Do Indian voters really care for economic growth?9 min read . Updated: 29 Nov 2015, 08:16 PM IST
The evidence is mixed on whether the country's electorate rewards parties that deliver high economic growth
The evidence is mixed on whether the country's electorate rewards parties that deliver high economic growth
Two weeks ago, the Bharatiya Janata Party (BJP) suffered its worst electoral setback since it swept to power at the centre in May 2014, losing by a wide margin a hotly contested election campaign in Bihar. The outcome, as readers will recall, was a victory by the Grand Alliance led by incumbent chief minister Nitish Kumar and former chief minister Lalu Prasad.
Why did the BJP lose and the Grand Alliance win? In particular, does the Bihar result cast any light on the hotly contested debate on whether incumbent governments are rewarded by voters for good economic policy, or good economic outcomes, such as high growth of gross domestic product (GDP) in the period preceding an election?
First, an important caveat. A single election result, amounting to one data point, cannot confirm or refute any particular theory. At best, one can say it is roughly consistent, or not consistent, with what one might have expected. There is an almost overwhelming temptation by data-free, armchair commentators to explain an election result in terms of a pet theory, amounting to a form of rationalization after the event, as I had argued in a Mint column in the immediate aftermath of the Bihar election result.
Having said that, much of the post-election analysis focused not on economics but on explanations relying on caste and communal arithmetic. Thus, the widely accepted wisdom in the past two weeks, repeated by an endless number of commentators in print, television and social media, is that the Grand Alliance’s rainbow caste coalition held firm, while the BJP’s attempt at communal polarization—trying to break caste alliances and consolidate the Hindu vote—did not succeed, or did not succeed well enough.
There is not much place for economics in such a story, but is this conventional story missing the real plot? After all, Kumar was a successful two-term incumbent chief minister going into the 2015 election, apart from the wrinkle of the brief stint when he was out of office following his party’s debacle in the 2014 Lok Sabha election. Kumar is widely credited with dramatically turning around governance and economic performance in what had previously been the “jungle raj" of Yadav (although, for a dissenting opinion, see bit.ly/1jpEh20). Albeit starting from a low base, Bihar was indeed one of India’s fastest growing states during Kumar’s tenure as chief minister.
Is it possible then that the subtext of the Bihar election, hidden under the clamour of the titanic clash between Modi and Kumar, is simply that Kumar was rewarded for his good economic stewardship of the state? What does the burgeoning academic and policy literature have to say on the subject of whether incumbent politicians are, indeed, rewarded for good economic outcomes?
The received assumption in most advanced democracies is that good economics makes for good politics. This was perhaps best exemplified by American political strategist James Carville’s immortal utterance, “It’s the economy, stupid!" during Bill Clinton’s successful run for the US presidency in 1992. This conventional view is strongly articulated, for instance, in a 2000 review paper by political scientists Michael S. Lewis-Beck and Mary Stegmaier. As it happens, academic studies do not uniformly support this received wisdom, and the results are decidedly mixed, as documented by a 2007 review paper by political scientist Christopher J. Anderson.
Yet, the general perception persists that, at least in the advanced economies, incumbents are rewarded for good economic performance. As with many aspects of classical democratic theory, it was widely taken as given, until recent decades, that India bucked this trend—that more basic factors such as anti-incumbency, identity politics, or clientelism trumped economic performance in determining election outcomes, or even more nihilistically that the reasons for election outcomes in India remained mysterious.
There is even a school of thought that incumbent politicians are punished, rather than rewarded, by voters if they campaign on an economic platform. Most famously, the unexpected defeat of the BJP and its allies in the 2004 Lok Sabha election was widely attributed to a backlash against their “India Shining" campaign, although several analysts, including Rupa Subramanya and myself argue that there is no basis in fact for this particular received wisdom.
In a 2015 research paper, political scientist Milan Vaishnav and his co-author Reedy Swanson aptly summarize thus: “By reputation, Indian voters are notorious for throwing governments out for reasons that remain elusive, even to veteran scholars."
Given this backdrop, a clutch of important recent papers, including the Vaishnav and Swanson paper, are starting to find at least qualified support for the notion that even the famously fickle Indian voters are starting to reward good economic performance, especially for state governments.
A key research paper by economists Poonam Gupta and Arvind Panagariya, published in 2014 but first presented at a conference in 2010, found a quantitatively large and statistically significant effect of state-level growth rates and the state incumbent party’s electoral prospects in the 2009 Lok Sabha election. In particular, in high-growth states, anywhere between 69% up to a whopping 90% of the incumbent party’s candidates were re-elected, depending on over what period the state growth rate was computed and on exactly how high, medium and low growth states were parsed. In contrast, in low-growth states, the incumbent party’s re-election prospects ranged between 36% and 50% depending on which specification was used.
The Vaishnav-Swanson paper studies elections in the 18 largest states of India from 1980 up until 2012 to determine if there is a relationship between state growth rates and the re-election prospects of incumbent governments. The novelty in their analysis is to introduce a temporal dimension. In particular, they find no correlation between growth and re-election prospects during the approximately three-decade time period if taken as a whole.
However, if the sample is split into the roughly three decades their data spans, they find evidence that in the 2000s, but not in the previous decades, there is evidence of a positive relationship between growth and re-election prospects—defined variously as the binary variable re-elected or defeated, or in terms of vote share or seat share. By contrast, in the 1980s, there is little relationship between growth and re-election, while it appears that in the 1990s, there is actually a negative relationship, so that incumbents were punished for good economic performance.
Vaishnav and Swanson conclude that there is, indeed, evidence that starting in the 2000s, Indian voters are increasingly rewarding incumbent state governments for good economic performance, although they caution that the effects are modest in magnitude. In other words, high growth does not guarantee re-election, but at least it increasingly means voters will not punish the governments of high growth states, as was the case in the 1990s. The reasoning would be that, in an era of rising growth and aspirations, voters would be more likely to reward good economic performance than in previous decades of economic stagnation. Although the data and methodology used are different, this, in spirit, matches the conclusion of Gupta and Panagariya.
It is important to note, however, a dissenting voice to this putative emerging consensus. My IDFC Institute colleague, Praveen Chakravarty, argues that studies based on using all or most of the states (to say nothing of the small Union territories), without adjusting for population or income share or in some other way weighting the states, are misleading, since there is such enormous variation across states. Small states that might be outliers in the language of statistics could be skewing the results.
Chakravarty’s preferred methodology is to consider only the 12 largest states—the four Hindi heartland states of Uttar Pradesh, Bihar, Madhya Pradesh and Rajasthan, the two western states of Maharashtra and Gujarat, the four southern states of Tamil Nadu, Karnataka, Kerala and (undivided) Andhra Pradesh, along with the two eastern states of West Bengal and Odisha. As he notes, these 12 states account for 80% of registered voters, 81% of Lok Sabha seats, 72% of state assembly seats, and 76% of national gross domestic product. Chakravarty has analysed data from all 38 state assembly and Lok Sabha elections in these 12 states from 2000 up until 2014.
When parsed this way, Chakravarty finds that incumbents were re-elected 15 times and defeated 23 times. There is no statistically significant difference between state growth rates in the cases of successful as against defeated incumbents; in fact, the relationship between growth and re-election appears to be random. Further, he has tested state growth rates against both re-election and vote share swings for the incumbent party: when weighted by each state’s population, no correlation is found.
While the research just described is as yet unpublished, preliminary results in a briefing paper by Chakravarty finds no correlation between re-election prospects and the consumer price index (CPI) inflation rate, thus finding no support for the folk wisdom that the price of onions is what determines election outcomes.
A Mint analysis of the 2014 Lok Sabha election results by Sumit Mishra and Pramit Bhattacharya also shows no apparent correlation between growth rates and the success of incumbent parties, using a version of the Gupta-Panagariya methodology.
If this appears to paint a confusing picture, the reader has every right to be confused. As we have seen, depending on the methodology and specific data that are used, it is possible to find support, or no support, for the intuitively appealing idea that good economic performance by incumbent governments is rewarded. It gets us no closer to determining if good economics helped play a role in Nitish Kumar’s re-election in Bihar two weeks ago.
The lack of definitive evidence in favour of the importance of economic growth may be disappointing for those who would like to believe that good economics is good politics. If governments themselves believe that delivering high growth will help get them re-elected, it will incentivize them to try to choose good economic policies in the hope of delivering high growth and thus getting re-elected, which of course every incumbent hopes for. If they believe that re-election is largely random or at any rate independent of economics, such an incentive may not exist.
The absence of the result we would hope to see in the data need not be as pessimistic as it appears. If there’s little or no correlation between growth and re-election prospects and therefore no incentive for politicians to be good economic stewards, equally there is no disincentive or fear of being punished, as appears to be the wrong lesson learned from the BJP’s 2004 defeat and the decade of populism and wasted reform opportunities seen under the Congress-led United Progressive Alliance.
One could hope, at least, that once elected or re-elected, for whatever reason, a party will do what is best for the economy or do its best more generally. Self interest may also collude with altruism or good luck. If a government believes that more rapid growth will, say, generate greater revenues for providing public services or patronage without resorting to the presumably unpopular move of raising taxes, it will pursue growth.
Thus, governments may be led, by hook or crook, to focus on sound economics, even if the jury is out on whether this will help them get re-elected. We shall see in the coming months and years if that logic prevails in the case of Bihar—and, indeed, of India more generally.
The author would like to acknowledge the valuable conversations on this subject with Praveen Chakravarty, Neelanjan Sircar and Milan Vaishnav without implicating any of them for the opinions expressed here.
Economics Express runs weekly, and features interesting reads from the world of economics and finance.
Never miss a story! Stay connected and informed with Mint. Download our App Now!!