In the 13 of the 20 states that voted for change since 2014, all have had a declining or zero rate of growth in per capita incomes for the five-year period prior to the election year.
Ease of business, Goods and Services Tax (GST), repo rates and demonetisation are all in the news, and so are the impending elections. A few months ago, the Reserve Bank of India confirmed that 99% of the demonetised notes (Rs 1000 and Rs 500) are back in its vaults. This is an indication that the demonetisation effort was “successful”.
Unfortunately, economic growth in the last quarter dropped by 2.2% over the same period last year.
Critics of demonetisation say that the falling growth is a vindication of their claim, while supporters of demonetisation feel that the slowdown is a larger continuing phenomenon. It would be hard to believe that lack of liquidity did not create a transmission shock, affecting the real economy, but it may also be a stretch to attribute the entire decline to demonetisation. Independent of this, the fall in growth rates alone could turn out to be bad news for the ruling dispensation if we were to anticipate that economic growth does impact election outcomes.
While pollsters are already predicting the future from their surveys, it is worth checking if economic outcomes and poll outcomes have been correlated in the recent past in India.
It is popularly believed that an incumbent party’s past performance is the principal factor determining individuals’ voting decision. Does this hold true for India where voting behaviour is influenced by social identities like ethnicity, family lineages, religion and caste? Since this is common knowledge to all contestants and voters, political competition should typically reduce ethnic rents of this nature to zero, presumably leaving one thing to distinguish between incumbents and their competitors – past performance or promise of better governance. Since past performance is often a good indicator, expectations are history dependent.
One of the quantifiable yard sticks to judge past performance of an incumbent party is growth in incomes experienced during the period of their governance. In a study of poll outcomes covering 422 candidates in India’s 2009 parliamentary elections, Gupta and Panagariya (2014) found that candidates of incumbent parties in high-growth states have much better prospects of victory than those in low-growth states.
This expectation has been re-affirmed in election outcomes since 2014. States that had positive growth re-elected the incumbent party and states that had declining growth rejected the incumbent party (see table). Twenty states cumulatively went to polls since 2014 (Andhra Pradesh, Arunachal Pradesh, Haryana, Jammu and Kashmir, Jharkhand, Maharashtra, Odisha and Sikkim), 2015 (Delhi and Bihar), 2016 (Assam, Kerala, Puducherry, Tamil Nadu and West Bengal) and 2017 (Goa, Manipur, Punjab, Uttar Pradesh and Uttarakhand).
In 2014, the Indian economy had seen a decline in income growth rate and we saw the incumbent Congress get a drubbing in the parliamentary elections. In the 13 of the 20 states that voted for a change since 2014, all (except Haryana) had a declining rate of growth in per capita incomes for the five-year period prior to the election year. Haryana had a neither an increase nor a decline.
In the six states where the incumbents were re-elected, three had an increasing growth (Manipur, Sikkim, West Bengal) and three had a declining growth (Arunachal, Odisha, and Tamil Nadu). This outcome in the first three states is in line with our claim but the latter group seems to contradict our claim – or there may be something different going on in these states that may have over-ridden the economic growth factor.
For example, Arunachal Pradesh has elected the Congress as the single largest party since 1984. Similarly, Odisha has elected the Biju Janata Dal as the single largest party since 2000. It is possible, therefore, that persistence or inertia (for lack of a better term) may be an explanator for Arunachal and Odisha where despite a declining growth rate, the electorate may have elected the incumbent.
Tamil Nadu, however, has not exhibited such long unbroken romance with AIADMK and in that sense, is an outlier. Goa and Manipur elected the Congress as the single largest party in the 2017 elections, but the Bharatiya Janata Party formed an alliance post-election that gave them a majority in the legislature. Bihar was an odd case. Prior to the elections, there was a split between the incumbent allies Janata Dal (United) and the BJP. The JD(U), Rashtriya Janata Dal and the Congress, for the first time, formed an alliance against the BJP. So it is open to interpretation whether the election of the alliance was one of re-electing the incumbent government or a change.
The election results may be reiterating a message that our political parties have long internalised – “roti, kapda our makan” or “sabka saath, sabka vikas”. What does this foretell for the seven states that are gearing up to go to polls over the next year – namely Gujarat, Nagaland, Karnataka, Meghalaya, Himachal Pradesh, Tripura and Mizoram and the ones that follow in 2019, including the parliamentary elections? Is the writing already on the wall or will there be any surprises? We will know soon.
Pranab Mukhopadhyay is with Goa University, Goa and Aparna Lolayekar is with DM’s College of Arts, Science & Commerce, Assagao, Goa.