Wednesday, March 11, 2015

Reinventing the criminal justice system (Part 1 of 2)

by Nandkumar Saravade.

Police and court inefficiency is the dominant feature of the Indian criminal justice system. The latest verdict on the L. N. Mishra murder case, which took 40 years to resolve even when the victim was a central minister, has once again brought home the point that things just don't work. Our system has repeatedly fallen short of its objective of being able to investigate and prosecute to ascertain the culpability of the accused and award deterrent punishments. Failure in this area is a significant inhibitor for achieving society's true potential, including economic progress.

After many decades of stagnation, there are now signs of important change in this field:

  1. There was a time when the dominant idea of the Indian State was a "mai baap" who would wipe every tear. The Indian State is now re-engineering itself towards a much narrower vision, of achieving State capacity in delivering core public goods, and leaving the pursuit of happiness to the private choices of each individual. Law and order looms large in the field of public goods. Law and order is the ultimate public good: it is non-rival (my consumption of safety does not diminish the safety available to you) and non-excludable (there is no way to exclude a new born child from the blanket of safety).
  2. Important new work has begun on building high quality courts in India. If these innovations work out well, and are then transplanted more widely into the working of courts all over India, this could have a major impact upon the criminal justice system.
  3. The 14th Finance Commission has given a great surge of budgets at State governments. This is an important time for State governments to rethink their expenditure priorities, and a key element of this should be a greatly increased focus on law and order.

Prioritisation of the criminal justice system is going beyond the intellectual elite to every citizen. One highly visible episode of rape in Delhi [link] appears to have had an impact on elections at the level of the state government and possibly in the general elections also. Politicians are now much more oriented towards succeeding on law and order. As an example, the manifesto of the newly elected AAP government in Delhi speaks of reducing judicial delays.

We may thus be at an important moment, at the early stages of reinventing the Indian criminal justice system. It is, then, particularly important to obtain a clear diagnosis of what is wrong, and of effective strategies for reform. This is what motivates this two-part article. The first part is on the state of affairs of crime, and of the police machinery. The second part will be on the strategy for police reform.

What is the extent of crime?

Crime must be measured as as the crime experience of the citizenry. E.g. the number of murders per 100,000 of population, per year, is a measure of crime. For our present purposes, it is useful to think of the work of the police which takes in a complaint that is filed at one end, and results in convictions at the other end. The workload that comes into the entry point is the number of FIRs filed.

During the 2003-2013 period, India's population grew by 15%, but was outpaced by the growth of crimes defined in the Indian Penal Code (IPC) by 54.3%. The number of IPC crimes registered as First Information Reports (FIRs) during the year 2013 was about 2.65 million.

While the IPC is the main law on crime, there are other laws called Special and Local Laws (SLL), which also keep police investigators busy. These include offences under, among others, the Motor Vehicles Act, Arms Act, Gambling Act and Narcotics Act. SLL figures reached 4 million in 2013. SLL grew less rapidly (5.7%) during the same period, probably due to the fact that most of cases booked under SLL result from a proactive and preventive approach, depending on the time and resources available to police leadership at the local level. For the sake of simplicity, we will focus only on IPC crimes.

With respect to IPC FIRs, investigators processed 3.5 million cases (including cases carried forward from the previous years) during 2013. They completed investigation in 2.5 million cases, of which 1.9 million (75%) cases were sent for trial, leaving 0.95 million cases under investigation to carry over to the next year.

Large as these numbers seem, it is well known that only a fraction of the criminal incidents get registered as FIRs. Indeed, non-registration of crimes is a major factor in citizens' dissatisfaction with the Indian Police. In a landmark experiment conducted in Jalpaiguri, West Bengal in 2007, it was revealed that the actual incidence of IPC crimes across different categories was 4 to 6 times of the recorded numbers. More recently, the Delhi Police leadership stepped up registration, with revealing results. Delhi Police had recorded 64,882 IPC cases in 1998, which fell to 54,287 in 2012. However, the numbers rose rapidly to 80,184 in 2013 and 147,230 in 2014, close to a three-fold jump over two years.

How do the police fare on investigative capacity?

The Bureau of Police Research and Development (BPRD) publishes data on police resources on a regular basis. It shows that the police-population ratio has improved from 139 (per 100,000 population) in 2002 to 181 in 2012. The state police comprises of civil police (police station and support unit staff) and armed police (reserve units). According to NCRB, the total sanctioned strength of the state civil police at the end of 2013 was 1.79 million, while the actual was 1.3 million, leaving about 25% of the positions vacant.

For the purpose of investigation, it is more important to look at the number of investigators who primarily come from the ranks of Assistant Sub Inspector (ASI), Sub Inspector (SI) and Inspectors. These are mere 11% of the civil police strength. However, these officers cannot consider investigation as their primary job, having to give priority to maintenance of public order, VIP security and court appearances.

There have been a few studies in recent years estimating the actual workload at police stations. Though not specifically aimed at the investigation function, these provide some clue about the existing gap. A recently released BPRD report on 8-hour shifts in police stations looked at the existing supply side situation and found that majority of police stations had its staff putting in 11- hour to 14-hour workdays. Further, most police station staff members could get only one or two days as weekly off during a month. The report estimates that to enable a 48-hour work week, with an assured weekly holiday, it will take 68% enhancement in the police station strength. Another highlight was that, of the total police strength, only a third was posted in police stations, though, admittedly, it was the most important part of the police organization.

A smaller workforce is directly linked to low crime registrations described earlier. They are seen as a necessary evil by police supervisors to ensure that work load remains manageable. Even when the top leadership is interested in more realistic registration, the police stations thwart the efforts by turning away complainants. In a territory-based approach, the local police station has a monopoly on registrations and uses it to the detriment of the poor victims, who are unable to get their grievances addressed. In other countries, this monopoly is taken away to a certain extent by a centralised emergency response system, such as the 911 mechanism in the US. No equivalent system has been created in India yet. Madhya Pradesh Police is attempting this currently, with the project cost of Rs.6.33 billion over a five year period.

Court delays and policing

Another factor impacting investigative capacity of the police is the pendency in courts. As described earlier, 1.9 million cases got added to the pool of cases under trial in different courts, swelling the number to 9.8 million. During 2013, only 1.3 million cases were disposed of, increasing the pendency by 0.6 million. Of the 1.3 million disposals, about 0.2 million were due to cases being compounded or withdrawn, 0.77 million discharged or compounded and only 0.52 million convictions. The success rate in offences like rape and murder was around 27% and 36% respectively. The prosecution was even less successful in economic crimes like criminal breach of trust (23%) and cheating (24%). The slow pace of the proceedings was apparent from the vintage of concluded cases, with 13% cases aged between 5-10 years and 3.2% older than 10 years.

Court pendency keeps the investigators unproductively occupied in attending courts where more time is spent on procedural matters, rather than deposing. It reduces the viability of success of prosecution, finally impacting the outcome of investigations and motivation of investigators. It adds to the burden of preserving evidence and tracking witnesses - jobs mostly done by the police - over a longer period of time. Delayed trials are also responsible for the lower rate of success, as a study of corruption cases under the Karnataka Lokayukta showed.

What is the level of funding?

Though manpower is an important factor, it is also relevant to look at the financial expenditure on police department, as an indicator of the relative importance given by the state governments towards bolstering the rule of law. State governments seem to be spending between 3-5% of their budgets on the police.

It would be interesting to put this low expenditure in context by looking at Maharashtra budget figures for FY 2015. The total number of employees under the state government was 0.64 million, of which police personnel were 0.2 million. Thus, despite being one third of the state government workforce, the budget allocation for police was less than 5%, of which sustenance spends (such as salary, fuel and office expenses) took away 98%. The low allocation of resources for training results in lack of upgradation of investigators' knowledge about forensic techniques and new modus operandi.

What is the level of data collection and analysis?

Poor funding has meant that police departments lack IT systems for managing workflow better and data analytical tools to crunch large data sets available in cases of terrorism and financial crimes. The funding for creating new data-sets on understanding crime victimisation, demographic and spatial mapping of criminals, statistical analysis of the investigative processes are non-existent.

One significant factor in doing meaningful analysis is the antiquated method of compiling and analysing crime statistics. The whole process is manual and prone to errors and manipulations. An enterprise level case management system, called Crime and Criminal Tracking Network and Systems (CCTNS) has been delayed by more than four years due to inadequate funding and poor project management. Till the crime data is digitally captured at source and is amenable to modern techniques and tools of business intelligence, mapping the problem of poor-quality investigation in adequate detail remains only an aspiration.

What incentives exist for better performance?

It is not impossible to conceive of an efficient police force within the limitations imposed by the manpower and resource constraints. For this to happen, performance appraisals need to be aligned to elicit appropriate behaviour. Little is known about what constitutes good performance, and how police fare on such metrics.

What metrics of performance exist, are not linked to effectiveness in investigating crime, since it figures low in the list of priorities. Also, trials take a long time to reach conclusion and a negative outcome is attributable to various factors, such as poor investigation, insufficient evidence, long delays and witnesses turning hostile. Existing internal processes to fix accountability for bad investigation seldom seem to work.


Even at current levels of registration of crime, the criminal justice system is overwhelmed and in a state of logjam. There is gross under-registration of crimes; a fact well known to the police leadership. There continue to be considerable delays in police investigation. The police continue to be under-staffed and under-funded. At the same time, performance measurement of the police is weak.

Simply adding more manpower, or increasing resource allocation without appropriate performance metrics and accountability frameworks will not get us the desired results. A reform strategy has to take into account the failures on each of these fronts and devise an approach that will bring in more accountability, structural modifications, adequate resourcing, training and process improvements. This will be the subject of Part II of this article.

1 comment:

  1. Very succinct article.

    One point (never appeared here) which however kept me glued to the article is the subject to non-independence of policing system in India. This phenomenon became ingrained and accentuated post Emergency. While digging the past may give a perspective (which I understand is not the purview of the article) the present ailment of political control of policing system compounded with resource constraints and lack of training gives more reliable grounds for formulating strategies for police reform. Ironical is that Police Act remains a colonial drawdown which lends ambiguities and "unfair" justice mechanisms.

    I was a bit disappointed to find no mention of organizational structure of Home Ministry and State Police that determines a lot about how the system thwarts and brings crime to justice. These are much complicated diagnostics which I feel would have taken the article to trilogy or a series longer than this but would have helped the readers with more insights than what we keep hearing about.


Please note: Comments are moderated. Only civilised conversation is permitted on this blog. Criticising me is perfectly okay; uncivilised language is not. I delete any comment which is spam, has personal attacks against anyone, or uses foul language. I delete any comment which does not contribute to the intellectual discussion about the blog article in question.

Please note: LaTeX mathematics works. This means that if you want to say $10 you have to say \$10.