India's manufacturing under-performance: A new clue from ‘multi-plants'

In Tuesday's piece, we highlighted the rise of the multi-plant phenomenon and showed that when account is taken of it, large plants in India have not grown over time despite a number of favourable developments such as policy liberalisation and contractualisation of labour. We also documented their lower productivity, which could explain in part why Indian firms did not become competitive internationally, especially in labour-intensive sectors.

- ABHISHEK ANAND, ARVIND SUBRAMANIAN & NAVEEN THOMAS

The next obvious question that arises is this: Why do we see the proliferation of multi-plants in the first place? One explanation is that land is difficult to acquire in contiguous parcels to sustain large plants. Another is that smaller plants can be closer to geographically dispersed labour pools, especially of women. The intriguing third possibility we explore in our recent paper is that firms proliferate plants in order that each plant can remain small. In economics jargon, the extensive margin increases so that the intensive margin does not have to.

A few pieces of evidence are instructive. Rising contractualisation of labour - from about 22 to 41 per cent over the first two decades of this century - has been an important response of Indian firms and their management to the regulatory environment. Firms such as TeamLease act as brokers, taking upon themselves the burden of complying with labour laws so that manufacturing firms themselves do not have to. But somewhat puzzlingly, we find that contractualisation is lower in labour-intensive industries than in non-labour intensive ones. This is puzzling because if contractualisation is a response to labour laws and their burdens, the incentives to do so should be greater in labour-intensive industries. For example, in labour-intensive industries, the share has risen from about 23 per cent to 31 per cent over two decades, but in other industries from 19 per cent to 47 per cent.

Unpacking this further, we find that recourse to contractual labour is greater in single plants than multi-plants. We also find that at the margin, the incentive to substitute contract labour for full-time employees rises with employment size in single plants but does not do so uniformly for multi-plant units. Both of these findings are illustrated in the accompanying graphics.