(Co-authored with Rajesh Raj S.N. and Kunal Sen)
The possibilities of a global economic reorganisation in a post-Covid-19 world has brought the attention of the Indian government back to their flagship ‘Make in India’ programme. Speaking to his council of ministers during the lockdown, Prime Minister Modi has asserted1 that the crisis was, in fact, an opportunity to boost this initiative. The Make in India programme, which was initiated by the Modi government in 2014, is an attempt to increase the ‘Ease of Doing Business’ in India by clearing up regulatory red-tape, and putting in place new business-friendly rules that are consistent with the World Bank’s Doing Business (DB) Reports. The programme has, however, achieved very limited success till now. In recent research (Raj et al. 2020), we highlight two important factors behind this outcome. First, using a framework developed by Hallward-Driemeier and Pritchett (2015), we demonstrate that the de jure (contractual) rules and regulations adopted by the government do not provide a realistic description of the business environment in India. Second, policy decisions made by the national government are not necessarily implemented with the same zeal in the Indian states, where most of the implementation of policies take place. We show how these factors have undermined this programme in the sections below.
Doing business in Indian states
Using Enterprise Survey data, we present the state-wise kernel density plots2 of days needed to get an operating licence in India (Figure 1). The vertical line in these figures represents the de jure DB values in these states. The data show that the median time taken to set up a business in India was 24 days, with a wide variation across states. The median time that it took to set up a business in Orissa and Andhra Pradesh were 1 and 3 days, respectively, whereas the same for Tamil Nadu and Punjab was 40 and 38 days, respectively. Moreover, we see remarkably wide within-state variation in the time it takes to get an operating licence. For example, in Bihar, the 10th percentile set of firms report obtaining an operating licence in one day, while the 90th percentile set of firms report obtaining an operating licence in 90 days. It is also very evident from Figure 1 that for a very large number of firms, the waiting time to get an operating licence is considerably lower than what the state-level DB indicators suggest.
Figure 1. Kernel density plots, operating licence
Notes: (i) The vertical line in the plots represent the de jure DB values of time needed for actual operation of the standardised business. As vertical line exceeds the sample observations for Orissa and Andhra Pradesh, vertical line is not presented for both states. As regards Gujarat, there are no observations on operating licence in the dataset, ii) for an explanation of kernels and bandwidth, see Footnote 2.
(ii) AP – Andhra Pradesh, UP – Uttar Pradesh, MP – Madhya Pradesh, TN – Tamil Nadu, WB – West Bengal.
How do we interpret these findings? Following Hallward-Driemeier and Pritchett (2015), we distinguish between ‘rule’-based outcomes, which would be consistent with the DB indicators, and outcomes based on ‘deals’ between businesses and the State, which are inconsistent with these ‘rules’. Deals are different from rule-based outcomes in two important ways. First, rules are impersonal interactions between the State and businesses, whereas deals are based on a personalised relationship between businesses and political leaders or the bureaucracy. Second, rules are usually the same for all relevant businesses whereas deals differ from case to case, depending on the nature of the personalised relationship. Clearly, the wide variation in the number of days needed to get an operating licence, and the lack of any relationship with the DB indicators, strongly suggest that de facto deals that are struck by firms with the State, rather than de jure regulatory rules, characterise the business-State relationship in Indian states.
Next, we characterise the quality of the environment for deals in each state in terms of how quickly firms in that state can obtain an operating licence3 . We define three kinds of deals – namely good, moderate, and bad deals – where a good deal is the case where firms get their operating licence within 15 days. Similarly, if firms get their licence between 15 and 45 days, we define it as a moderate deal. Finally, if it takes more than 45 days for firms to obtain the licence, we define it as a bad deal. As is clear from Figure 2, there is wide variation in deal-making across states in India. For example, a state like Uttar Pradesh has a high degree of good deals (100%) as compared to bad deals (0%), while a state like Kerala has a low degree of good deals (33%) as compared to bad deals (48%).
Figure 2. Types of deals, by state
Note: Good deals: number of days to get an operating licence (ND) <=15; moderate deals: 15<ND<=45; and bad deals: 45<ND.
Does better governance ensure better deals?
What explains this systematic variation in deal-making across Indian states? We focus on the role of governance, which is widely seen as a key determinant of the business climate in the developing world. There are two possible ways in which governance may impact deal-making. One way could be that states with higher administrative capacity are able to shorten the time it takes to issue an operating licence or construction permit, due to more efficient approval or implementation procedures. This would be a case of stronger states being able to administer their policies for regulatory approval in a much more efficient manner. A second way could be that better deals are observed in states with weak governance as firms in these states have developed collusive relationships with the bureaucrats responsible for investment approvals. This could a case of capture of weak bureaucracies by more powerful business actors.
In order to test these alternate hypotheses, we examine the relationship between the quality of the deal environment with alternative measures of the quality of governance. We find that firms in better-governed states experience longer delays in obtaining operating licences. Our results do not support the general perception that better governance leads to good deal-making in India. Rather, they indicate that good deal-making is an outcome of high levels of state capture and administrative corruption, and that good deals are more prevalent in an environment of weak governance. In other words, good deals come through where bureaucrats and politicians are willing to engage unofficially with the private sector, and are possibly the result of a weak or corrupt state administration.
Do better deals ensure more productivity?
Our findings challenge the conventional wisdom that good deals are reaped by firms operating in better-governed states. It may be argued, however, that if it is the more productive firms who are getting their operating licences or construction permits within a shorter time due to this governance failure, then such deal-making is not necessarily bad, as it has positive implications for productivity and growth. Is there any proof that this is indeed what is happening? To understand this, we need to look at the relationship between good deals and the performance of firms. Specifically, are firms that are successful in cornering the better deals also the ones that exhibit higher productivity?
We test this hypothesis by examining the relationship between labour productivity of firms and the quality of deal-making at the state level. We also look at the role of the quality of governance on deal-making – both on its own and in conjunction with the quality of deals. In our analysis, we take into account the size of firm, and the industry in which the firm is located. We find that good deals are associated with better firm performance. We also find that the positive association of good deals with firm performance tend to diminish as the quality of governance increases. Furthermore, a simple calculation based on our results shows us that good deals lead to lower productivity for most states in India. Thus, we can conclude from our results that for most states in India, better deals lead to lower productivity.
Our paper shows that de facto deals rather than de jure rules characterise the business-state relationship in Indian states, with the actual days needed to get an operating licence far lower for many firms than what is given by the de jure rules and regulations. Furthermore, better deals environments are proportionately more observed in states with weak capacity, suggesting that most state governments – particularly those that have poor governance – are captured by business interests. Finally, better deals go to the less-productive firms in most states, suggesting that good deals are not necessarily growth-enhancing. This jeopardises the healthy growth in the Indian manufacturing scenario as the most unproductive firms are able to undercut and outcompete more productive firms by manipulating the regulatory environment. Moreover, such regulatory capture also creates structural disincentives for improving the governance capability of these state governments. This perpetuates a vicious cycle of poor governance in Indian states and unproductive growth in the Indian business sector. Thus, our results suggest that as long as the business environment in India is characterised by these types of deals, the institutional reforms initiated by the government in order to promote the ‘Make in India’ programme is unlikely to succeed.
Hallward-Driemeier, M and L Pritchett (2015), “How Business Is Done in the Developing World: Deals versus Rules”, Journal of Economic Perspectives, 29(3): 121-140.
Natarajan, Rajesh Raj, Kunal Sen and Sabyasachi Kar (2020), “Unmaking ‘Make in India’: Weak Governance, Good Deals, and their Economic Impact”, Economic and Political Weekly, 55(11):43-53.
World Bank (2017), ‘Doing Business 2018: Reforming to Create Jobs’, World Bank Report, 31 October 2017, Washington, DC.
World Bank (2016), ‘Doing Business 2017: Equal Opportunity for All’, World Bank Report, 25 October 2016, Washington, DC.
2. A kernel density plot is a visual representation of how the values of one variable (in our study, the proportion of firms) are distributed over a continuous scale measuring another variable (in our study, the number of days needed to get an operating license). This visual representation is based on an estimation exercise which can choose alternative smoothing functions (called kernel functions) and smoothing parameters (called bandwidths) that determine the size of the separate bins over which this estimation is carried out.
3. In the study, we also undertake a similar exercise using construction permits.
Sabyasachi Kar is Professor (RBI Chair), NIPFP, New Delhi, Rajesh Raj S. N. is Associate Professor, Sikkim University and Kunal Sen is Director, UNU-WIDER.
The views expressed in the post are those of the authors only. No responsibility for them should be attributed to NIPFP.
This article was first published in IGC Ideas for India, July 10, 2020.