How various regions fare in Modi’s Cabinet

There has been much talk of the politics behind Prime Minister Narendra Modi ‘s 5 Jul Cabinet reshuffle. It hasn’t escaped anyone’s attention that several new appointees came from Uttar Pradesh and Gujarat, both states in which the BJP has big stakes and that have elections due in 2017.

But how important is regional representation in the appointment of ministers? One way is to look at how under- or over-represented various regions are. This map (from the business news daily Mint) shows that Hindi heartland and western Indian states have the strongest representation:

Screen Shot 2016-07-08 at 12.35.48 PM.png

A good first cut, but this only gives us absolute numbers. And it shouldn’t surprise that the more populous states get the most ministerial positions. One way to check under- or over-representation is to compare a state’s presence in the Council of Ministers with a baseline expectation of how much it should have. One way to do this is to look at the ratio of a state’s share of ministries with the proportion of Lok Sabha seats that it has (a number that is also roughly aligned with its population share).

And here is what we find:

Best represented states in the Council of Ministers (ministry share/LS seat share)

Screen Shot 2016-07-08 at 1.53.03 PM.png

Surprise! Goa and Arunachal Pradesh are at the top of the list, since the appointments of Defence Minister Manohar Parrikar and Minister of State for Home Affairs Kiren Rijiju give each 3.5 times (or 350% of) their parliamentary presence (proportionate representation is equivalent to 1, or 100%). However small states skew these results because the appointment of only one or two ministers can give them huge over-representation.

Sticking to the larger states, Uttar Pradesh and Bihar now stand demoted. The best represented are Haryana (2.1), Madhya Pradesh (1.9),  Rajasthan (1.7), Gujarat (1.6), Jharkhand (1.5), Bihar (1.4), Uttar Pradesh (1.3) and Karnataka (1.2). The poorest represented are Tamil Nadu (0.2), West Bengal (0.3), Odisha (0.3), Telangana (0.4), Assam (0.5), Chhattisgarh (0.6), Andhra Pradesh (0.8) and Maharashtra (0.9).

The states with the highest representation include those with well-established state BJP units that comprise the party core (with Haryana a notable outlier). Conversely, the states with the lowest representation are mostly those where the BJP is weak (with the interesting exceptions of Assam and Chhattisgarh where the BJP is the ruling party).

Regional representation isn’t the only factor in allocation ministerial positions, administrative ability and identity group balancing being other plausible drivers that could explain some of the deviations from 1 (or 100%). But the broad pattern seems consistent with a patronage model of politics in which political parties need to keep their base happy, motivated and delivering benefits to supporters.

Update

Another way to predict how much representation a state “deserves” would be to look at its contribution of parliamentary seats to the ruling coalition, in this case the 331-member National Democratic Alliance (NDA). The pattern now appears somewhat different:

Best represented states in the Council of Ministers (ministry share/NDA seat share)

Screen Shot 2016-07-08 at 7.42.53 PM.png

This time the small states are joined at the top of the list by large states in which the NDA has a limited presence. West Bengal and Tamil Nadu score high because they account for two and one ministers respectively, the same as the number of MPs they sent to the Lok Sabha. Among the larger states, the best represented are West Bengal (4.2), Tamil Nadu (4.2), Odisha (2.1), Haryana (1.8), Madhya Pradesh (1.4), Punjab (1.4) and Karnataka (1.3).

Other than Madhya Pradesh, which remains in the overrepresented column, the Hindi heartland and western states now seem more fairly represented, staying close to their weightage in the NDA: Bihar (1.1), Jharkhand (1.1), Rajasthan (1.0), Gujarat (1.0), Uttar Pradesh (0.9). Under-represented NDA states include Chhattisgarh (0.4), Maharashtra (0.6) and Assam (0.7).

From the NDA’s point of view, Modi’s Council of Ministers provides a good balance. No major state other than Kerala gets short shrift (though some in BJP-ruled Chhattisgarh, Maharashtra and Assam might grumble).

Rush to judgement

The Association of Democratic Reforms has analysed 280 of 469 Lok Sabha candidates announced by the Bharatiya Janata Party (BJP) and the Indian National Congress (INC) and discovered that:

The data do not seem to be up as yet on the ADR website, so here is quick breakdown by party using the numbers quoted in news reports:

Screen Shot 2014-03-22 at 10.52.15 am

It would appear that the proportion of candidates with criminal cases in 2014 has gone up in comparison with 2009 — from a fourth to close to a third. The INC’s proportion stays constant at 27% but the BJP’s seems to have risen by 7 percentage points to 35%.

But don’t race to any conclusions until parties have declared all their candidates. We don’t know the basis on which the sample of 280 candidate affidavits was chosen, and that too from a partial list of 469 candidate declarations. The two biggest parties together had 833 candidates standing for election to the Lok Sabha last time.

Should opinion polls be banned?

Despite my critique of their use of communal violence data, the India Today Group graciously invited me to participate in a panel discussion on opinion polling at the India Today Conclave 2014 on Saturday, Mar 8.

I made the following points:

  1. Parliamentary seat projections should be treated with skepticism because there is no simple relationship between vote shares and seats won (as I have argued previously).
  2. About a fifth of the seats in the 2009 Lok Sabha election were won with a margin of 3 per cent or less, which means that fluctuations within the margin of error of most surveys can dramatically change the number of seats a party wins.
  3. Between 20 and 30 per cent of voters make up their minds about whom to vote for a couple of days prior to voting, and last minute vote swings can make a big difference.
  4. Of course I didn’t think that opinion polls should be banned but made more transparent and evaluated by better-informed consumers.

The full video of the event is here: India Today Conclave 2014: Need to regulate opinion polls, say psephologists.

Here is an edited version:

Political parties’ Facebook activity

In response to my post on Political parties’ Facebook “likes”, Pragya Tiwari correctly pointed out that an additional flaw in that measure is the likelihood that parties and politicians purchase “likes”, and that people “talking about this” might be a less manipulable figure.

So here is some data (an evening’s snapshot on Mar 1):

Politicians here appear to attract more chatter than political parties, although we don’t know what people are saying. No doubt there are commercial tools available to figure that out, and I’d be interested in getting such data were it to be available.