Productivity & Innovation

Manufacturing, management and mysteries

  • Blog Post Date27 July, 2012
  • Print Page
Author Image

Nirvikar Singh

University of California, Santa Cruz

boxjenk@ucsc.edu

The private sector is crucial to India's development. This column asks how it can be more effective. Looking at a large sample of Indian manufacturers it suggests that what many firms may be missing is good management.

In 2006-2007, I worked with Shubhashis Gangopadhyay and Manisha G Singh on a study of Indian manufacturing firms (Gangopadhyay et al. 2008), using data from thousands of company plants made available through the Annual Survey of Industries (ASI). The goal of the study was to examine the effects of investment in hardware and software on plant performance, including profits, value added and employment. It turned out that there was a clear positive effect of information technology (IT) capital on all these outcomes, for instance a larger stock of hardware and software leading to higher profits, even controlling for other factors. At the same time, many plants in the sample used no IT at all, raising the question as to why they did not, if the payoff was so clearly positive. Examining the factors that influenced decisions on whether or not to invest in IT suggested that market imperfections and certain kinds of input constraints (particularly for electric power) played an important role.

One drawback of the ASI data at the time was that plants could not be identified, so we could not examine performance of plants over time. This also meant that we could not control for unobservable plant-specific effects, with managerial quality being an obvious possibility. Recently, Shruti Sharma and I (Sharma and Singh 2012) have revisited the earlier analysis using newer data, which allows us to dig deeper. Echoing the earlier empirical results, we find some evidence that plants with higher gross value added have higher levels of IT capital stock, controlling for other inputs. However, this effect is considerably reduced when controlling for some of the unique characteristics of the plants themselves - what economists call 'plant-level fixed effects'. We could plausibly interpret this result as an indication that unobserved managerial quality is an important factor in the impact of IT capital on productivity.

Management matters

When we examine the demand for IT investment, we again find that plant level fixed effects underlie what seems to be a positive impact of IT capital stock, profits and availability of skilled labour on IT investment demand. This result and the previous one would suggest that management may have an important role to play in the performance of Indian manufacturing plants, and may drive decisions about key input choices that affect productivity. This interpretation is consistent with two other empirical analyses of Indian management practices. Bloom and van Reenen (2010) find that Indian firms with strong management practices are comparable to the best US firms on this dimension. However, there is a thick tail of badly-run Indian firms(by their measure of management practices), which often neglect basic tasks such as collecting and analysing data, setting clear performance targets, and linking pay to performance. Bloom et al. (2012) perform a controlled experiment by helping a sample of Indian textile firms implement better management techniques, and indeed find that the treatment firms improved productivity by 17% over the control group. This provides very direct evidence that ‘management matters’ for at least a subset of Indian firms. While our work cannot provide such a direct test, our results are certainly suggestive of a similar phenomenon, in a much larger sample of Indian manufacturing plants.

Skills and imported goods

We also explore the effects of skill composition and use of imported intermediate inputs on the productivity of IT capital. Work by my current co-author (Sharma 2012) suggests that imports of intermediate goods have been important in affecting the skill composition of the workforce in Indian manufacturing plants. We find that plants that use intermediate goods imports are more productive on average, as measured by gross value added (GVA). We also find that IT capital and imported intermediates are substitutes to some extent. When we examine the effect of skill composition, where we might expect that a higher proportion of skilled workers (as roughly measured by salaried employees, as opposed to production-line wage workers) improves productivity, we do not find a positive effect. However, the interaction of skill composition with IT capital is positive, consistent with an expected complementarity found in developed country studies.

More questions to answer

We view our results as suggestive, and there remains a significant amount of analysis to be done. Certainly, there is a large degree of ignorance about what drives the performance of Indian manufacturing. For example, another new paper, by Bollard et al. (2012), examines ASI data all the way from 1980 to 2007 (we use only five years, from 2003-2007). They find a large but imprecisely measured acceleration in productivity growth starting in the early 1990s and trace it to productivity growth within large plants (200 workers or more), as opposed to reallocation across such plants. Furthermore, when they look across industries they cannot robustly relate productivity growth to prominent reforms such as industrial de-licensing, tariff reductions, FDI liberalisation, or lifting of small-scale industry reservations. The authors discuss several alternative explanations for this lack of a smoking gun, including lags (on lags, see the work of Geng 2010, in particular), previous infant-industry-type policies, investments in intangible capital, non-uniform impacts of reforms across industries, and so on. It is also possible that some firms or plants were better able to use or acquire the skills or technology needed to benefit from the reforms. This last conjecture of the authors is most consistent with the story we tell in our work.

Yet another analysis largely based on ASI data is that of Ghani et al. (2011). They focus on new business creation and employment. They point out that self-employment alone is not enough for new business creation to have significant employment effects. In India, much of the self-employment is in the so-called informal sector, and does little to generate new jobs. Instead, it is new formal firms that make a difference. Ghani et al.show that India’s density of new business registration is below average, conditional on per capita GDP. So India lags in the kind of entrepreneurship that could help growth.

Aggregating the ASI data to the level of the states reveals a striking pattern. Those states which had higher levels of new business creation in 1989 had faster manufacturing employment growth over the subsequent 16-year period, 1989-2005. The major state with the worst performance on both dimensions was West Bengal. Among the states with the best record was Tamil Nadu. A partial exception to an otherwise strong relationship between business creation and subsequent employment growth was Uttar Pradesh, which did well in setting up new plants, but lagged badly in job creation. Ghani et al. show that the same pattern is evident, perhaps even more strongly, when one disaggregates to region-industry clusters. It could be that the same factors are driving both patterns. State-level policies that encourage entrepreneurship may also make it easier for new firms to grow and hire more workers. State-level policies may encourage the creation of regional industry clusters, which attract new firms and make it easier for those firms to grow.

There are other initial conditions that matter, beyond ones that might plausibly be conjectured to arise directly from governance or policy. Certain industries are more likely to grow faster, and attract entry. But even controlling for such factors, the pattern of new business creation leading to subsequent periods of higher employment growth is still evident. Other channels of influence also emerge from Ghani et al.’s analysis of the data. The rate of entrepreneurship in organised manufacturing is positively affected by the share of population with a graduate education, and by closeness to a city. The strength of local supplier networks also is a plus for setting up new establishments. On the other hand, the stringency of local labour laws has a negative effect on entrepreneurship. This last result is reminiscent of Besley and Burgess (2004), but the latter was based on state-level data, rather than the plant-level data of the newer analysis.

Policy in need of focus

There are many more empirical studies of Indian manufacturing firms, but my sense is that we do not yet have a robust integrated understanding of what drives their performance. And this translates into a lack of clear policy guidance. My sense is that India’s new National Manufacturing Policy does not have a clear focus. I would like to see that emerge from a synthesis of existing work on Indian manufacturing firms, and a resolution of our ignorance in areas such as the basic determinants of performance and the drivers of industrial growth in India.

Further reading

  • Besley, Timothy and Robin Burgess (2004) “Can labour regulation hinder economic performance? Evidence from India”, Quarterly Journal of Economics, 119(1):91-134
  • Bloom, Nicholas, Benn Eifert Aprajit Mahajan, David McKenzie, and John Roberts, (2012), “Does Management Matter?”, Policy Research Working Paper 5573, The World Bank
  • Bloom, Nicholas, and John Van Reenen (2010), “Why Do Management Practices Differ Across Firms and Countries?”, Journal of Economic Perspectives, 203-224
  • Bollard, Albert, Peter J Klenow, and Gunjan Sharma (2012), “India’s Mysterious Manufacturing Miracle”, Stanford University Working Paper, June
  • Gangopadhyay, Shubhashis, Manisha G Singh, and Nirvikar Singh (2008), Waiting To Connect: Indian IT Revolution Bypasses The Domestic Industry, Lexis-Nexis-Butterworth.
  • Geng,Nan (2010), “Adjustment of Indian Manufacturing Firms to Pro-Market Economic Liberalizing Reforms, 1988-2006: A Time-Varying Panel Smooth Transition Regression (TV-PSTR) Approach”, Santa Cruz Institute for International Economics, Working Paper, #10-08
  • Ghani, Ejaz, William R Kerr, andStephen D O´Connell (2011), “Spatial Determinants of Entrepreneurship in India”, NBER Working Paper No. w17514, October
  • Sharma, Shruti, (2012), “Do Firms Skill Upgrade with a Decline in Input Tariffs? A Psuedo-Panel Analysis”, UCSC working paper in progress.
  • Sharma, Shruti and Nirvikar Singh (2012), “IT Investment and Productivity in Indian Manufacturing”, paper prepared for India Policy Forum Conference, July.
No comments yet
Join the conversation
Captcha Captcha Reload

Comments will be held for moderation. Your contact information will not be made public.

Related content

Sign up to our newsletter