Explaining intra-party organizational variation at the sub-national level in India

ABSTRACT

Existing scholarship on democratic theory has emphasized the role of political party's organization in establishing institutionalized party systems. If an intra-party organization is understood to be a key explanatory variable, then it is important to explain why some political parties have a strong internal organization in comparison to others. However, there is no single study, of which I am aware, that explains variations in the level of organizations within political parties in India. Using Chhibber, Jensenius and Suryanarayan's party organization dataset alongside a qualitative example from the sub-national level in India, I show that there exists a ‘U’ shaped relationship between parties controlling the executive office and the strength of individual party organization. There exists a strong negative effect between time spent in the executive office and individual party organization. It is only in the very long run that parties in government reap the small benefits of the executive office for their individual party organizations.

KEYWORDS:

• India
• party organization and executive office

Acknowledgments

I would like to thank Oliver Heath, Arvind Kumar, Pranav Gupta, Chris Hanretty and the Quantitative Politics Reading Group at Royal Holloway, University of London, for their comments on an earlier version of this paper. I am also extremely grateful to Louise Tillin and both the anonymous reviewers for their detailed and extremely constructive feedback. The article is significantly improved because of their valuable time and input.

Table A2. Party organization and political defections.

Table A3. Determinants of strong party organization (discounting junior coalition members).

Table A4. Determinants of strong party organization (continuous executive office).

Table A5. Party organization and alternative explanations.

Table A6. Party organization as an ordinal variable.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 It is important to mention that the measure of executive office years is a cumulative one as opposed to a continuous one.

2 The Indian states included in their dataset are the following: Andhra Pradesh, Assam, Bihar, Gujarat, Haryana, Karnataka, Kerala, Madhya Pradesh, Maharashtra, Odisha, Punjab, Rajasthan, Tamil Nadu, Uttar Pradesh and West Bengal.

3 As an additional robustness check on Chhibber, Jensenius, and Suryanarayan (2014) party organization dataset, I test the association between Chhibber, Jensenius, and Suryanarayan (2014) party organization scores and the proportion of defections from an individual political party. If, indeed, Chhibber, Jensenius, and Suryanarayan (2014) party organization dataset measures the strength of individual party organization, then their scores should correlate negatively with the proportion of defections from the individual political party. The results for the tests are displayed in Table A2 in the appendix.

4 As a further robustness check, I present findings in Table A3 where I show that the main results are robust to excluding executive office years accumulated because of junior coalition membership.

5 Here, it is important to mention that my measure of the executive office is a cumulative measure of executive office years as opposed to a continuous one. Theoretically speaking, we should expect the continuous executive office measure to have a stronger negative effect on a party's organization in comparison to a cumulative executive office measure. This is because, in a cumulative executive office measure, parties will have incentives to strengthen their organization when they are out of power. This opportunity is not present for parties that are continuously in power. See Table A4 in the Appendix where I replicate Table 1 but with a continuous executive office measure as opposed to a cumulative one.

6 Please see Table A5 in the appendix.

7 The results do not change substantively if I switch an individual party's voteshare with seatshare.

8 See Table A6.

9 The threshold value is the value of the independent variable at which the impact of the independent variable on the dependent variable changes signs.

10 The state-election year dummy variable subsumes the need for including state and year fixed effects because of perfect collinearity.

11 I include Bharatiya Jana Sangh (BJS) along with the BJP.

12 I include the CPI along with the (Marxist)-CPI (M) as the organizational differences between the two are more a matter of degree than of kind (Nossiter 1988).

Additional information

Funding

This work was supported by Economic and Social Research Council [ES/T007451/1] and The Leverhulme Trust [ECF-2020-381].