as an independent variable because of the lack of systematic published
date on bad loans.
 H. Ramcharran / The Quarterly Review of Economics and Finance 65 (2017) 1624
efcient use of the countrys abundant factor; a negative EL does not
bode will with this objective. Tybout (2000), in analyzing the role
of skilled workers in efciency, notes that exibility in the produc-
tion process and the ability to absorb new technology are positively
related to the stock of human capital.
8. Policy implications and conclusion
Many studies have shown that access to nance is closely
linked with the performance and efciency of SME in developing
countries. Within the framework of money as an input in the
production function, the NHPF estimates indicate improvement in
scale efciency the SME sector of India mainly due to the positive
and increasing productivity of bank loans. Other studies discussed
in this paper before also report a positive impact of external
nancing on growth. Future lending to this sector will depend on:
(i) the rising cost of credit, (ii) economic recession that could slow
the demand for output from this sector, and (iii) banks compliance
with international regulations to increase capital. Most of the loans
to SME are from state owned banks which account for about 75 per
cent of the banking assets of the country. The prospects of nancing
from other sources are not promising because of the following: (i)
venture capital is primarily for rms with high growth potential,
and (ii) the national equity markets will not accommodate small
rms. A recent proposal to reform the banking industry has the
potential of increasing loans to the SME sector from private and for-
eign banks. The new measures will allow new licenses to increase
foreign investment and allow private companies to transform
their nancial services department to full-edged banks.10 Addi-
tionally, a recent report indicates that there is a growing number
of bad loans owed to government banks by the top Indian com-
panies; this is considered a signicant risk factor to the nancial
system.11
Some analysts have expressed reservations about credit ef-
ciency regarding the priority lending mandate to SME in light of
the high number of sick units. Tybout (2000) indicates that the
potential gains from scale economy exploitation may be dampened
by policies that prop up sick rms and thereby saturating the mar-
ket with inefcient producers and discouraging better rms from
entering. Many studies (Boot, Greenbaum, & Thakor, 1993; Rajan &
Zingales, 1998) have favored market lending since private banks
have better screening and monitoring methods; this could possibly
eliminate the existence of many sick banks.
Labor productivity, though it increases, is still negative. Business
policies to address the low level of skills and surplus workers are
needed to improve efciency. It is estimated that by 2016, 50 per
cent of Indias labor force will be in the age group of 1525 years;
thus, there will be tremendous pressure for this group to be ef-
ciently employed. There are also complementary/external factors
to be addressed; these include the non-availability of some raw
materials, power shortages, transport and nancial bottlenecks,
production disruption and demand shortfalls.
This research adds signicantly to the empirical literature on the
effects of bank lending to SMEs; the estimation technique enables
us to identify the sources of efciency/inefciency. This enables the
formulation of appropriate policies to enhance productivity. The
slight limitation of the model, like other studies, is some aggre-
gation bias since there are rms (units) with different sizes, ages,
and operating in different locations. Systematic published data at
10
India Eases Its Hold on Bank Lending, Financial Times, J. Crabtree, December
20, 2012.
11
Bad Loans at State-Run Banks Add to Indias Woes by Vikas Bajaj, New York
Times, March 15, 2012.
23
the disaggregate level are not published and unavailable, plus the
latest data are not very timely.
References
Ahluwalia, M. S. (2002). Economic reforms in India since 1991: Has gradualism
worked? Journal of Economic Perspectives, 16, 6788.
Ahluwalia, M. S. (1994). Indias quiet economic revolution. Columbia Journal of World
Business, Spring, 112.
Aiyar, S. S. A. (2011, July). The elephant that became a tiger. CATO Institute, Center for
Global Liberty and Prosperity.
(20022003). Annual Report. Government of India, Ministry of Small Scale Industries.
(20052006). Annual Report. Government of India, Ministry of Small Scale Industries.
Ayyagari, M., Beck, T., & Demirguc-Kunt, A. (2007). Small and medium enterprises
across the global. Small Business Economics, 29, 415434.
Ayyagari, M., Demirguc-Kunt, A., & Maksimovic, V. (2012). Financing of rms in devel-
oping countries. Policy research working paper 6036. The World Bank.
Bairam, E. I. (1997). Homogenous and nonhomogeneous production functions. Avebury
Press.
Bajaj, V. (2012, March). Bank loans at state-run banks add to Indias woes. In New
York Times.
Beck, T., Demirguc-Kunt, A., & Maksimovic, V. (2005). Financial and legal constraints
to growth: Does rm size matter? Journal of Finance, 60(1), 137177.
Beck, T., Demirguc-Kunt, A., & Martinez Peria, M. S. (2011). Bank nancing for SMEs:
Evidence across countries and bank ownership types. Journal of Financial Services
Research, 39, 34.
Beck, T., Demirguc-Kunt, A., & Maksimovic, V. (2008). Financing patterns around the
world: Are small rms different? Journal of Financial Economics, 89, 467487.
Beck, T., Demirgu-Kunt, A., Laeven, L., & Levine, R. (2008). Finance, rm size and
growth. Journal of Money Credit and Banking, 40, 13791405.
Berger, A. N., & Humphery, D. B. (1997). Efciency of nancial institutions: Interna-
tional survey and directions for future research. The Wharton Financial Institution
Center.
Berger, A. N., Hunter, W. C., & Timme, S. G. (1993). The efciency of nancial insti-
tutions: A review and preview of research past, present, and future. Journal of
Banking and Finance, 17, 221249.
Berger, A. N., & DeYoung, R. (1996, August). Problem loans and cost efciency in com-
mercial banks. Working paper. Washington, DC: Broad of Governor of the Federal
Reserves System.
Bhavani, T. A. (1991). Technical efciency in Indian modern small scale sector:
An application of Frontier production function. Indian Economic Review, 31,
149166.
Bloom, N., & Van Reenen, J. (2007). Measuring and explaining management practices
across rms and countries. Quarterly Journal of Economics, 122(November (4)),
13511408.
Bloom, N., Mahajan, A., McKenzie, D., & Roberts, J. (2010). Why do rms in devel-
oping countries have low productivity? American Economic Review, 100(2),
619623.
Boot, A. W. A., Greenbaum, S. I., & Thakor, A. V. (1993). Reputation and discretion in
nancial contracting. American Economic Review, 83, 11651183.
Brooks, C. (2002). Introductory econometrics for nance. Cambridge University Press.
Calice, P., Chando, V., & Sekioua, S. (2012). Bank nancing to small and medium enter-
prises in East Africa: Finding of a survey in Kenya, Tanzania, Uganda and Zambia.
Working Paper Series, Working Paper No. 146. African Development Bank Group.
Chakrabarty, K. C. (2012, February 4). Empowering MSMEs for nancing inclusion and
growth The role of banks and industry associations. Mumbai: BIS Central Bankers
Speeches.
Christensen, L., Jorgenson, D., & Lau, L. (1973). Transcendental logarithmic produc-
tion Frontiers. Review of Economics and Statistics, 55(February), 2845.
Chung, J. W. (1994). Utility and production functions. Blackwell Publishers.
Clarke, G., Cull, R., Peria, M. S. M., & Sanchez, S. M. (2005). Bank lending to small
businesses in Latin America: Does bank origin matter? Journal of Money, Credit,
and Banking, 37, 93118.
Crabtree, J. (2012, December 20). India eases its hold on bank lending. Financial Times.
De la Torre, A., Martinez Peria, M. S., & Schmukler, S. (2010). Bank involvement
with SMEs: Beyond relationship lending. Journal of Banking and Fiancing, 34,
22802293.
Dewatripont, M., & Maskin, E. (1995). Credit efciency in centralized and decentral-
ized economics. Review of Economic Studies, 62, 541555.
Enders, W. (2010). Applied econometric time series. John Wiley & Sons.
Enders, W. (1995). Applied econometric time series. John Wiley & Sons.
Falls, G. A., & Natke, P. (1988). The demand for liquid assets: A rm level analysis.
Southern Economic Journal, 54(3), 630642.
Falls, G. A., & Natke, P. (2010). Economies of scale and the demand for money. Small
Business Economics, 35, 283298.
Finnerty, J. D. (1980). Real money balances and the rms production function: Note.
Journal of Money, Credit and Banking, 12, 666671.
Fishman, R., & Love, I. (2007). Financial dependence and growth revisited. Journal of
the European Economic Association, 5(23), 470479.
Fried, H. O., Knox Lovell, C. A., & Schmidt, S. S. (2008). The measurement of productive
efciency and productivity growth. Oxford University Press.
Friedman, M. (1969). The optimum quantity of money and other essays. University of
Chicago Press.
Granger, C. W. J., & Newbold, P. (1974). Spurious regressions in econometrics. Journal
of Econometrica, 2, 111120.
24 H. Ramcharran / The Quarterly Review of Economics and Finance 65 (2017) 1624
Granger, C. W. J. (1986). Developments in the study of co-integrated economic vari-
ables. Oxford Bulletin of Economics and Statistics, 48, 226.
Gujarati, D., & Porter, D. C. (2009). Basic econometrics. Irwin: McGraw Hill.
(20102011). Handbook of Statistics on the Indian Economy. Reserve Bank of India.
Hasan, M. A., & Mahmud, S. F. (1993). Is money an omitted variable in the production
function? Some further results. Empirical Economics, 18, 431445.
Intrilligator, M. (1978). Econometric models, techniques, and applications. Englewood
Cliffs, NJ: Prentice Hall.
Johansen, S. (1988). Statistical analysis of co-integrating vectors. Journal of Economic
Dynamics and Control, 12, 231254.
Kamesam, V. (2003). Financing for entrepreneurship and SMEs An Indian perspec-
tive. BIS Review, (40), 15.
Khan, A. H., & Ahmad, M. (1984). Real money balances in the production function of
a developing country. Review of Economics and Statistics, 336340.
Kim, Y. H. (1992). The Translog Production Function and Variable Returns to Scale.
Review of Economics and Statistics, 74, 546552.
Kumar. C. What is the role and importance of small scale industry in India? (www.
preservearticles.com).
King, R., & Levine, R. (1993a). Finance, entrepreneurship, and economic develop-
ment. Journal of Monetary Economics, 32, 513543.
King, R., & Levine, R. (1993b). Finance and growth: Schumpeter might be right.
Quarterly Journal of Economics, 108, 717738.
Laumas, P. S., & Mohabbat, K. A. (1980). Money and the production function: A case
study of France. Weltwirtscaftliches Archiv, 116, 685696.
Levine, R., & Zervos, S. (1998). Stock markets, banks, and economic growth. American
Economic Review, 88, 537558.
Levine, R. (1997). Financial development and economic growth: Views and agenda.
Journal of Economic Literature, 35, 688726.
Little, I., Mazumdar, D., & Page, J. M., Jr. (1987). Small manufacturing enterprises: A
comparative analysis of India and other economices. NY: Oxford U. Press.
MacKinnon, J. G. (1996). Numerical distribution functions for unit root and cointe-
gration tests. Journal of applied econometrics, 11, 601618.
Maddala, G. S., & Kim, I. M. (1998). Unit roots, cointegration, and structural change.
Cambridge University Press.
Mead, D., & Liedholm, C. (1998). The dynamics of micro, and small enterprises in
developing countries. World Development, 26, 6174.
Mel, S. D., McKenzie, D., & Woodruff, C. (2008). Returns to capital in microenter-
prises: Evidence from a eld experiment. Quarterly Journal of Economics, 123(4),
13291372.
McKenzie, D., & Woodruff, C. (2005). Experimental evidence on returns to capital
and access to nance in Mexico. World Bank Economic Review, 22(3), 457482.
Mester, L. J. (1996). A study of bank efciency taking into account risk-preferences.
Journal of Banking and Finance, 20, 10251045.
Mester, L. J. (1997). Measuring efciency at U.S. banks: Accounting for heterogeneity
is important. European Journal of Operational Research,
Rajan, R. G., & Zingales, L. (1998). Financial dependence and growth. American Eco-
nomic Review, 88, 559587.
Ramaswammy, K. V. (1994). Technical efciency in modern small-scale rms in
Indian industry: Applications of stochastic production Frontiers. Journal of Quan-
titative Economics, 10, 309324.
Ramcharran, H. (2001). Productivity, returns to scale and the elasticity of factor
substitution in the USA apparel industry. International Journal of Production Eco-
nomics, 73(3), 285291.
Ramcharran, H. (2011). The pharmaceutical industry of Puerto Rico: Ramications
of global competition. Journal of Policy Modeling, 33, 395406.
Ramcharran, H. (2012). Estimating the production efciency of US foreign direct
investment. Managerial and Decision Economics, 33, 273281.
Reserve Bank of India. (2012). Master circular Lending to priority sector, July 02.
Revankar, N. S. (1971). A class of variable elasticity of substitution production func-
tions. Econometrica, 39(1), 6171.
Schiffer, M., & Weder, B. (2011). Firm size and the business environment: Worldwide
survey results. IFC Discussion Paper, 43. Washington, DC: The World Bank.
Sinai, A., & Stokes, H. H. (1981). Money and production function: A reply to Boyes
and Kavanaaugh. Review of Economics and Statistics, 63(2), 313318.
Tybout, J. R. (2000, March). Manufacturing rms in developing countries: How well
do they do and why? Journal of Economic Literature, 1144.
Vinod, H. D. (1972). Non-homogeneous production functions and applications to
telecommunications. Bell Journal of Economics and Management Science, 3(2),
531543.
World Bank. (2010). In K. Kushnir, M. Mirmulstein, & R. Ramalho (Eds.), Micro, small
and medium enterprises around the world: How many are there and what affects
the count?.