Karnataka Elections: Spoils of Fragmentation

One of the questions that forms the backdrop to Karnataka’s elections is the impact of former chief minister B.S. Yeddiyurappa’s Karnataka Janata Party (KJP) and mining baron and former minister B. Sriramulu’s BSR Congress on the performance of the BJP, the Congress and the Janata Dal (Secular) [JDS]. Do established parties lose vote share when new parties enter the political arena, affecting some established parties more than others, irrespective of who the entrant views as its primary competitor? What does the electoral history of Karnataka tell us about new parties that pose credible threats?
Yeddiyurappa has been a thorn in the BJP’s flesh for some time now, and his departure has clearly had an impact on the strategies of his loyalists and opportunists on when and how to jump ship. BSR Congress seems to have a lot of money, and while money cannot guarantee success, it definitely does not guarantee failure.

As a preliminary exploration of these questions, we used electoral data from 1978 to 2004 to trace the relationship between vote share and party fragmentation. Political scientists use a measure of party fragmentation called the Effective Number of Parties index (ENP) to answer such questions rather than a simple count of the number of parties because not all parties get equal vote shares. If four parties in a constituency get 25% vote share each, the ENP is four and the simple count is four; but if one party gets 50% and the other three parties split the remaining 50% equally among themselves, using the formula, the ENP will be at three even if a simple count still suggests four.
In the four graphs, the party fragmentation (ENP) is on the horizontal axis and party vote shares on the vertical axis.  Each dot represents a constituency in an election (1978-2004), and the lines of best fit show the average relationship. Since our inferences are based on 1978-2004 data (prior to the latest delimitation), they tell us how fragmentation in the past has affected party performance in Karnataka.
The figures show that on average the BJP neither gains nor loses vote share as party fragmentation increases. In contrast, the Congress and the JDS tend to lose vote share as party fragmentation increases, and the rate of loss for the Congress is greater than for the JDS. And lastly, independents gain the most with party fragmentation.
The figures are suggestive along several lines. Since the BJP’s vote share is relatively stable irrespective of party fragmentation, it suggests the presence of a relatively strong cadre and group of “core” voters. This suggests that, irrespective of fragmentation, the BJP is here to stay in Karnataka, unlike in other states of south India. More precisely, whoever captures the cadre and organisational base of the party, the BJP will be around.
Second, the figures suggest that the other two major parties (Congress and JDS) lose votes when new entrants demand their slices of the pie, indicating that they have a relatively greater share of “swing” voters compared to “core” voters. Historically, the Congress has been the predominant party, and therefore, is also the one with the most to lose because of new entrants. The JDS is also at a disadvantage when faced with new entrants, but its rate of loss is not as steep as the Congress's, perhaps because of its relatively smaller geographical spread.
Finally, independent candidates gain the most with party fragmentation, as common sense would suggest. When there are anywhere between 3 and 4 effective numbers of parties, the chances are highest that an independent [or a group of independents] would gain 20 percent of the votes. It is also interesting to note that at low levels of party fragmentation or ENP, the vote share of independents is close to zero and only rises when the ENP is more than three – which means that independents do better when there are three or more effective parties and have no impact otherwise.
What can these conclusions tell us of broad trends that could have a say on the outcome of the ensuing elections in Karnataka?
The BJP's performance in the elections depends on who controls the cadre and the group of “core” voters. The party’s votes are historically unaffected by the entrance of new parties, but this time at least one of the two break-away factions is from the organisation’s own cadre. So this election might turn out to be the make or break election for the BJP. If its breakaway factions have weaned away some of the cadre, fragmentation will certainly undermine the BJP’s performance, and thereafter political bargains will decide whether the BJP decides to toe the line of its provincial leaders. If its breakaway factions have not managed to take away its cadre, then the BJP has not lost everything in state.
The data suggests that the Congress is the most vulnerable to ENP, and public opinion seems to project the party as the likely majority party this time around. If the Congress wins the highest number of seats leading to a majority, those victories would be based on comparably thin margins, considering electoral fragmentation. And so they might have to work much harder at governance to ensure a second term or at the very least avoid coalitional rotations in the same term. Otherwise the party's victory might only add fuel to future anti-incumbent sentiments.
The JDS also loses vote share as fragmentation increases, but unlike for the Congress, its decline trend line remains flat when the ENP is between 3 and 4. The party faces losses with fragmentation like the Congress, but only up to a point since it also has certain strongholds where, beyond a certain level of fragmentation, its share tends to remain constant.
These trends also show that if  three- or four-way contests are here to stay, the party cadre of the BJP and the JDS tend to become fairly stable over time. But the Congress seems to suffer from over reach and needs to reassess which constitutencies it contests from to avoid losses. It needs to look at its constituencies and cadre a lot closer to avoid volatility.
Independents, on the other hand, should possibly target seats where the number of parties are fairly well established so as to capture swings. This may sound counter intuitive, but what the graph suggests is that independents would be far more successful in gaining higher vote shares in electorally competitive constituencies rather than in strongholds of big names.