The road to Election 2014 has now become curved with the advent, and the Delhi-assembly elections success, of the AAP. With the AAP's intention to contest at least 300 seats (maybe, all 543!) in the forthcoming Lok Sabha elections, the Congress and the BJP appear to be a worried lot. This discomfort stems from the huge surprise victory of the AAP in the recent Delhi assembly elections, where the Congress suffered a massive thrashing by the jhadoo (broom, the AAP's symbol). It is not that the AAP won, but that everybody else lost; hence, its victory is in quotes. Ultimately, the question on everybody's mind is, "How many seats will the Aam Aadmi Party win in 2014"
Regardless of the seats it might win, the AAP might hurt either one or both of the major parties/alliances. The fact remains that the AAP is an important force in the forthcoming elections, especially in terms of the vote share they will be able to obtain. Their overall effect is analysed in this first of two articles; vote shares in the first part, and the all important seat 'estimates' appearing tomorrow (January 25, 2014). It bears emphasis that the estimates are not forecasts; these estimates are not based on opinions or opinion polls, but rather on assumptions generated by Indias electoral history.
Conventional analysis looks at swings in vote shares. A near-equivalent way of looking at vote swings is via vote give-ups. The vote for a new entrant like the AAP has to come from the existing pool of voters; from an existing Congress/BJP/regional party/independent voter. Existing here is defined as the 2009 elections. The give-up of vote share is the percentage of vote share lost by a party with respect to its previous election performance. For example, if the Congress obtained 40% of the vote in a constituency in 2009, then a give-up of 40% to the AAP implies that in that constituency, the Congress will obtain only 60% of the vote, i.e., 24 percentage points, and the AAP will gain the remaining 16 percentage points from the Congress.
All analysts are constrained by historythere is no parallel to the unique formation of the AAP. There have been instances of new political parties pulling off surprising victories (e.g., NT Rama Rao in Andhra Pradesh in 1983); however, hardly any regional party has gone on to make an impact at the national level.
What give-up can we expect from either the Congress or the BJP to the AAP And, will both be the same What about the possibility that there will be give-up from the AAP to the BJPthe Modi effect Recall that the Delhi opinion polls had suggested that at least a third of the AAP voters might vote for Modi.
Looking at past assembly elections in terms of give-ups of the Congress party, Kejriwal's sweeping victory over the Congress is one of the highest on record with an average give-up of 39%, i.e., out of every 100 voters of the Congress in the 2008 Delhi assembly elections, 39 voted for the AAP in 2013. In the case of BJP, the give-up was only 9 out of 100. In the case of the remainder (BSP and independents), the give-up was 50 out of 100. Historically speaking, in terms of assembly elections, the Delhi Congress give-up was one of the worst in history, the worst being the 60% give-up to Asom Gana Parishad in the 1985 Assam elections. For Lok Sabha elections, the worst Congress give-up was in 1996, when, on average, 24 out of 100 Congress voters voted for other parties. (Note that in the 1977-debacle, the Congress give-up average was 19%.)
This provides us with a historical basis for analysing the AAP's possibilities in 2014. The attempt is to construct scenarios which guarantee the AAP the largest possible share of theoretical votes. Towards this end, the following assumptions are made. The Congress gives up a minimum of 20% and a maximum of 40% of the vote share obtained in the 2009 election. If the Congress did not contest a seat in 2009, it obtains zero votes from the Congress in that constituency; if it obtained 60% of the vote share (as it did in Delhis Chandni ChowkCC, herafterconstituency), and the assumed Congress give-up is 40%, then the AAP obtains 24% of the vote. In addition, in this base estimation, the AAP obtains 10% of the BJP vote. In CC, the BJP obtained 34% in 2009, so the AAPs vote share goes up to 27.4%; finally, the independents and third parties yield 50% to the AAP; so the vote share of the three major parties (the Congress, the BJP and the AAP) in CC is 36, 30.6, and 30.4, respectively. The surprising conclusion, in this scenario (a replica of the Delhi Assembly 2013 give-ups), the Congress wins in Chandni Chowk!
Of course, different give-ups will yield different results; in the above example, if the Congress gave up 45% of its vote share, and the BJP 15%, the AAP would win in Chandni Chowk. These variations are experimented with in 12 different simulations of vote shares. Seven of these simulations have the following range of assumptionsthe Congress give-up: 30-35%; BJP give-up: 0-10%; and other parties give-up, at an average of 20%.
The results are revealing. The average vote shares obtained by the three parties are all very close to each other. The accompanying chart reports the national vote shares of the two major parties for all elections since 1989. However, and this is where the AAP votes in 2014 may meet electoral reality, there is virtually no meaningful relationship between the vote shares and the seats obtained by each party. In 1989, Congress obtained one of its largest vote shares (37.7%), yet it obtained fewer seats than in 2009 (vote share of 26.1%!). As is well known, this lack of a relationship is due to two factorsmost importantly, the first-past-the-post electoral system and coalition politics. The latter may have made us forget the Lahiri-Roy formulation of IOU (Index of Opposition Unity) over the last two decades. By making it a three-party/coalition fight, the AAPs entry brings back IOU with a vengeance.
An alternate electoral system prevails in many European countries and in some developing countries, e.g., Brazil. This is the proportional voting system, where seats obtained are proportional to the vote obtained. If this system had prevailed in India, then in 1999, Congress would have formed the government. (As it happened, Congress got its lowest ever seats114in 1999).
This provides a perspective on the projected vote shares for the three parties in 2014, and the seats they will obtain. The projections suggest that it will be a very close fight, given our assumptions, for votes. On average, the AAP obtains the lowest at 17.7%, the Congress 18.8%, and the BJP the highest, at 19.1%. So, if proportional representation was in existence today, the AAP would gain near identical seats to the BJP and the Congress. But with the first-past-the-post systemwait till tomorrows article. HintIOU makes a big dent to common assumptions about the AAPs seat popularity.
Surjit S Bhalla
First of two columns based on work with Prasanthi Ramakrishnan
and Sriramjee Singh. The author is chairman, Oxus Investments
and a senior advisor to Zyfin. Twitter: @surjitbhalla