Gandhi Institute of Development Research, Mumbai. Figure 2 plots employ-ment (indexed log val-ues), real gross value added (GVA) (output) and real gross fixed capital formation (GFCF) (fixed investment) in only 0.8%, relative to growth rates of 6.9% in Figure 2: Employment, Output and Investment in the Factory Sector (1982-2009) 120 90 Employment 1982 1986 1990 1994 1998 2002 2008 GVA GFCF 4 2 0 1982 1986 1990 1994 1998 2002 2008 DISCUSSION
10
ment in 13 years, over
the previous peak in
Number
8
1995-96.
6
Workers Employees 115
the factory sector for
Chandan Mukherjee’s astute reading of the same period. Over
Index of log values
Growth in Organised Manufacturing Employment: A Comment
R Nagaraj
output and 4.5% in fixed investment. Undoubtedly, the recent boom in fixed investment is noteworthy, peaking at onethird of the domestic fixed investment in 2008-09, as per the National Accounts. So, over a longer period, and relative to output and investment, employment growth is surely modest. Evidently, the refrain of “jobless growth” is perhaps not (at least as yet) completely out of place. Table 1 reports the trend growth rates for all the three variables since 1981-82 for different time periods.
Validity of Tests
Reviewing the competing propositions to explain the employment growth (or lack of it), Goldar attributes the recent boom to the variations in state level labour reforms, based on a positive correlation coefficient between (i) employment elasticity of output during 2003-04 and 2008-09, and (ii) the labour reforms index across 20 states. The estimated value of the
Table 1: Trend Growth Rates for Factory Sector (% per year)
Time Period Employment Output Fixed of Workers Investment
Bishwanath Goldar’s argument that the rapid growth in employment in organised manufacturing between 2003-04 and 2008-09 can be explained by labour reforms at the state level does not stand up to close statistical scrutiny. Employment growth did accelerate but when viewed over a longer period the accretion as yet remains small.
B
We take a closer look at these findings as they have far-reaching policy implications.
Figure 1 reports employment in the factory sector (most of which is in registered manufacturing) since 1981-82. The trends for workers and employees (workers plus supervisors) are similar. The impressive turnaround since 2003-04 largely represents a recovery of the employment lost over the previous
Figure 1: Factory Employment (1981-82 to 2008-09) (in million)
nine years and, a mere
12
12% rise in employ
| 1982-2009 | 0.8 | 6.9 | 4.5 |
| 1991-2009 | 0.6 | 6.3 | 3.4 |
| 2000-2009 | 5.1 | 13.4 | 14.0 |
Real output and investment series are obtained by deflating the nominal series by implicit GDP deflators for registered manufacturing and gross fixed capital formation respectively with the base year 1999-2000.
110
an earlier draft of this note, and Anadleeb close to three decades
105
Rahman’s research assistance are gratefully
since 1981-82, on a trend
100
95
Economic & Political Weekly
EPWmarch 19, 2011 vol xlvi no 12
DISCUSSION
Table 2: Correlation Coefficient between Employment Elasticity and Labour Reforms Index
Rank Correlation Simple Correlation Coefficient
All 20 states 0.279 0.353
17 major states 0.292 0.399
Source: Table 2 of Goldar’s paper.
Table 3: Inter-correlation Matrix
Workers GVA GFCF Labour Market Index
Workers 1
GVA 0.514* 1
GFCF 0.438** 0.828* 1
Labour market index 0.219 0.002 0.049 1
For the first 3 variables we have used the average of annual growth rates between 2003-04 and 2008-09. ** Significant at 90% level; * at 95% level.
coefficient at 0.35 is not statistically significant, but by dropping the observation for Andhra Pradesh (which has the second highest value in the labour reforms index), Goldar has obtained a statistically significant result.
To clarify the ambiguity of the above result, and by applying a more appropriate non-parametric test, we have computed a Pearson rank-correlation coefficient with the same variable using Goldar’s data. The results are reported as applied to all the 20 states, as well as for a subset of 17 “major” states that individually accounted for at least 1% of factory employment in 2003-04 (Table 2). The estimated coefficients in both the cases are positive but not statistically significant, though the association is stronger for the major states. Thus, Goldar’s inference of a positive effect of labour reforms in explaining inter-state variation in employment elasticity of output is not statistically valid.
In seeking to understand the recent employment boom in terms of demand factors, it is perhaps more appropriate to include the unprecedented investment boom (as evident from Figure 2), along with output growth. To find out if the labour market reforms at the state level have an influence, we have also included Goldar’s labour reforms index. A simple correlation matrix of these four variables for 17 major states, reported in Table 3, shows the following:
In sum, Goldar’s observation of a turnaround in factory employment between 2003-04 and 2008-09 is significant enough to warrant reporting. Indeed the growth is a welcome change in the employment scene. But when viewed over a longer period, and in relation to the trends in output and investment, the turnaround appears meagre in relation to the growth in the workforce.
On a closer statistical scrutiny, however, his finding of a positive association between the employment boom and the labour law reforms at the state level is not tenable.
EPW Research Foundation (A UNIT OF SAMEEKSHA TRUST)
YYYGRYTſVUKP +PFKC 6KOG 5GTKGU
'294( KU JCRR[ VQ CPPQWPEG VJG KPVTQFWEVKQP QH CP QPNKPG UGTXKEG EJTKUVGPGF CU Ŏ+PFKC 6KOG 5GTKGUŏ 6JG RTQLGEV GPXKUCIGURTQXKUKQPQHQPNKPGFCVCUGTXKEGKPVYGNXGOQFWNGUFKURNC[KPIVKOGUGTKGUQPCYKFGTCPIGQHOCETQGEQPQOKE CPF ſPCPEKCN UGEVQT XCTKCDNGU KP C OCPPGT EQPXGPKGPV HQT TGUGCTEJ CPF CPCN[VKECN YQTM 6JKU KU VCTIGVGF VQ DGPGſV RCTVKEWNCTN[ UVWFGPVU TGUGCTEJ UEJQNCTU CPF VJG CECFGOKE EQOOWPKV[ DQVJ KP +PFKC CPF CDTQCF
6JKU QPNKPG UGTXKEG KU C RCTV QH VJG RTQLGEV HWPFGF D[ VJG 7PKXGTUKV[ )TCPVU %QOOKUUKQP
7)% CPF GZGEWVGF D[
VJG 6CVC +PUVKVWVG QH 5QEKCN 5EKGPEGU
6+55 /WODCK CPF VJG 'EQPQOKE CPF 2QNKVKECN 9GGMN[ '29 0QYVJGſPCPEKCNOCTMGVUOQFWNGKPYYYGRYTſVUKPJCUDGGPOCFGHWNN[QRGTCVKQPCN6JGFGOQXGTUKQPECPDG CEEGUUGF D[ HTGG TGIKUVTCVKQP 6JG GZKUVKPI OGODGTU CNTGCF[ TGIKUVGTGF YKVJ WU CPF CEEGUUKPI OGODGT UGTXKEGU CV YYYGRYTHKP YKNN TGSWKTG PQ HTGUJ TGIKUVTCVKQP 6Q ICKP HWNN CEEGUU VJG UWDUETKRVKQP TCVGU CTG
5WDUETKRVKQP4CVGU2GT#PPWOVQ#EEGUU(KPCPEKCN/CTMGVU/QFWNG
+PFKXKFWCNU
7PKXGTUKVKGUGGOGF
1VJGT +PUVKVWVKQPU
7PKXGTUKVKGU QT %QNNGIGU
#EEQTFKPI VQ 0WODGT QH
#EEQTFKPI VQ 0WODGT QH
%QPEWTTGPV 7UGTU
%QPEWTTGPV 7UGTU
7R VQ
VQ
/QTG VJCP
7R VQ
VQ
/QTG VJCP
+PFKC
KP 4U
(QTGKIP
KP 75
6JGUKVGYKNNDGUJQTVN[GZRCPFGFVQEQXGTVJTGGOQTGOQFWNGUQP$CPMKPI5VCVGQOGUVKE2TQFWEVCPF#ITKEWNVWTG 'KIJV OQTG OQFWNGU YKNN DG CFFGF UQQP VJGTGCHVGT
(QT CP[ HWTVJGT FGVCKNU QT ENCTKſECVKQPU RNGCUG EQPVCEV
KTGEVQT '29 4GUGCTEJ (QWPFCVKQP % #MWTNK +PFWUVTKCN 'UVCVG #MWTNK 4QCF -CPFKXNK
'CUV /WODCK
RJQPG QT OCKN VQ GRYTHXUPNEQO
march 19, 2011 vol xlvi no 12
EPW
