Sunday, October 25, 2009

The Civil-Criminal Ratio Puzzle

Commentators have often noted that the High Courts, District Courts, and other subordinate courts are rarely systematically studied in India. In such a large country, it's difficult to easily make many statements about what is going on in these lower courts. One way to attempt to (relatively easily) study these courts is to look at the statistical data available about them. The Supreme Court issues quarterly reports about the pendency, institution, and disposal numbers in these courts (which is forwarded to them from the High Courts). The data is fairly basic, but it does separate out civil and criminal cases from each other.

Chief Justice Balakrishnan commented yesterday at a national consultation for reducing pendency and delays that:
“the litigation rates in the various States do not bear a consistent correlation with their respective populations. This means that in some States, a large proportion of the population has been approaching the courts as compared to other States. What is especially worrying is the immense disparity between the number of civil and criminal cases instituted in backward and insurgency-hit areas.

A perusal of the pendency figures indicates that while there are more civil cases filed in developed areas, the reliance on the civil justice system is shockingly low in Bihar, Jharkhand, Chhattisgarh, Jammu and Kashmir as well as the northeastern States. This disturbing trend could have two explanations — one, that the number of courts is grossly inadequate, and secondly, ordinary citizens are consciously not bringing their civil disputes before the judicial system. If the second of these explanations holds good, then it indeed calls for targeted interventions.”

I agree with the Chief Justice that there are striking differences between civil/criminal pendency figures across states in the country and the rates that the judiciary is being accessed by the population. I also think he is generally right that filing rates and low civil/criminal filing ratios are correlated to backward and insurgency areas, but there are outliers that deserve further investigation. Indeed, the picture is a lot more murky when one looks at the statistics, which indicates either the statistics are inaccurate or we need to have more detailed studies and theorizing about this issue.

Let's take up the question of filing rates to population first. In India on average in 2008 for every 616 people one case is filed in a High Court. The five best state ratios in 2008 (if you think more litigation is a good thing) are Himachal Pradesh (239); Madras (258); Kerala (388); Orissa (431); and Jammu and Kashmir (471); the five worst are Assam (1300); Chattisgarh (1198); Andhra Pradesh (1127); Jharkhand (1032); and Calcutta (908). Orissa and Jammu and Kashmir being in the top five might be a surprise and Calcutta and AP in the bottom five as well, but more or less the statistics seem to show some correlation between development and filing rates at the High Court level.

If we look at the district and subordinate courts amongst the best are Delhi (27); Kerala (30); Himachal Pradesh (39); Madras (40); and Gujarat (41). Amongst the worst are Bihar (245); Jharkhand (220); Orissa (134); AP (114); and Chattisgarh (94). Again, the correlation seems to hold up and we might assume that if we average out a few years the correlation might even get stronger and potential anomolies (Orissa being in the best for High courts and worse for subordinate courts) might even out. Generally, the states with poor filing rates also have poor judge to population and judge to filing ratios, so one could argue there is a supply side deficit that's effecting the demand. These results are somewhat interesting, but not particularly surprising, although if certain anomolies (like Orissa) remained that could call for further study.

Now let's turn to civil and criminal ratios. Nationally the civil/criminal ratio in H. Ct.'s is about 66%/34% in 2008. The CJI suggests that if the civil rate goes too low that might be an indicator of broader trouble with that court and citizens are deciding not to turn to it for civil redressal. For example, in Bihar the civil/criminal ratio is 32/68 for the High Court. Meaning 32% of the cases filed were civil, while 68% were criminal (this is a whopping 34% off the national average). Bihar has the worst ratio that year (again, if you think that a low civil filing rate compared to criminal is a bad thing). Following Bihar were Jharkhand (36/64) Punjab and Haryana (55/45), Gujarat (57/42) and UP (59/41). The "best" civil/criminal ratios were in Jammu and Kashmir (92/8), Himachal (86/14), Karnataka (85/15), AP (81/19), and Bombay (81/19).

In the subordinate courts amongst the worst were West Bengal (11/89), Jharkhand (14/86), Bihar (15/85), Chattisgarh (16/84), and Uttarkhand (16/84). Amongst the best were TN (51/49), AP (64/54), Karnataka (34/66), Himachal (33/67), and Delhi (25/75). Here the surprises would be finding Jammu and Kashmir as having the best ratio on the High Court side, and finding Punjab and Haryana and Gujarat amongst the worse ratios. In the subordinate and district courts having West Bengal have the worst ratio comes as a bit of a surprise, along with Uttarkhand being towards the bottom.

You will notice that there seems to be some correlation between having a high filing rate for ones population and a good civil/criminal ratio. AP is the outlier here having a poor filing rate, but a very healthy civil/criminal ratio.

Further, 3 out of 5 of the top and bottom scorers in filing rate in the High Court also scored in the top or bottom 5 in the subordinate courts. For the civil/criminal ratio the correlation was only 2 out of 5 of each the bottom and top bracket. Indeed, the civil/criminal ratio seems to be more murky overall, often following development indicators, but not as tightly as the filing rate. For example, Delhi and Chattisgarh High Court have about the same civil/criminal ratio (67/33), but per capita income in Delhi is about 4 times as much. Something else seems to be at work here - i.e. development is an ok indicator of a court's civil/criminal ratio, but there seems to be some other factor (or several other factors) as well.

Even if we accept development as the primary driver of the civil/criminal ratio it's not perfectly clear what about development drives the ratio. For example, is it that as the CJI suggests in under-developed states litigants aren't as likely to bring their civil suits to court (presumably because either they don't think the court can efficiently or fairly deal with them, or they can't afford the court)? Is it that there just aren't as many civil suits because there isn't much economic development? Is it that there is higher crime or more criminalization of activity by the state? Is it that development is correlated to a lack of judges and so it's a supply side issue? I would be possible to get a satisfactory answer with some more crunching and hopefully the proposed National Arrears Grid will be able to shed more light on these sorts of questions.

5 comments:

Kannan said...

I had done a simple analysis some months back here. I found what seemed to me to be a strong correlation between the rate of institution of civil cases in the district and subordinate courts in different states and the backlog of civil cases in those courts. The backlog was quantified by the number of years it would take to be cleared at the current case disposal rates, assuming no new institutions. There were two takeaways from the correlation. One - people were less inclined to file civil suits as the average time for disposal of cases increased. (This behavior seems eminently reasonable) Two - states in the eastern part of the country - Bihar, Orissa, W. Bengal, Assam, Jharkhand - showed very low institution rates. These states are considered to be lagging on various development indicators. This seems to bear out that "low civil litigation institution rate" is also an indicator of low development.

Nick Robinson said...

Thanks Kannan. I hadn't seen your article, but found it really helpful, a good read, and I think on point. You show there is a strong correlation between civil backlog and low civil filing rate. I just ran the numbers after reading your article and come to similar conclusions for 2008 (I'm still crunching the other years and trying to figure out other potential variables). However, you seem to jump with the CJI and conclude that people must be shunning litigation because of the backlog. This might very well be true (and has an intuitive appeal about it), but it could also be that these states are just poorer and badly run. The badly run part could explain the backlog and the poorer the lower filing rates. What would really prove the correlation I think is if there was a wealthy state that had a low filing rate and high backlog or a poor one with a high filing rate and low backlog. There isn't a clear case of this though so the correlation might not be as clear.

I just ran the numbers for the High Courts and the best civil institution states are Himachal Pradesh (277, backlog 1.9 years)), TN (347, 2.5 yr.'s), Kashmir (512, 2.7 years), Kerala(536, 1.3 yr.'s), and Delhi (557, 1.9 yr.'s). The worst are Jharkhand (2849, 3.3 yr.'s) Bihar (2364, 2.9) Assam (1901, 2.8 yr.'s), Chattisgarh (1785, 1.9 yr.'s), and Gujarat (1509, 1.5 yr.'s). As you can see the correlation doesn't seem as strong at the High Court level, although it's certainly still there.

The theory does seem to hold up more or less on the civil/criminal ratio. On the civil/criminal ratio of the ones I highlighted in post here are the year backlog stats for subordinate courts: worst - West Bengal (11/89, 4.3 yr.'s), Jharkhand (14/86, 2.9 yr.'s), Bihar (15/85, 5.4 yr.'s), Chattisgarh (16/84, 1.5 yr.'s), and Uttarkhand (16/84, 1.2 yr.'s). Best were TN (51/49, 0.8 yr.'s), AP (46/54, 1.5 yr.'s), Karnataka (34/66, 2 yr.'s), Himachal (33/67, 1.3 yr.'s), and Delhi (25/75, 2.6 yr.'s). There are definitely a few anomalies here, but it more or less holds.

Similarly, at the High Court level, worst were:
Bihar (32/68, 2.9 yr.'s), Jharkhand (36/64, 3.3 yr.'s) Punjab and Haryana (55/45, 3.8 yr.'s), Gujarat (57/42, 1.5 yr's) and UP (59/41, 6.9 yr.'s). The best civil/criminal ratios were in Jammu and Kashmir (92/8, 2.7 yr.'s), Himachal (86/14, 1.9 yr.'s), Karnataka (85/15, 1.4 yr.'s), AP (81/19, 3.5 yr.'s), and Bombay (81/19, 2 yr.'s).
(continued)

Nick Robinson said...

Backlog is clearly correlated to civil rates and ratios, but so is state's economic development and perhaps other factors (such as historical or cultural practice, etc.). I hope to run a regression analysis eventually and hopefully off a larger number of years to give weights to these factors. (we'll see how much time I can get though . . . )

If you bring in the criminal filing rate that brings in another interesting factor. For example, Jammu and Kashmir H Ct. has the lowest criminal filing rate in the country, which more than anything explains its really high civil/criminal ratio (i.e. this dynamic is probably being driven by the criminal and not the civil side). Similarly, the AP H Ct has the second lowest criminal filing rate, probably helping explain its top five status. Jharkhand and Bihar H. Ct.'s actually have about average criminal filing rates, so that means their poor civil/criminal ratio is probably being driven by the civil side. In the subordinate courts Bihar and Jharkhand have really low criminal filing rates, helping explain why their civil/crim ratio isn't as pronounced at the high court level as at the civil. Similarly, AP has a really low criminal filing rate in the subordinate courts perhaps helping explain why it's civil/criminal ratio looks so good (i.e. it's actually driven by a poor criminal filing rate). There actually is about an average backlog on the criminal side for AP, while the backlog is amongst the worst on the criminal side for Jharkhand and Bihar. However, I don't know if we can say the backlog is driving the low institution numbers for criminal cases (although it could be). At any rate, this all deserves a closer look - along more charts, and less words. This writing is fairly dense and I apologize to those who were looking for clearer writing in this post.

SRIDIP NAMBIAR said...

Is there any statistical data that shows the percentage of cases in which State is involved (on either side). Does the low institution of cases against the State in some states mean good governance (based on Kannan's logic).

Nick Robinson said...

Good question. I have never seen a number floated about what % of cases the government is involved in. Presumably, in criminal it is close to a 100%, but on the civil side all I've seen are comments by judges, litigators, etc. that it's "very high." I don't know what "very high" means though, or as you point out, it's unclear how you would even judge what's too high. I don't think the Court keeps track, but presumably the governments lawyers would know how many cases they are involved in. Another way to get some ballpark number would be to just search Judis by party name. This would probably give you a more accurate figure at the Supreme Court, where the reporting is better, than many of the High Courts.