Selection bias and land acquisition data

What proportion of industrial projects are being held up by land acquisition challenges? A somewhat abstruse debate entered the mainstream after Rahul Gandhi cited Centre for Monitoring Indian Economy (CMIE) data (provided by the Finance Ministry) in a 12 May Lok Sabha speech:

Although I have argued elsewhere that the 8% figure may be an exaggeration in the context of Narendra Modi’s land amendments, the more common view is that it understates the extent to which land acquisition difficulties inhibit manufacturing in India, as Maitreesh Ghatak has explained most clearly in Quartz:

“It’s a classic underreporting problem,” said Maitreesh Ghatak, a professor at the London School of Economics who has studied land acquisition law in India. “There may be projects that never got started because they anticipated these problems. Also, the ones that did get started are likely to have been selected because the risk of land acquisition problems was low for them for whatever reason.”

This is a perfectly fair point, but it doesn’t necessarily follow that the 8% number is an understatement. Imagine a businessperson who is contemplating setting up a plant to manufacture wickets. One could easily imagine her investigating the feasibility of setting up such a plant, but giving up because she failed to procure land.

But one could also imagine her giving up because she couldn’t get a large enough bank loan, or a sufficient supply of workers, or even permission to cut trees to manufacture the wickets. To use Ghatak’s language, she may have chosen not to start because she anticipated any, or several of these problems. One simply does not have enough information about the universe of projects that were contemplated but not started, which doesn’t justify throwing out the data about projects that were started but subsequently ran into trouble. Because the ones that got started are also likely to have been selected (into the CMIE sample (pdf) of 804 projects) because the risk of labour, capital or market problems was low for whatever reason.

But let’s assume for argument’s sake that the number of unobserved projects delayed by land acquisition issues is in fact significantly high. It doesn’t follow that Modi’s land amendments will have any impact on their viability. The original 1894 land act was in operation until 31 December 2013, until which time the government was able to compulsorily purchase land for private companies without landowner consent or carrying out a social impact assessment, just as in the amended law. Yet land acquisition problems abounded. It would be ludicrous to argue that a law that was in operation for a single year — and excluded nuclear energy, mining, railways, national highways and petroleum pipelines from its purview — was responsible to delaying land acquisition in the prior decade.

Advertisements

Is the Modi Effect waning? (What Jammu and Jharkhand tell us)

The recent state elections in Jammu & Kashmir and Jharkhand went well for the Bharatiya Janata Party (BJP), but the results have also sparked speculation as to whether the Narendra Modi effect is beginning to wane. Congress Party MP Rajeev Satav seems to think so (and was retweeted by Aam Aadmi Party leader Arvind Kejriwal):

If only things were so simple.

Take Jharkhand. As Satav says the BJP’s vote share dropped about 9 percentage points between the general election in April-May 2014 and the state assembly election in November-December 2014. But that’s not unexpected: the chart below shows the BJP has lost 7-9 percentage points in vote share between national and state elections held in close succession since 2004.

Screen Shot 2014-12-28 at 7.52.10 pm

 

This is also true of state assembly constituencies won (see chart below). In both 2009 and 2014, the BJP in Jharkhand won 19-20 more state assembly segments in national elections than it did in the state elections that followed a few months later. Note that 2004/05 is an exception partly because the BJP stood alone in the 2004 parliamentary election but allied with the Janata Dal-United in 2005 for the state election.

Screen Shot 2014-12-26 at 8.53.44 am

This pattern occurs presumably because many Jharkand voters chose to vote for the BJP in national elections but switched to a regional party such as the Jharkhand Mukti Morcha that seemed a more viable option in state politics.

The BJP’s position in Jammu appears unchanged between the national and state elections. While it’s true that the party’s overall vote tally in Jammu and Kashmir has fallen, its grip on Jammu hasn’t really weakened. The BJP won 24 assembly segments (in the Jammu and Udhampur parliamentary constituencies) in the general election, and won 25 assembly seats in Jammu in the subsequent state election.

There is therefore no real evidence to suggest that the Modi effect (or anti-Congress sentiment) has waned either in Jammu or Jharkhand.

The Rajya Sabha will remain a headache for the NDA

With the Lok Sabha, the lower house of Parliament, securely in the hands of the Bharatiya Janata Party (BJP) and its allies, the locus of political opposition is set to move to the Rajya Sabha, the upper house, where the BJP remains in a minority. Most bills need to be passed by both houses to become law, and this has sparked speculation as to how the BJP might push through its legislative agenda without a Rajya Sabha majority.

It is evident from the chart below that the the United Progressive Alliance (UPA) controls many more seats in the Rajya Sabha than does the BJP-led National Democratic Alliance (NDA).

Screen Shot 2014-06-01 at 5.03.14 pm

There are two obvious paths before the BJP:

  1. Regional parties. The ruling party could make deals with unaligned regional parties (such as the Biju Janata Dal and the AIADMK) to win support for specific bills. Since the NDA is 59 seats short of a majority it will have to convince six or seven of the largest regional parties to support it in the Rajya Sabha.
  2. Joint sittings. The BJP could arrange for joint sittings of the two houses under Article 118 of the constitution. The NDA is slightly over the halfway mark of the two houses’ 777 combined seats. However this is a rare step, carried out only thrice in Indian parliamentary history, and would require the consent of the President of India.

With two ordinances passed in its first two days, the Modi government has signalled that it may not be overly concerned about precedent. Even so, frequent joint sittings of the two houses could anger opposition parties that have become quite efficient at disrupting parliament in recent years.

The desirability of each strategy depends in part on how the number of seats controlled by the NDA grows over the next five years. If the NDA can woo new allies easily then it is less likely to take recourse to the more drastic step of calling joint parliamentary sessions. Recall that the Rajya Sabha reflects the balance of power in India’s states, and that a third of its members are elected every two years.

How then might the balance in the Rajya Sabha shift in the coming years? Let us assume that the NDA’s strong 2014 Lok Sabha performance translates into state election victories in 2014 (Delhi, Haryana and Maharashtra), 2015 (Bihar) and 2016 (Assam) in roughly the same proportion. This is a contestable assumption because the national level “Modi wave” will have less relevance in a state election in which local factors gain prominence.

Still, maintaining that assumption, here is what the Rajya Sabha might look like in 2016 (excluding the 11 current vacancies in the upper house):

Screen Shot 2014-06-01 at 5.08.31 pm

This is quite clearly not enough to tilt the scales in the NDA’s favour. Even after the 2016 Rajya Sabha elections, the ruling alliance will be 47 seats short of a majority and will need the support of regional parties such as the Samajwadi Party and the All India Trinamool Congress.

The NDA’s prospects begin to sharply improve in its fourth year if the 2014 election results are mapped on to state elections in 2017. Keep in mind that this is a very optimistic scenario for the alliance: the BJP’s popularity could well have ebbed by then and the local factors mentioned above will again be in play in a state election. We are also assuming that party state assembly seat shares translate quite simply into Rajya Sabha votes; in practice there is often voting across party lines and strategic voting in favour of third candidates that can make the outcome less predictable.

But assuming that the broad patterns of the general election apply to the 2017 state elections in Goa, Gujarat, Himachal Pradesh, Manipur, Punjab, Uttar Pradesh and Uttarakhand, this is what the Rajya Sabha could look like in 2018:

Screen Shot 2014-06-01 at 5.09.12 pm

The NDA is still 30 seats short of the halfway mark in the Rajya Sabha, and the ruling alliance will need to win over at least one of the three biggest regional parties (Samajwadi, Trinamool and AIADMK) plus an assortment of smaller players. The NDA is well over the halfway mark of a joint sitting at this point, so its preferred strategy will depend on how easily it can woo those parties.

Either way, the NDA will remain short of a majority in the Rajya Sabha, although its rising tally will require it to attract fewer allies by 2018. But 2018 is close to the end of the its term, which means that for most of its tenure the NDA will be forced to reach out to a large number of regional parties to pass bills in parliament.

Is AAP India’s most criminal party?

With candidates declared for six of the nine rounds of voting in this never-ending general election, a partial analysis can now be done of how many stand accused of criminal activity and in what proportion parties are awarding tickets to such folks (data from National Election Watch, also see preliminary analysis here).

The chart below ranks parties by the proportion of candidates that have serious criminal charges against them:

Screen Shot 2014-04-22 at 1.45.48 pm

As before, parties from Bihar and Maharashtra have the highest percentage of Lok Sabha candidates that face serious criminal charges. In 2009, the Indian National Congress (INC) and Bharatiya Janata Party (BJP) were comparable; the BJP is doing a little worse so far in the 2014 election with 17% of its candidates facing serious charges compared with 13% of the INC’s.  The Bahujan Samaj Party (BSP) is performing better this time round: it has fewer accused candidates than the BJP, INC or the Samajwadi Party (SP). The Aam Aadmi Party (AAP), as might be expected, looks pretty good here.

What about levels of criminality? To calculate the intensity of the alleged criminality of each party’s candidates, we count the number of candidates with more than one serious charge, add up the number of charges and calculate the total as a percentage of the total candidates per party. And here is what we find:

Screen Shot 2014-04-21 at 11.40.40 am

Shock and horror! AAP turns out, by far, to have the highest levels of alleged criminality of all of India’s parties. The Shiv Sena and Nationalist Congress Party come a poor second and third respectively.

As it happens, AAP’s ranking is driven by two statistical outliers: Anti-nuclear activists SP Udayakumar and M Pushparayan respectively account for 382 and 380 of the 829 serious charges that AAP candidates face. The next such candidate is Trinamool’s Kameshwar Baitha with “only” 48 charges, suggesting that the numbers for Udayakumar and Pushparayan are unusual. After all, is Udayakumar really 48 times more “criminal” than Sri Ram Sene leader Pramod Mutalik who faces eight serious charges?

To limit the effect of such outliers, let’s give a score of 10 to all candidates who were had more than nine serious criminal charges. The result is:

Screen Shot 2014-04-22 at 1.46.24 pm

The AAP now drops in the rankings and ends up with about the same intensity of criminal charges as the Congress Party, although it still does worse than the BSP and Biju Janata Dal. The “criminality” of the BJP slate of candidates is more than double that of the INC and AAP, up from being 28% greater than the INC in 2009. The Congress score in 2014 is about the same as it was in 2009, but the BJP’s is sharply higher this time.

As before, the BSP does well in this ranking, although its score is not strictly comparable with the SP’s because the latter is focused much more on Uttar Pradesh and has only two-fifth as many candidates. Communist Party of India (Marxist) supporters can breathe a sigh of relief: it moves from the top of the list in 2009 to somewhere in the middle.

Finally the overall incidence of serious criminal charges is higher (so far) in 2014: 302 of the 533 candidates facing serious criminal charges (the 57th percentile) have more than one charge against them compared with 355 of 1,114 (the 32nd percentile) in 2009.

Remember however that this is still an incomplete list: we await candidate data for the final rounds of voting.

Why Modi may have a fight on his hands in Varanasi

It seems clear that Uttar Pradesh (UP) with its 80 Lok Sabha seats is the Bharatiya Janata Party’s (BJP) biggest bet in this election and that Narendra Modi is standing from Varanasi in order to boost the BJP’s prospects across eastern UP.

The BJP’s man in UP Amit Shah certainly seems gung-ho:

With a couple of exceptions, many people are assuming that Modi will easily win in Varanasi even if Arvind Kejriwal joins the contest. For instance CNN-IBN’s Sagarika Ghose tweeted this yesterday from Varanasi:

But don’t take anything for granted, as the following analysis by two political scientists suggests. Modi will probably still pull off a win in the face of adverse local politics, but he may have a fight on his hands.

Guest post by Philip K. Oldenburg, Research Scholar, South Asia Institute, Columbia University

The implications of the choice of the BJP to field Narendra Modi in the Varanasi constituency need to be explored a bit more.  It is not simply that the BJP has to do very well in UP, and that Modi contesting in the state gives them a much better chance (which I would agree with), and that this is a signal that Hindutva will remain central to the BJP (possibly true). 

It is not at all clear that Modi would win the seat, and certainly will have to expend a good deal of effort if he wants to have a chance.  In 2009, the party-wise vote shares were: BJP (30.5%), Bahujan Samaj Party (27.9%), Samajwadi Party (18.6%) and Indian National Congress (10.0%). Jolie Wood, who teaches political science at Allegheny College, writes (on 19 Mar):

As someone who has spent some time doing fieldwork in Varanasi and has met many major local political figures there, I agree . . . that it is not at all clear that Modi will win in Varanasi.

The city is often viewed as a “BJP stronghold”, but it is not exactly that. First, Muslims make up about 27% of the city population and at least 30% of the district’s population (according to the admittedly outdated 2001 census). Joshi won his seat with just over 17,000 votes over a Muslim candidate who was not previously very prominent and is widely seen as a mafioso. The business community, probably the best organized occupational group in the city, is not entirely pro-BJP – the SP also has a very strong organisation among them. Many in the business community (including those loyal to the BJP) are quite wary of anyone who might inflame communal sentiment and in fact discouraged more right-wing elements of the BJP from pushing the Gyanvapi Masjid/Kashi Vishwanath Mandir issue because it was bad for business.

If we look at other reasons why the city is seen as a BJP stronghold, we can also see some reason for doubt. Take, for example, the current Members of Legislative Assembly (MLAs). The MLA of Varanasi South is Shyamdeo Rai Chaudhary, who has held that seat for something like 6 or 7 terms now. He is personally popular and is seen as someone who is decent, trustworthy, non-corrupt, and (importantly) not super partisan or “communal”. Jyotsana Srivastava is not a major political player by herself – she alternates terms (more or less) with her very popular, much higher profile, husband Harishchandra Srivastava. These MLAs win consistently on the basis of their personal stature, as much as or more than on the basis of their party identification. Finally, the MLA from Varanasi North, Ravindra Jaiswal, is a business leader who is a first-time MLA, and who holds a seat in a district that has usually gone for Muslim candidates linked to the Ansari community, which is particularly strong in that constituency. I would not assume that he or any BJP candidate can easily hold onto that seat.

So, my point is that the BJP’s current dominance of Varanasi’s elected offices is not necessarily due to the strength of the BJP per se. Add to this that the very large Muslim community tend to support SP, Congress, or BSP. Congress has done very poorly in the city for the last several elections. The SP and BSP have much stronger organizations in Varanasi than does Congress, so I’m not sure why some discount their role in the coming election, though it is true that they’ve never fielded a strong candidate for the MP seat.

The big question for me is how they will approach this election. If they form an alliance with AAP, it could weaken the BJP considerably. [But] I really wonder if AAP could coordinate any kind of agreement with BSP and SP. And while Mukhtar Ansari may have won a lot of votes in 2009, I actually don’t know if the BSP is that much stronger in Varanasi than it was in 2007, when it had no presence at all, even when Mayawati won an outright majority in the state. I do know that Congress had become a total non-player by then.

I do think that if Muslim voters backed AAP in a more or less unified way, then Kejriwal might have a chance, but if BSP fields Mukhtar Ansari again, that probably won’t happen. (Also, many Shia Muslims in Varanasi support BJP.)

Jolie’s analysis seems sound to me, and it raises an important question: If this is a seat for which Modi must fight, and if fighting for it means spending significant time in Varanasi that he might otherwise have spent travelling throughout the country, then perhaps this choice of seat might weaken the BJP’s campaign.

And if it so happen that Modi loses – and here the importance of Jolie’s point that in effect “communalism can be bad for business” is particularly important. If Modi tones down his Hindutva rhetoric, then what happens in those parts of UP where that would draw votes and, more important, a hyperactive “base” of political workers? And if the BJP and its pre-election allies and supporters fall short of a majority, then having an alternate BJP prime minister becomes a serious possibility.

How opinion polls fared in the last two general elections

As more and more opinion polls hit the headlines, it’s worth recalling how they fared in the two most recent general elections:

2004

2009

Note: The figures in parentheses under the NDA and UPA columns are the respective seat tallies of the BJP and Congress Party.

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: