ISSN (Print) - 0012-9976 | ISSN (Online) - 2349-8846
-A A +A

The AAP Effect

Decoding the Delhi Assembly Elections

Chitranshu Mathur (chitranshu.m@gmail.com) is with the Indian Institute of Management, Bangalore.

A scrutiny of voting trends in Delhi over the last two decades and an analysis of the spectacular debut of the Aam Aadmi party may provide clues to the challenges and opportunities for other political parties, particularly the Bharatiya Janata Party and Congress party in the forthcoming Lok Sabha elections.

The electoral debut of the Aam Aadmi Party (AAP) in the Delhi assembly elections in December 2013 has attracted significant positive as well as negative commentary. On the positive side, it is said that AAP’s victory represents the start of a new era in Indian politics, as this fledgling party of activists has ousted the Indian National Congress (INC) from Delhi after 15 years in power, and given the Bharatiya Janata Party (BJP) a run for its money. Meanwhile, critics say that AAP’s rapid rise may be followed by an equally rapid decline, or worse, that it is a conspiracy by one of the two major parties to divide votes of the other.

This is an attempt to look into the drivers and implications of AAP’s electoral performance, not just in Delhi but in national politics as well, now that AAP has announced that it will contest the Lok Sabha elections this year from as many seats as possible. There has also been a lot of discussion about AAP’s electoral promises, ideology, policies and actions since it has assumed power in Delhi, but those will not be discussed here.

Pre-election Surveys and Actual Results

The opinion polls (except the parties’ internal ones) held in Delhi between September to November 2013 had predicted 6 to 25 seats for AAP (out of a total of 70 seats), 22 to 36 seats for the BJP and 19 to 37 seats for INC. When the actual election results were declared, AAP got 28 seats, the BJP got 31 and INC got just eight. While the BJP performed largely as predicted, AAP performed far better than any expectations except its own, and the INC fared much worse. Besides these 67 seats, one seat went to the Shiromani Akali Dal (SAD), an ally of the BJP, one seat to the Janata Dal (United) [JD(U)], a former ally of the BJP, and one seat to an independent candidate.

Drivers of AAP’s performance

While AAP’s performance can be attributed to several factors, such as the support they gathered during the anti-corruption movement, their efforts in grassroots mobilisation, the general disenchantment with traditional political parties and with the incumbent INC government in particular, and the growing political participation and expectations of the youth, it is important to quantify such factors to draw any reliable conclusions. Therefore, we will try to answer three related questions:

      ‒What are the different types of voters who voted for AAP?

      ‒How these voters would have voted if AAP had never existed, and in that case, what would the Delhi assembly               election results look like

      ‒Given these answers and the developments since the Delhi assembly election results, what will be the likely impact       of AAP on the upcoming general elections?

To answer these questions, we list five possible categories of voters for AAP. If it were not for AAP, these voters may have:

‒mostly voted for the INC. They may have been typically centrist or left-of-centre or from certain groups which are traditionally INC voters and not attracted to the right-of-centre BJP.

‒mostly voted for the BJP. They may have been typically right-of-centre or right-wing but did not find the BJP or its state-level leaders in Delhi as attractive as Arvind Kejriwal and his team.

‒“swing voters” who may have voted either for the INC or BJP, depending on various factors like anti-incumbency, specific policies or leaders and the prevailing public opinion. Many voters from the previous and this category may have been  those  who wanted Arvind Kejriwal as Delhi’s chief minister and Narendra Modi as prime minister.

‒voted for parties other than the INC and the BJP, e.g. Bahujan Samaj Party (BSP), Samajwadi Party (SP), the communist parties (CPI and CPM), other parties and independents. They may have been disenchanted with the INC and BJP or were actively courted by one of the other parties, e.g. dalits who had hitherto voted for the BSP.

‒not voted at all because of disenchantment with the traditional political parties. They were mainly influenced by AAP’s fresh approach to politics.

We shall now try to ascertain what percentage of AAP’s votes came from each of these types of voters.

Trends in Delhi Assembly Elections, 1993-2013

We start by looking at the performance of major parties in Delhi assembly elections since 1993.

Table 1.1: Seats and Vote Shares

 

1993

1998

2003

2008

2013

 

Seats

Vote Share

Seats

Vote Share

Seats

Vote Share

Seats

Vote Share

Seats

Vote Share

AAP

-

-

-

-

-

-

-

-

28

29.49%

INC

14

34.48%

52

47.76%

47

48.13%

43

40.31%

8

24.55%

BJP

49

42.82%

15

34.02%

20

35.22%

23

36.34%

31

33.07%

BSP

0

1.88%

0

3.09%

0

5.76%

2

14.05%

0

5.35%

SP

0

0.50%

0

0.48%

0

0.65%

0

0.49%

0

0.22%

CPI + CPM

0

0.59%

0

0.21%

0

0.13%

0

0.13%

0

0.11%

Other Parties

4

13.81%

1

5.74%

2

5.25%

1

4.76%

2

3.65%

Indepen-dents

3

5.92%

2

8.70%

1

4.86%

1

3.92%

1

2.93%

None of the above (NOTA)

-

-

-

-

-

-

-

-

 

0.63%

TOTAL

70

100%

70

100%

70

100%

70

100%

70

100%

The data in the table is sourced from The Election Commission of India website. The link http://eci.nic.in/eci_main1/ElectionStatistics.aspx  provides data for the period 1993 to 2008, and the link  http://eciresults.nic.in/ provides data for the year 2013. 

Table 1.2: Number of Votes, Candidates and Voting Percentage

 

1993

1998

2003

2008

2013

AAP

-

-

-

-

2,322,417

INC

1,224,361

1,952,071

2,172,062

2,489,816

1,933,020

BJP

1,520,675

1,390,689

1,589,323

2,244,629

2,604,320

BSP

66,796

126,197

259,905

867,672

420,931

SP

17,717

19,418

29,224

30,073

17,042

CPI + CPM

21,230

8,633

5,888

7,878

8,307

Other Parties

490,383

234,645

237,245

295,276

287,759

Independents

210,086

355,773

219,225

242,000

230,427

NOTA

-

-

-

-

49,892

TOTAL

3,551,248

4,087,426

4,512,872

6,177,344

7,874,115

Number of candidates

1,316

815

817

875

810

Voting percentage

61.75%

48.99%

53.42%

57.58%

66.17%

The data in the table is sourced from The Election Commission of India website. The link http://eci.nic.in/eci_main1/ElectionStatistics.aspx  provides data for the period 1993 to 2008, and the link  http://eciresults.nic.in/ provides data for the year 2013. 

The BJP won comfortably in 1993, but presided over a turbulent tenure with three chief ministers. When the Sheila Dixit-led INC swept into power in 1998, the BJP’s seats and vote share were more or less the same as what the INC had obtained in 1993.  The INC also probably gained votes in 1998 due to fragmentation of the Janata Dal in the mid-1990s. In 2003, the BJP gained 1.2% vote share and five seats, and in 2008, it gained 1.12% vote share and three seats. The INC marginally increased its vote share in 2003 and then lost nearly 8% to BSP in 2008. It also lost five seats in 2003 and four seats in 2008.

Meanwhile, the changes in party affiliation of the seats was not very high. About 40% of seats remained with the same party in 1998, despite the huge swing from the BJP to INC, and[AB1]  about 75% of seats remained with the same party in 2003 as compared to 1998. Delimitation changed seat boundaries in 2008, but even among those which did not change about 80% of the seats were retained by the party that had won in 2003. Thus the political scenario in Delhi remained relatively stable for two decades, with many legislators getting continuously re-elected since 1993. Even the rapid rise of the BSP from 1993 to 2008 had seemingly reached a plateau, as its core vote-bank of dalits formed ~17% of Delhi’s population, and it faced an electoral reversal in the neighbouring state of Uttar Pradesh in 2012 assembly elections. We also note that the total voting percentage increased by ~4% in 2003 and 2008.

Scenario Without AAP, and Where AAP Got its Votes From

To arrive at an alternate scenario of the 2013 elections if AAP had not existed, we make the following assumptions:

‒The total voting percentage, following the trend from 1998 to 2008, increases to 61.75% in 2013, i.e. equal to the previous high of 1993.

‒The NOTA option is exercised by 1.90% of voters (the same percentage as in Rajasthan and Madhya Pradesh), instead of the 0.63% who actually used it with AAP in the fray.

‒The vote share of the BSP, SP and CPI + CPM remains stable at 14.05%, 0.49% and 0.13% respectively.

‒The vote share of “other parties” and “independents” continues to spiral  downwards as seen over the  years and is 4.27% and 2.98% respectively.

‒The voters who would mostly vote for the INC represent 34.48% of the total votes, i.e. the minimum vote share of INC from 1993 to 2008.

‒The voters who would mostly vote for the BJP represent 34.02% of the total votes, i.e. the minimum vote share of the BJP from 1993 to 2008.

‒The “swing voters” make up the remaining 7.68% of votes.

Given these assumptions, we have the following scenario, where we also juxtapose the actual votes from 2013 and the difference:

Table 1.3: Scenario Without AAP

 

Scenario without AAP

Actual 2013 results

Difference in votes

 

 

Number of votes

Vote share

Number of votes

Vote share

 

AAP

-

-

2,322,417

29.49%

 

 

INC

2,533,677

34.48%

1,933,020

24.55%

600,657

 

BJP

2,499,875

34.02%

2,604,320

33.07%

-104,445

 

Swing voters (INC or BJP)

564,346

7.68%

-

-

564,346

 

BSP

1,032,429

14.05%

420,931

5.35%

611,498

 

SP

36,006

0.49%

17,042

0.22%

18,964

 

CPI + CPM

9,553

0.13%

8,307

0.11%

1,246

 

Other Parties

313,770

4.27%

287,759

3.65%

26,011

 

Independents

218,978

2.98%

230,427

2.93%

-11,449

 

NOTA

139,617

1.90%

49,892

0.63%

89,725

 

TOTAL

7,348,250

100%

7,874,115

100%

525,865

 

Voting percentage

61.75%

 

66.17%

 

 

The data in the table is sourced from The Election Commission of India website. The link http://eci.nic.in/eci_main1/ElectionStatistics.aspx  provides data for the period 1993 to, and the link  http://eciresults.nic.in/ provides data for the year 2013. 

From the above table (1.3), we begin to see what the probable composition of AAP’s voters is. To make it clearer, in the following table (1.4) we combine BSP, SP, CPI + CPM, other parties, independents and NOTA in a single category called “Others”, and focus only on the “difference in votes” column. Here the row for “TOTAL” is now called “New Voters”.

Table 1.4: How Each Category of Voters Differed from the Scenario Without AAP

 

Difference in Votes

INC

600,657

BJP

-104,445

Swing voters (INC or BJP)

564,346

Others

735,995

New Voters

525,865

The data in the table is sourced from The Election Commission of India website. The link http://eci.nic.in/eci_main1/ElectionStatistics.aspx  provides data for the period 1993 to 2008, and the link  http://eciresults.nic.in/ provides data for the year 2013. 

These five rows add up to 2,322,417, which is exactly the number of votes received by AAP in 2013. Since the “swing voters” represent those who could have voted for either the INC or BJP, we offset the difference of -104,445 for the BJP against this category, i.e. 104,445 of the “swing voters” actually voted for BJP even when AAP was present as an option. Thus, we get the following table (1.5), where we also show the percentage of AAP’s votes coming from each category of voters:

Table 1.5: Number and Percentage of AAP’s Votes From Each Category

 

Difference in Votes

Percentage of AAP’s votes

INC

600,657

25.86%

BJP

0

0%

Swing voters (INC or BJP)

459,901

19.80%

Others

735,995

31.69%

New Voters

525,865

22.64%

The data in the table is sourced from The Election Commission of India website. The link http://eci.nic.in/eci_main1/ElectionStatistics.aspx  provides data for the period 1993 to 2008, and the link http://eciresults.nic.in/ provides data for the year 2013. 

Thus, we can see that 22.64% of AAP’s voters were probably those who may not have voted at all if not for AAP’s presence in the elections. Another 31.69% represent those who probably voted for other parties besides the INC and BJP in the last election, mostly for the BSP. Another 25.86% of votes came from traditional INC voters. The balance 19.80% came from “swing voters” who would have otherwise voted either for the INC or BJP depending on various factors. Meanwhile, no votes for AAP came from traditional BJP voters.

We can also look at the impact of AAP on each of the parties and categories of voters by finding the percentage of votes lost to AAP as follows:

Table 1.6: Impact on Other Parties Due to AAP

 

Votes in Scenario without AAP

Votes lost to AAP

% of votes lost to AAP

INC

2,533,677

600,657

23.71%

BJP

2,499,875

0

(104,445 swing votes gained)

0%

(18.51% of swing votes gained)

Swing voters (INC or BJP)

564,346

459,901

81.49%

Others

1,750,353

735,995

42.05%

The data in the table is sourced from The Election Commission of India website. The link http://eci.nic.in/eci_main1/ElectionStatistics.aspx  provides data for the period 1993 to 2008, and the link  http://eciresults.nic.in/ provides data for the year 2013. 

(Note that the “Others” category has wide differences among its constituents, e.g. while 50-65% of the votes from BSP, SP and NOTA moved to AAP, the effect on CPI + CPM, other parties and independents was less than 15%).

How closely do these results represent the reality, and how realistic are the assumptions made above? While it is difficult to answer this question precisely, one can make a reasonable attempt by comparing actual constituency-wise data from 2008 and 2013 for three major parties, and noting the following statistics:

Table 1.7: Changes in Votes Per Constituency Received by Leading Parties

 

INC

BJP

BSP

 Mean change in votes received per constituency

-7,954

+5,719

-6,382

 Median change in votes received per constituency

-9,126

+3,151

-5,012

 Mean change in votes received where party was incumbent

-8,469

+1,564

-19,835

 Mean change in votes received where party was opponent

-7,135

+7,753

-5,986

The data in the table is sourced from The Election Commission of India website. The link http://eci.nic.in/eci_main1/ElectionStatistics.aspx  provides data for the period 1993 to 2008, and  the link http://eciresults.nic.in/ provides data for the year 2013. 

Thus, we can see that while the INC lost votes across most constituencies whether or not it was the incumbent, the BJP gained votes in most constituencies, and its performance where it had lost in 2008 significantly improved relative to the seats that it had won in 2008. The BSP on the other hand lost a large number of votes in both constituencies that it won in 2008 and lost a lesser, but still significantly large, number of votes across the rest of the constituencies. This confirms the alternative scenario and its analysis above to some extent.

Conclusions

This aggregate-level analysis hides some important differences, especially across economic classes and constituencies, for e.g. that AAP gained the most in inner Delhi constituencies, especially ones with a higher number of slums (Ramani: 2013) and did not fare so well in outer Delhi. It is also difficult to ascertain exactly how valid is the assumption that “swing voters” are only 7.68% and the “traditional” voters of the INC and BJP are 34%, or whether the reality is different, which would in turn affect the subsequent analysis and also any conclusions about how many seats and votes did AAP gain from the INC or BJP.

However, this is just an attempt to understand electoral trends based purely on historical and recent election data. In the case of AAP in Delhi, it can be said that more than half of their votes probably came from voters who would not have voted otherwise or would have voted for non-INC, non-BJP options. The remaining voters were most likely either “swing voters” or “traditional” INC voters, and the impact on BJP’s “traditional” vote base was minimal, if at all. AAP’s presence also ensured that far fewer voters opted for the newly introduced NOTA option, and the total number of candidates in the fray was lesser than it had ever been. In a way, this represents a consolidation of opinion among all those who have been hitherto disenchanted with mainstream electoral politics. AAP’s performance in Delhi has also established its “winnability”, which could in turn attract voters who hitherto  did not  want to “waste” votes on parties they thought would lose.

Since the Delhi elections, AAP’s actions and rhetoric have pitted it more strongly against the BJP than any other party, and thus, at the national level, one can expect a consolidation of BJP’s voters on one hand and a division of the remaining voters among AAP, INC and various other parties. The eventual results may depend on various state-level factors, and the vagaries of the first-past-the-post system make it even more difficult to make any predictions, but if the rise of AAP continues, one can expect its impact on existing parties to be in similar proportions as discussed above. In this manner, AAP may fundamentally alter India’s politics in the months and years to come.

References

Ramani, Srinivasan (2013): “The Aam Aadmi Party's win in Delhi: Dissecting it through Geographical Information Systems.” Economic and Political Weekly. Vol. XLVIII No. 52, 28 December 2013, available at http://www.epw.in/web-exclusives/aam-aadmi-partys-win-delhi-dissecting-it-through-geographical-information-systems.htm?ip_login_no_cache=e660f768be3113afbf87a3717e0e9f80, accessed on 13 February 2014. 

 

Comments

(-) Hide

EPW looks forward to your comments. Please note that comments are moderated as per our comments policy. They may take some time to appear. A comment, if suitable, may be selected for publication in the Letters pages of EPW.

Back to Top