A few weeks ago, Satyapal Singh, the Mumbai commissioner of police, resigned from his job one year ahead of retirement and joined the BJP. This party has traditionally attracted retired policemen, including, in the 1980s and ’90s,  Shrish Chandra Dixit, former DGP of Uttar Pradesh (and then Varanasi MP) and Bharatendu Prakash Singhal (former DGP of UP turned Rajya Sabha MP and former VHP president Ashok Singhal’s brother). Indeed, the BJP claims that it pays a particular attention to law and order as one of the pillars of its plan for “good governance”. The mutual attraction between police officers and the BJP raises an obvious question: do BJP-ruled states have better records than others when it comes to law and order?

Christophe Jaffrelot
Jaffrelot’s core research focuses on theories of nationalism and democracy, mobilization of the lower castes and Dalits (ex-untouchables) in India, the Hindu nationalist movement, and ethnic conflicts in Pakistan.
More >

Based on the figures of the National Crime Records Bureau for the years 2006-2010, the national average of the number of cognisable crimes in India for that period was 17 for 100 inhabitants. No state ruled by the BJP during this period, except Uttarakhand, performed better than the national average: Chhattisgarh and Gujarat stood at 19, Karnataka (where the BJP governed after 2008) at 21 and Madhya Pradesh at 28. But many Congress-ruled states did not do better. Andhra Pradesh stood at 21, for instance. In contrast, the Uttar Pradesh government clocked in a remarkable seven, partly because this state is very populous, but also because the BSP — at the helm after 2007 — mobilised against the “goonda raj”.

Experts point out, however, that the category of “cognisable crimes” must be disaggregated, since it brackets together many different things, including murder, dacoity, arson. Let’s focus on murder, one of the most important items on this list, and see where the BJP-ruled states stand vis-à-vis the others. When we do so, it is not easy to come to a clear conclusion. For Gujarat, Madhya Pradesh and Karnataka, the number of murders per year and for every 10 inhabitants was 1.97, 2.24 and 2.87, respectively, over the period when the party was in office (1998-2011, 2003-11 and 2008-11, respectively — 2011 is the last year for which I have robust data from the NCRB). The Congress did worse in Andhra Pradesh over the period 2004-11, with a ratio of 3.13, and in Delhi over the period 1998-2011, with a ratio of 3.22. But in Maharashtra, over the years 1999-2011, it is neck and neck with the BJP government of Karnataka, with a ratio of 2.85.
 One may argue, however, that the parties’ performances should be compared within the same state. This comparison shows that in Himachal Pradesh, the BJP has done better than the Congress when they alternated in power, with 1.7 for the Congress over 2003-07 as opposed to 1.5 for the BJP over 2008-11. This is also true in Rajasthan, with 1.8 for the BJP in 2003-07 and 2 for the Congress in 2008-11. But in UP, the BJP compares very badly with the BSP, with a surprisingly high 4.6 over the years 1997-2002 against the BSP’s 2.3 over the years 2007-11.

Do these figures reflect contrasting policies? In particular, have the BJP-ruled states tried harder to improve law and order by recruiting more policemen, one of the things states can do under the Indian Constitution? In 2012, the Bureau of Police Research and Development released revealing data regarding vacancies in the state police across India. The national average is shocking since the actual strength of the state police, 1,585,117, represented roughly 74 per cent of the sanctioned strength (2,124,596), as if politicians had a vested interest in under-staffing the police to save money and/or to help private security agencies prosper. States traditionally ruled by the BJP did not appear to make a systematically bigger effort to reduce these vacancies.

There appears to be no correlation between the strength of the state police and the number of reported crimes. As mentioned above, UP registered the lowest number of crimes and also has the highest vacancy rate — close to 60 per cent. Does it mean that there are more crimes when there are more policemen simply because they take haftas — or that we cannot rely on the number of reported crimes, most of them remaining unreported? (To answer the second question, one should be able to compare the ratios of the FIRs that are filed).

Another important variable that needs to be factored in is the efficiency of the judiciary in charge of trying the alleged criminals arrested by the police. If cases remain pending and alleged criminals on bail, the latter can more easily continue to take the law in their hands. We must turn here to another vacancy ratio, that of high court judges, as proxy for measuring the attitude of the states to the judicial machinery as a whole. The national average is worse than in the state police —  30 per cent against 26 per cent — which partly explains the huge backlog of cases pending in the courts all over India. But, again, there is no correlation between the number of judges and the identity of the ruling party or the number of crimes. In fact, some traditionally BJP-ruled states, including Chhattisgarh and Gujarat, have more than the national average of vacancies (with 33 per cent in both cases), while others are doing better (Madhya Pradesh stands at around 20 per cent only). Congress-ruled states are similarly distributed, with Andhra Pradesh at 34.7 per cent and Maharashtra at 20 per cent.

Assessing the “legal structure and property rights” in 20 states in 2007, Bibek Debroy and Laveesh Bhandari gave contrasting grades to states that were ruled by the BJP for at least four years. Madhya Pradesh was second out of 20, Chhattisgarh came in at number four, Rajasthan was at number five and Gujarat — the state already benefiting from the longest BJP rule  — was at number nine.

Last but not least, the ranking of the state so far as corruption is concerned must be considered, since it tells us something about the quality of governance. To measure corruption is not easy and few attempts have been made at the state level. The most convincing ones are slightly old, unfortunately. In 2008, Transparency International surveyed different Indian states for corruption by interviewing more than 22,700 BPL households. The organisation classified Jammu and Kashmir, Bihar, Madhya Pradesh and Uttar Pradesh in the “alarmingly corrupt” category, Karnataka, Rajasthan and Tamil Nadu as “very highly corrupt”, Chhattisgarh, Delhi, Gujarat, Jharkhand, Kerala and Orissa as “highly corrupt” and Andhra Pradesh, Haryana, Himachal Pradesh, Maharashtra and Punjab as “moderately corrupt”. This list suggests that BJP-ruled states were represented in each of the categories.

We may have to conclude that former policemen probably do not join the BJP only because of its ability to enforce law and order.

This article was originally published in the Indian Express