Small is not beautiful: India's bias against large factories is damaging

India’s persistent underperformance in manufacturing, particularly labour-intensive manufacturing, is one of the major issues in Indian economics. In fact, it is not so much a puzzle as it is over-determined, with multiple possible and overlapping explanations. Inconsistent trade policy, poor gender relations, problematic law and order, and an ineffective educational and skilling system all play a role. But one major contributor, it has long been understood, is over-regulation, particularly of labour and employment. For at least two decades, the effect of regulations on firm size in India has been observed. Labour regulations increase unit costs by more than a third, and the incentives to stay below certain thresholds to avoid intrusive regulation are high. The consequence of this is a skewed distribution of factories. The Annual Survey of Industries has long been quoted in this context, showing that micro, small, and medium enterprises make up over 96 per cent of Indian factories, and micro-enterprises with less than 10 employees are 99 per cent of those in turn. Factories with over 2,000 workers are a tiny fraction of the total.

Recent work by economist Arvind Subramanian and others, recently summarised in this newspaper last week, shows the problem is even worse than the ASI data suggests. In fact, the number of large factories has been overstated by the ASI, not understated. Many companies, it turns out, create multiple production units within a single state. These, while effectively separate factories owned by a single company, can be reported to the ASI as a single factory, introducing errors into the data. The prevalence of these “multi-plants”, the authors show, gives India’s largest manufacturing plants the appearance of being bigger than they are in reality. Larger plants are more productive for multiple reasons. The authors estimate multi-plants’ productivity is about 5 per cent lower than that of single plants. Given the low margins that prevail in most labour-intensive sectors, a productivity differential of 5 per cent can be damning. There are also other problems with smaller plants, and they need to be examined in detail. For example, larger plants are able to take on the bulk orders associated with global fashion brands. Capital investment in a larger plant can also be more productive, and they can thus move quicker up the value chain.

Firms’ choice to build “multi-plants” is related to regulatory and political risks, which keep them small. Smaller plants provide more flexibility to companies dealing with difficult labour regimes. More general political risks, such as those that come from vexatious or malicious litigation, can also be hedged against by having multiple plants rather than one. The presence of multi-plants must be understood as another equilibrium response to a state that simply penalises any attempt to build scale. But without scale, India’s attempts to create manufacturing jobs are doomed to fail. It has long been understood that India’s small-scale bias is a problem for growth in manufacturing. It is now known that the problem is in fact worse than previously thought. Its competitor nations — Bangladesh and Vietnam, for example — have significantly larger median plant sizes in labour-intensive industries. Deregulation and the reduction of political risk have long been on the policy agenda. But efforts so far have clearly not been enough. Starting with truly effective, national, and flexible labour codes, the government must work on deregulation with a will.