THE EFFECTS OF ECONOMIC REFORMS ON MANUFACTURING DUALISM: EVIDENCE FROM INDIA

1
THE EFFECTS OF ECONOMIC REFORMS ON MANUFACTURING
DUALISM EVIDENCE FROM INDIA

• Vinish Kathuria
• IIT Bombay
• (Rajesh Raj S N
• CMDR Dharwad
• Kunal Sen
• University of Manchester)
• PROJECT FUNDED by Economic and Social Research
  Council, UK
• ESDS Conference Nov 28 2011

2

1. Introduction

• Dualism in the organization of production
  activities is a permanent feature of developing
  countries, where a large informal sector
  coexists with a relatively small formal sector.
• Despite strong economic growth, the persistence
  in the size of informal sector along with large
  differences in productivity between the formal
  and informal sectors has remained a matter of
  concern

3
Objective

• This paper examines the effects of economic
  reforms on manufacturing dualism
• Dualism differences in efficiency levels
  between informal and formal manufacturing firms
• Considered absolute (AE) and relative technical
  efficiency (RE)
• AE - extent to which firms in the manufacturing
  sector are producing the maximum possible
  output, for a given bundle of inputs, in a given
  industry (improvement in level)
• RE - extent to which other firms are close to the
  most efficient firm in a given industry
  (improvement in equity)

4
If dualism has reduced.

• Our objective is to investigate whether economic
  reforms have led to an increase in absolute
  technical efficiency and a decrease in RTE of
  informal firms relative to formal firms
• Reforms would have reduced manufacturing dualism
  if absolute TE of informal firms have improved
  relative to formal firms while RTE has fallen
• This implies that reforms would have had a
  positive effect on both the level and
  distribution of efficiency in the informal
  sector as compared to the formal sector, and
  therefore, would have strong pro-poor growth
  effects

5
Present study

• focuses on Indian manufacturing sector where
• about 80 per cent of manufacturing employment
  and 17 per cent of manufacturing output is in
  the informal sector
• we have large cross-sections of firm-level
  data-set for both the informal and formal
  manufacturing sectors for four years, beginning
  in 1989-90 and ending in 2005-06

6
2. Methodology Sample Selection Framework

• We employ a method developed by Greene (2010)
  that corrects for self-selection bias to obtain
  firm level measure of absolute and relative
  technical efficiency
• Theoretical models predict that firms choose
  whether to be in the informal or formal sector
  depending on factors which are mostly available
  to the formal sector (Dessy and Pallage 2003,
  Straub 2005, Ulyssea 2010)
• Efficiency comparison needs to address the
  endogeneity of firm location
• Else bias upwards the efficiency levels of
  formal firms if these levels depended on the
  firm being located in the formal sector.

7
3. Empirical Specification

• First stage analysis
• Firms choose between formal or informal sector
  subject to a set of variables that capture the
  benefits and costs of formalization
• We estimate probit models of the following type
• F aZi wi
• Where F is a dummy variable that takes the value
  1 for formal sector firms and 0 for informal
  sector firms,
• Z is a vector of variables explaining the
  decision to formalize,
• a is a vector of parameters, and wi is the white
  noise error term

8
3. Empirical Specification

• Drawing from the recent theoretical literature,
  we considered
• labour laws,
• priority sector lending,
• price of power supply and
• firm size
• as the variables that determine the benefits and
  costs of formalization in Indian manufacturing
• Choice of these variables also exploit the fact
  that there are important differences in
  institutions relating to labour regulation,
  access to credit and the provision of
  infrastructure across Indian states and over time

9
3. Empirical Specification

• Second stage analysis
• We modelled the production behaviour of formal
  and informal sector firms using a simple
  Cobb-Douglas function as follows
• ln(QiT) ?0 ?1ln(KiT) ?2ln(LiT) (viT -
  uiT)
• Where T1989-90, 1994-95, 2000-01 2005-06 and
  i is the firm.
• Q is output, K is capital, L is labour, and ?s
  are parameters to be estimated
• viT captures the random shocks that are beyond
  the control of firms
• uiT represents technical inefficiency which is
  the combined outcome of non-price and
  organizational factors that constrains a firm
  from achieving their maximum possible output

10
3. Testing for Impact of Reforms on Duality

• Effect of reforms on duality
• an interaction term for our reform variable with
  that of the formal firms is introduced
• TEijt a ß1FORMALijt ß2REFORMSjt
  ß3FORMALxREFORMSijt dj ?t eijt
• If dTE/dFORMAL ß1 ß3REFORMS lt 0
• ? reforms have reduced duality
• We used estimates of absolute and relative
  technical efficiency alternately as our
  dependent variable

11
4. Data and Variables

• ASI data for the formal sector and
• NSSO data - surveys on the unorganized
  manufacturing sector for the informal sector
• Unit level data for four years, 1989-90,
  1994-95, 2000-01 and 2005-06 repeated
  cross-sections, and not in panel form
• Confined the analysis to those informal firms
  that employ at least one hired worker
• serious limitations of the data on output and
  capital stock for firms that employ only family
  labour
• interest is in firms that are likely to modify
  their behaviour in response to policy changes as
  family firms in informal sector are often in
  business to bring in additional income with
  little additional effort and are unlikely to
  expand or invest in their businesses

12
4. Data and Variables Reform Variables

• Tariff
• trade and industrial output data of the World
  Bank Trade Data-base (World Bank 2006)
• Delicensing
• Data on number of industries delicensed is
  obtained from Aghion et al (2006) and different
  notifications of GoI.
• constructed as the total number of four-digit
  industries delicensed in a year to that of total
  number of four-digit industries in the sector
• Dereservation
• list of number of products dereserved is obtained
  from different notifications of the GoI
• constructed as the cumulative number of products
  dereserved in respective two-digit industries to
  that of total reserved products

13

4. Trend of Reform Variables

1. most of the delicensing reforms were over by 1995
2. dereservation reforms started after 1995
3. by 2005-06, only 4 of industries required
   delicensing, nearly one-third products needed to
   be dereserved and average tariff was 29 (ranging
   between 24 to 37), much below 100 tariff
   around 1991

14
5. Impact of Reforms on Duality
Impact of reforms on duality of Indian
manufacturing industry
VARIABLES Absolute Technical Efficiency Absolute Technical Efficiency Absolute Technical Efficiency Absolute Technical Efficiency Relative Technical Efficiency Relative Technical Efficiency Relative Technical Efficiency Relative Technical Efficiency
VARIABLES (1) (2) (3) (4) (5) (6) (7) (8)
Formal -0.223 (0.004) -0.220 (0.00469) -0.00277 (0.00199) 0.205 (0.00455) 17.977 (0.310) 39.41 (0.297) 3.972 (0.111) -2.318 (0.285)
REFORMS 0.009 (0.000) -0.166 (0.007)
FormalREFORMS 0.005 (0.000) -0.240 (0.006)
Delicense 0.00415 (4.84e-05) -0.0724 (0.00247)
Formaldelicense 0.00317 (5.37e-05) -0.409 (0.00328)
Dereserv 0.00208 (9.49e-05) -0.102 (0.00618)
Formaldereserv 0.00205 (6.73e-05) 0.109 (0.00439)
Tariff 0.000885 (0.000110) -0.105 (0.00683)
Formaltariff -0.00338 (8.64e-05) 0.158 (0.00569)
Constant 0.247 (0.004) 0.269 (0.00362) 0.357 (0.00304) 0.333 (0.00901) 36.937 (0.240) 29.55 (0.194) 40.18 (0.194) 43.79 (0.620)
Ind. dmy Y Y Y Y Y Y Y Y
Yr dmy Y Y Y Y Y Y Y Y
N 173108 173108 173108 173108 173108 173108 173108 173108
R-squared 0.19 0.179 0.139 0.119 0.23 0.314 0.215 0.215
15
5. Results

• For the average values of REFORM variable, the
  efficiency difference between formal and
  informal firms is 0.04 for absolute efficiency
  and 5.35 for RTE.
• At the mean value of absolute technical
  efficiency (0.44), this suggests that reforms
  have increased the efficiency of formal firms by
  9 per cent vis-à-vis informal firms.
• Similarly, relative to a mean RTE of 24.5, the
  efficiency gap between the representative formal
  firm and the MEF in a particular industry has
  increased by 21.9 as compared to the efficiency
  gap between the representative informal firm and
  the MEF in the same industry.

16
5. Results TE vs. Reforms
REFORM vs efficiency relationship - stronger in
the formal sector as compared to the informal
sector
17
5. Results RTE vs. Reforms
REFORM vs efficiency relationship - stronger in
the formal sector as compared to the informal
sector
3/13/2015
17
Economic reforms and manufacturing dualism
18
6. Conclusions

• Do economic reforms reduce or exacerbate
  manufacturing dualism?
• Our regression results suggest that these
  reforms have had an unambiguous positive effect
  on the absolute efficiency levels of formal
  firms relative to informal firms, and a negative
  effect on the relative efficiency levels

19
6. Conclusions

• Economic reforms (as measured by individual
  reform variable or composite index) had an
  unequalising effect on the manufacturing sector,
  as it exacerbated the productivity differentials
  between average firm in the informal formal
  sectors, and
• at the same time, it reduced the gap in
  efficiency levels between formal firms as
  compared to informal firms
• ? Dualism has increased in Indian mfrg. after the
  reforms, both between formal and informal firms
  and within the informal sector