Previous Table of Content Next
4

Court Efficiency and Business Outcomes: Measuring the Performance of District Courts

Manaswini Rao

Introduction

Theoretical literature in economics has emphasised the role of institutions in promoting economic development.1 As a society develops and becomes complex, informal institutions start playing a complementary role to formal institutions that become more central. Codified rules of governance and transactions become important as the relationship between societies and organisations increase. The judiciary, in its role as an enforcer of rights and contracts, is responsible for maintaining the rule of law. Presence of rights—including property rights— frees up one’s own resources from defending their interests, which can be better allocated or invested in making improvements to their production processes. For example, having strong property rights reduces the time devoted to guard labour and increases employment (and thereby welfare) as found by Field2 in her study of changes in land titling laws for slum dwellers in Peru. Second, as an enforcer of contracts, an efficient judiciary is likely to be more important for formalisation of industrial and service sector activities in large economies, particularly leading to specialisation and driving innovation. Consequently, judicial inefficiency may force production into the informal sector or give rise to intermediaries driving up transaction costs.3

Earlier empirical work used cross-country variation in institutional structure, including the judiciary4,5,6, to identify the relationship between institutions and economic development. Over time, researchers have used increasingly available subnational data7,8,9,10, to examine the effects on specific markets, mainly the credit markets. These papers find that well-functioning judicial institutions increase credit availability at lower rates on an average but can also lead to distributional concerns in the presence of credit supply constraints where these benefits are only enjoyed by large businesses.

In this chapter, I examine the nature of judicial institutions in India—specifically, district courts— to uncover their functioning at the smallest level that has a plausible bearing on the functioning of businesses. I use the detailed and novel data from the e-courts system across 195 District and Sessions Courts to construct measures of court performance at the district level for each year between 2010 and 2018. Figures 2.4.1 and 2.4.2 shows the availability of data through histograms. A sharp increase in the number of cases filed and disposed starting from 2010 can be seen, although there are some variations between the states. For example, Karnataka and Maharashtra have data on cases filed and disposed in the years before 2010, whereas for many others, these start from 2010.

figure 2.4.1.Data Availability (By Year of Filing)

figure 2.4.2.Data Availability (By Year of Disposal)

Using this data, I present the statistical relationship between court performance and business outcomes. While there has been considerable attention given to the issue of large pendency in Indian courts in policy and media debates, there has been little exploration and discussion on the consequences of pendency. DAKSH’s ‘Access to Justice Survey’ (2016)11 reports that substantial costs are borne by private individual litigants—around 􀚒 500 per day on travel to courts and 􀚒 850 to 900 in the form of forgone wages. The cost to businesses and economy, on the other hand, is hard to estimate and could be even larger. This estimation is hard because the channels linking court performance and business outcomes could be both direct (for litigating companies) as well as indirect, in the form of overall business environment, trust in institutions, and belief in the enforcement of rights and contracts.

I estimate this relationship by examining the business outcomes—namely, annual revenue from sales, annual wage bill, annual profits adjusted for inflation, and annual legal expenses—of all companies incorporated before 2010,12 and registered in the same location as the district court. Prowess dx academic database published by the Centre for Monitoring Indian Economy (referred to as ‘Prowess’ hereafter), covers annual financial and other performance data on 49,200 companies registered under the Companies Act, 1956. While companies may operate in many more locations than their registered headquarters, the data allows me to only match at this level. Considering that the distribution of all businesses in India leans heavily towards the small firms13,14 operating mainly in one location, this strategy seems reasonable. Therefore, my dataset allows me to comment on the combined (both direct and indirect) relationship between court performance on the sample of businesses with the same registered location.

briefly summarise the findings before discussing them in detail below. First, court performance is not a monolith but comprises many dimensions, of which I am only able to present a few. These are average case duration—which is the life span of a case at the time of disposal; court level disposal rate—which is the ratio of cases disposed relative to active case load (pending cases + new institution); and dismissal rate—which is the proportion of cases disposed that are dismissed. Future research can focus on creating an index of court performance that includes these different components with appropriate weights. Second, I show that there exist patterns of lagged association between business outcomes and court performance; that is, outcomes move one to two years post the movement observed in court performance. Since I include all companies co-located as the court district and not just the litigating ones in my analysis, the lag is understandable, similar to lagged response to macro-economic variables. I break the patterns down by smaller geographic units and show that they generally hold, except in some cases. Finally, I add caveats to the findings, that they should not be construed as causal, since more research is required to establish the causal link. Having said so, the patterns do speak of the importance of district court performance on business outcomes. Given the heterogeneous associations across states and districts, policy response should be tailored to the local context, beginning with experimentation and pilots before embarking on large-scale reforms.

Context and Data

Judiciary in India is a three-tier unitary system, with the Supreme Court at the apex followed by High Courts at the state level and finally first instance courts at the level of a district and below. My research question concerns the performance of the District and Sessions courts, which are typically the first point of contact for filing cases involving firms, determined by monetary and geographic jurisdiction of the case.

Well-functioning judicial institutions feature as one of 10 key indices in the ‘Doing Business’ indicator developed by the World Bank, where each index pertains to a specific feature of an economy important for the functioning of businesses. These include aspects of the regulatory environment and the functioning of the bureaucracy, including procedures and costs involved in starting a new business15 as well as measures of judicial quality, such as contract enforcement, resolving insolvency, and protecting minority stakeholder interests. One can also argue that a well-functioning judiciary also has indirect effects—for example, through credit markets that require effective contract enforcement to address the twin problems of adverse selection and moral hazard.

India has consistently ranked low in overall ranking as well as ranking within the specific area of contract enforcement. Even as the overall ranking has miraculously jumped from 142 in 2014 to 77 in 2018, the ranking under contract enforcement continues to remain poor. India was ranked 163 in 2018, only better than Sri Lanka, Afghanistan, and Bangladesh in the region.16

I construct the dataset on the court functioning by scraping publicly available case records from 195 administrative districts (see Table 2.4.1) on the e-courts system,17 detailing case level metadata as well as proceedings from each hearing.18

Table 2.4.1.Study E-Courts Sample District Coverage

From the case level dataset, I build measures of court performance, namely case duration, and disposal rate. Case duration is measured as the average life of a case by year of disposal, that is, the average life span of cases that are disposed in a given year. Disposal rate is the ratio of number of cases disposed relative to total active cases (pending and newly instituted) in a given year. One could construct additional court speed measures, but I restrict my attention to these two in this chapter. Additionally, I present a plausible measure of court quality—the proportion of cases disposed as a result of being dismissed on account of procedural or substantive reasons. This measure is constructed using the details contained in the variable—‘nature of disposal’—as recorded on e-courts. A case can be disposed on many accounts, including final judgment upon the completion of trial or dismissal at any stage prior to the completion of full trial typically on procedural grounds, such as lack of evidence, absence of either of the parties for an extended period, or conciliation between the parties, etc. Therefore, I view dismissals as an indicator of plausible improved quality, especially in instances of false claims or out-of-court settlements.

For business outcomes, the Prowess dataset represents over 60 per cent of the economic activity in the organised sector in India, which although a small subset of all industrial activity, accounts for about 75 per cent of corporate taxes and 95 per cent of excise duty collected by the Government of India.19 I focus on business performance variables that include annual revenue from sale, wage bill, legal charges, and profits adjusted for inflation. The choice of these variables is driven by theoretical models linking institutional environment and firm performance built on Besley and Persson20 and Lilienfeld-Toal et al. (2012). The basic idea stems from the view that most businesses are either revenue or profit optimisers. They choose capital and labour as inputs subject to constraints arising out of various markets, institutional, and informational asymmetries. Court performance is modelled as an institutional constraint that limits their ability to utilise optimal levels of inputs for production relative to the unconstrained situation. Therefore, I focus on these measures in this chapter, in addition to legal charges, which is a measure of the direct costs borne by companies to address the institutional constraints including fees paid to legal advisors and law firms.

I exclude some large urban agglomerations, including Delhi and Mumbai, from the sampling to isolate the influence of agglomerative forces that could plausibly affect court capacity in the first place. This yields 13,298 companies in the Prowess database with registered offices in the study’s district sample, described in Table 2.4.2 below. The table presents the mean and standard deviation21 of the variable listed on the left-hand side across the sample of districts in my study.

A large share of the companies are privately owned (75 per cent), publicly listed (64 per cent), and in non-finance sector (79 per cent). Among sectors represented, 43 per cent are in the manufacturing sector, 30 per cent in business services (of which a large share is financial service companies), and 15 per cent in trade, transport, and logistics sector.

In the next section, I present the different court performance measures as well as look at the association between court performance and business outcomes over time.

Table 2.4.2.Distribution of Companies in Study Sample

Court Performance Measures

Case duration (blue dots in Figure 2.4.3) varies between 750 and 1,100 days across the entire sample exhibiting a drop in 2015, and a spike in 2017, followed by a drop again, as shown in Figure 2.4.3. This measure varies substantially between states, with a few hundred days in Haryana, Punjab, Karnataka, to over 1,000 days in West Bengal and Odisha, while between 500–1,000 days in the remaining states. Disposal rate (red dot) remains fairly stable at 15 per cent, whereas the share of case dismissal (green dot) varies between 35 and 50 per cent. However, there is a wide variation in these measures across the states. Disposal rates are consistently low at around 20 per cent of the active case load that is gradually, but slowly, increasing over the years. Some exceptions are the spikes we notice in Punjab (2012), Tamil Nadu (2012), Haryana (2013), Puducherry (2015), and West Bengal (2017). On the other hand, share of disposed cases that are dismissed vary widely between the states. West Bengal tops this metric, dismissing close to 70 per cent of the cases it disposes in a given year. However, when viewed along with the case duration metric, it appears as though the dismissals come much later during the case life cycle than earlier on. That is, while the dismissal rate is close to 70 per cent, the duration of cases at the time of their disposal vary between 1,000–2,000 days in West Bengal. On the other hand, Tamil Nadu and Punjab have a relatively high dismissal rate (at about 50 per cent) and when viewed along with case duration at the time of disposal that is under 500 days on an average, it looks like that dismissals are made earlier in the case life cycle. Another interesting juxtaposition is comparing disposal rates with case duration at the time of disposal in a year. Cases in both Odisha and West Bengal tend to have a long life, while disposal rates are low at around 20 percent. While many other states have low disposal rates as well, the duration of cases at disposal are much lower in other states relative to Odisha and West Bengal. One implication of this could be that judges may be prioritising relatively older cases for disposal in some states.

figure 2.4.3.Court Performance Measures

Key Take-away

Court performance has multiple dimensions, which could and should be measured. Juxtaposition of these measures throws some light on the decision-making processes on how cases are resolved across different courts and over time. Some of these measures may have more easily detectable associations with socioeconomic outcomes of interest whereas others may suffer from a lack of sufficient statistical power to clearly identify the associations.

I discuss the association between case duration, that is, the average life-span of cases by their year of disposal and business outcomes in further sections.

Associations between Court Performance and Business Outcomes

I transform all business outcome variables, except profits, into their logarithmic equivalent so that I can interpret the outcome in terms of year-onyear percentage change. Under the logarithmic transformation, a simple difference between the values across consecutive years gives us the change in the underlying variable in proportion to its initial value. For example, looking at the top-left panel of Figure 2.4.4, we notice that log sales revenue is about 5.2 log units in 2010 and about 5.3 log units in 2011. This implies that the underlying variables—sales revenue—grows 10 per cent (5.3 − 5.2 = 0.1, or 10 per cent when measured as percent change) between 2010 and 2011. However, because profits also take on negative values, I cannot use logarithmic transformation and, therefore, present them in their raw form after adjusting for inflation. All raw outcome measures are reported in INR million, adjusted for inflation.

figure 2.4.4.Business Outcomes and Case Duration: All Sample

Figure 2.4.4 depicts a lagged association between business outcomes and court performances. That is an improvement, and a reduction in case duration between 2014 and 2016 is associated with an improvement in the business outcomes one to two years later (an inflection in 2017).

Legal charges grow with time, but show lagged inflections associated with changes in case duration. It is important to bear in mind is that such legal charges are incurred by litigants not just in the form of court fees, travel and board, but also in the form of lawyers’ fees, and other preventive and compliance measures that are dictated by larger macro-economic and policy conditions.

Next, I examine these patterns by state and by district (of a specific state, Maharashtra, for the purposes of illustration22) to explore any heterogeneous patterns in this association.

figure 2.4.5.Business Outcomes-I and Case Duration: By States and Districts

Sales revenue does not vary substantially between states and shows a modest growth during the period of analysis (with the exception of Chhattisgarh and Telangana). On the other hand, there is greater variation between districts within a state as well as over time. Focusing on the association between business outcomes and court measures between districts, it can be seen that the growth in sales revenue is negatively associated with case duration (see Osmanabad and Latur, for a clear example).

A similar pattern can be observed between states as well, although due to the issue of scale for representation, these variations aren’t very obvious.

Wage bills also show a similar pattern. As the case duration increases, the companies appear to cut back on wage expenditure, which could arise either from stalling recruitment or stalling wage increases. Unfortunately, the data presents only wage expenditure and not number of workers/ employees over time to uncover the mechanisms behind changes in wage expenditure.

figure 2.4.6.Business Outcomes-II and Case Duration: By States and Districts

Legal charges show more variation between and within states but do follow an overall increasing trend. In some states and districts, this matches with an increased trend in case duration (for example, Chhattisgarh, Andhra Pradesh among states, and Nashik and Osmanabad among districts) whereas in others, it is associated with decreasing case duration (for example, Gujarat among states, and Ahmednagar, Chandrapur, and Solapur). A positive correlation between case duration and legal charges is understandable, as the longer a case takes, the more a company has to spend on retaining lawyers, spending on travelling to court, and associated fees. On the other hand, a negative correlation—declining legal charges with increasing duration or vice-versa—is confusing, but perhaps can be rationalised if higher legal charges are mainly incurred from preventing litigation and compliance purposes in areas with low case duration or if lower legal charges in areas with higher case durations can be rationalised with larger periods of non-activity between hearings that are spread out.

Real profits are generally negatively correlated with case duration, for example, in Odisha, Telangana, and Chhattisgarh among states, and Ahmednagar and Solapur among districts. On the other hand, Chandrapur is an exception where real profits fall with a decrease in case duration. Perhaps, this could be due to unobserved idiosyncratic factors.

Share of case dismissals and disposal rate show a similar pattern (see Appendix) but the latter is relatively weaker. As mentioned earlier, the court performance variables need to exhibit substantial variation between the sample units to have sufficient statistical power to identify these types of associations.

Caveat

While the approach mentioned earlier provides a good measure of the association between court performance and business outcomes, it is possible that unobserved district-level dynamic variables could be driving both business outcomes and court performance. For example, industry or sectoral conditions or migration patterns that evolve differently across districts over time, could affect business outcomes as well as court performance (for example, influx of migrant labour). There is also the possibility of simultaneous causation if firms in slow court districts are also more likely to litigate.23 I try to address these concerns mainly by not including highly urbanised districts from my analysis and retaining focus on rural districts. However, these challenges do present avenue for future research.

Conclusion

To summarise, this chapter presents different ways of measuring court performance and urges policymakers and practitioners to focus on more aspects of performance than pendency alone. Second, I show the patterns of association between court measures and business outcomes—sales revenue, wage bill, real profits, and legal charges. The main business outcomes—sales revenue, wage bills, and profits—are negatively associated with the average case duration at the time of disposal, lagged by 1–2 years, in the overall sample. Breaking this down by states and districts also reveals a similar patterns subject to some idiosyncratic variations within the geographic unit. While I present some caveats to interpreting these associations as the causal response of businesses to district court institutions, these patterns reveal that the institutional environment in the form of effective courts at the entry level is important for business performance.

Appendix

figure 2.4.7.Business Outcomes and Disposal Rate

figure 2.4.8.Business Outcomes-II and Disposal Rate

Notes

  1. O.E. Williamson. 1998. “Transaction Cost Economics: How it Works; Where it is Headed,” De Economist, 146(1): 23–58.
  2. E. Field. 2007. “Entitled to Work: Urban Property Rights and Labor Supply in Peru,” The Quarterly Journal of Economics, 122(4): 1561–1602.
  3. M. Fafchamps. 1996. “The Enforcement of Commercial Contracts in Ghana,” World Development, 24(3): 427–448.
  4. D. Acemoglu, S. Johnson, and J.A. Robinson. 2001. “The Colonial Origins of Comparative Development: An Empirical Investigation,” American Economic Review, 91(5): 1369–1401.
  5. D. Acemoglu and S. Johnson. 2005. “Unbundling Institutions,” Journal of Political Economy, 113(5): 949–995.
  6. S. Djankov, R. La Porta, F. Lopez-de Silanes, and A. Shleifer. 2003. “Courts,” The Quarterly Journal of Economics, 118(2): 453–517.
  7. S. Visaria. 2009. “Legal Reform and Loan Repayment: The Microeconomic Impact of Debt Recovery Tribunals in India,” American Economic Journal: Applied Economics, 1(3): 59–81.
  8. U.V. Lilienfeld Toal, D. Mookherjee, and S. Visaria. 2012. “The Distributive Impact of Reforms in Credit Enforcement: Evidence from Indian Debt Recovery Tribunals,” Econometrica, 80(2): 497–558.
  9. M. Chemin. 2012. “Does Court Speed Shape Economic Activity? Evidence from a Court Reform in India,” The Journal of Law, Economics, and Organization, 28(3): 460–485.
  10. J. Ponticelli and L.S. Alencar. 2016. “Court Enforcement, Bank Loans, and Firm Investment: Evidence from a Bankruptcy Reform in Brazil,” The Quarterly Journal of Economics, 131(3): 1365–1413.
  11. Harish Narasappa, Kavya Murthy, Surya Prakash B.S., and Yashas C. Gowda. 2016. “Access to Justice Survey: Introduction, Methodology, and Findings,” in Harish Narasappa and Shruti Vidyasagar (eds), State of the Indian Judiciary: A Report by DAKSH, pp. 137–155. Available online at dakshindia.org (accessed on 11 October 2017).
  12. I do this in order to study the association with existing businesses. On the other hand, the institutional environment could also influence the start and exit of companies, which I haven’t discussed here.
  13. C.-T. Hsieh and P.J. Klenow. 2009. ‘Misallocation and Manufacturing TFP in China and India’, The Quarterly Journal of Economics, 124(4): 1403–1448.
  14. C.-T. Hsieh and P.J. Klenow. 2014. ‘The Life Cycle of Plants in India and Mexico’, The Quarterly Journal of Economics, 129(3): 1035–1084.
  15. Property registration, construction permits, access to electricity, etc.
  16. The World Bank. 2018. Doing Business 2018: Reforming to Create Jobs. Washington, D.C.: The World Bank.
  17. This data has been made available for public use since late 2014 through web portals such as www.ecourts.gov.in and https://njdg.ecourts.gov.in.
  18. These include date of filing, registration, first hearing, decision date if disposed, nature of disposal, time between hearings, time taken for transition between case stages, litigant characteristics, case issue, among other details.
  19. P.K. Goldberg, A.K. Khandelwal, N. Pavcnik, and P. Topalova. 2010. ‘Imported Intermediate Inputs and Domestic Product Growth: Evidence from India’, The Quarterly Journal of Economics, 125(4), 1727–1767.
  20. T. Besley and T. Persson. 2009. ‘The Origins of State Capacity: Property Rights, Taxation, and Politics’, American Economic Review, 99(4): 1218–1244.
  21. Standard deviation is a measure of variance, which when read with the mean (or the statistical average) provides a measure of spread of the corresponding variable. For example, while 1,854 is the average number of companies per district in my sample, some districts have more than this number while some other districts have less than this number. A standard deviation of approximately 1,947 companies implies that there is a large spread in the number of companies registered within a district in my sample. This is not surprising since industrialised districts like Ahmednagar (MH) will have more companies compared to less industrialised districts like Chandrapur (MH).
  22. It is hard to present the patterns across 195 district courts graphically and so, I present cross-state comparison and cross-district comparison of one specific state, Maharashtra.
  23. For example, an improvement in overall economic environment would lead to an increase in the number of economic transactions which could mechanically increase the number of disputes and litigation in courts.