Will higher turnout hurt the ruling party in Assam?

After Assam, Tamil Nadu also recorded its highest-ever voter turnout in the recently concluded assembly election. In West Bengal, where the election is still underway, an unprecedented number of people turned out to vote in the first phase as well. The big jump in turnout observed in these states has once again led many to speculate that a large number of people have voted to change the government.

 

In fact, such a notion has been around for a long time in Indian politics.It rests on the assumption that the additional number of people who turn out to vote in the succeeding election are driven by their frustration against those in power and as such vote against them. Had they been satisfied with the ‘state of affairs’, to further stretch the assumption, they would not have turned out to vote this time as they did not do so in the preceding election as well. The corollary that naturally follows from this is that the probability of the incumbent losing rises in proportion to the increase in turnout.

 

So, does higher turnout hurt the incumbent? Let’s see what three decades (1991-2021) of election data from Assam has to tell.

 

If state level voter turnouts and outcomes are compared between 1991 and 2021, a couple of cases would be observed where the election that saw a higher turnout also resulted in a change in government. For example, compared to the 1991 election, there was a three percentage points increase in turnout in 1996 and it also saw Asom Gana Parishad (AGP) dislodging the Congress from power. Again, the 2016 election saw turnout rise by over eight percent and it also resulted in the Congress being voted out of power in Assam. Most of those who argue that higher turnouts adversely impact the incumbent do so by making such kind of comparisons. While in the first glance this may seem plausible but ultimately such inferences are nothing more than fortuitous conjectures.

 

In the electoral system that India follows, both turnout and incumbent change at the national and state level are ultimately a product of what transpires at the constituency level. As such, whether turnout indeed impacts outcome can be plausible determined only after looking at trends at that level. Table 1 classifies constituencies of Assam according to turnout and incumbent party. On change. On the basis of it, a few observations can be made.

 

Firstly, between 1991 and 2021, a considerable number of constituencies that experienced a higher turnout compared to the previous election also witnessed a change in the incumbent party. For example, except for 2016-2021, constituencies that have seen a turnout increase of under 2 percent also demonstrate a probability of incumbent party change that is over 50 percent. This should not however lead anyone to hastily conclude that higher turnout does indeed produce anti-incumbent outcomes. This is because, as the same data point reveals, there is an almost equal probability that incumbent party did not change in around the half the constituencies despite seeing a similar level of turnout increase.

 

Secondly, and even more importantly, the probability of incumbent party change does not proportionately increase with a proportionate increase in voter turnout. Take the case of 2001 -2006. There is a 51.85 percent probability of incumbent party change when there is an under two percent turnout increase. It increases to 66.67 percent when turnout rises to two to four percent before falling and then rising again. If higher turnout indeed led to anti-incumbent outcomes, the probability of change should have kept increasing as the turnout percentage went up. This severely weakens the commonly held notion.

 

Lastly, there are numerous instances of incumbent party change even when turnout has decreased compared to the last election. In 1991-1996, in constituencies where turnout dipped, the probability of incumbent party change stood at 42.86 percent. Five years later, that figure stood at 53.54 percent.

 

The data from Assam’s assembly constituencies show that the relationship between higher turnout resulting in incumbent change is much weaker than that what is often made out to be. Incumbent party change takes places at the local level due to a variety of reasons. Variations in turnout are thus not a reliable indicator in this regard as the corresponding outcomes are neither uniform nor proportionate. At best, only a limited claim may be made that compared to a decrease in turnout, the chances of incumbent party change are somewhat better when there is an increase in turnout.

 

Table 1: Relationship between Turnout Change &Electoral Outcomes

Year	Turnout change	Total constituencies	Probability of change of incumbent party (%)
1991-1996	Under 2%	29	55.17
	2% to under 4%	28	53.57
	4% to under 6%	24	50.00
	6% and above	27	51.85
	Decrease in turnout	14	42.86
1996-2001	Under 2%	15	73.33
	2% to under 4%	5	40.00
	4% to under 6%	1	100.00
	6% and above	2	50.00
	Decrease in turnout	99	53.54
2001-2006	Under 2%	27	51.85
	2% to under 4%	27	66.67
	4% to under 6%	19	31.58
	6% and above	11	45.45
	Decrease in turnout	42	45.24
2006-2011	Under 2%	32	62.50
	2% to under 4%	20	65.00
	4% to under 6%	10	60.00
	6% and above	7	71.43
	Decrease in turnout	57	49.12
2011-2016	Under 2%	2	50.00
	2% to under 4%	8	75.00
	4% to under 6%	19	68.42
	6% and above	97	70.10
	Decrease in turnout	0	-
2016-2021	Under 2%	17	35.29
	2% to under 4%	4	25.00
	4% to under 6%	0	0
	6% and above	1	0
	Decrease in turnout	104	34.62

Note: The table should not be used anywhere else apart from this article without the permission of the author.