Privileging One’s Own? Voting Patterns and Politicized Spending in India

• Author: Francesca R. Jensenius,Pradeep Chhibber
• DOI: http://doi.org/10.1177/00104140221109430
• Published date: 01 March 2023
• Date: 01 March 2023
• Subject Matter: Articles

Article

Comparative Political Studies

2023, Vol. 56(4) 503–529

© The Author(s) 2022

Article reuse guidelines:

sagepub.com/journals-permissions

DOI: 10.1177/00104140221109430

journals.sagepub.com/home/cps

Privileging One’s Own?

Voting Patterns and

Politicized Spending in

India

Francesca R. Jensenius

1

and Pradeep Chhibber

2

Abstract

How do politicians allocate public resources? Despite the extensive literature

on distributive politics, we have limited micro-level evidence for why and

under what circumstances politicians choose various allocation strategies.

India’s discretionary constituency development scheme (MPLADS) provides

an excellent opportunity to study the spending choices of individual politi-

cians. Drawing on an original dataset linking voting patterns across 227,507

villages in the 2009 general elections to MPLADS allocations 2009–2014, we

f‌ind that politicians generally channel more projects and resources to villages

that vote for them in higher numbers. We then leverage a natural experiment

created by the delimitation (redistricting) of electoral boundaries in 2008 to

provide causal evidence that spending choices are driven by short-term

electoral incentives. Finally, we show that allocation patterns differ by the

type of party that has brought a politician to power. Our f‌indings contribute to

the growing literature on heterogeneity in politicians’distributional choices.

Keywords

India, political economy, distributive politics, clientelism, natural experiment

1

University of Oslo, Norwegian Institute of International Affairs, Oslo, Norway

2

University of California Berkeley, Berkeley, CA, USA

Corresponding Author:

Francesca R. Jensenius, Professor of Political Science, University of Oslo, and Research Professor,

Norwegian Institute of International Affairs, Oslo, Norway.

Email: f.r.jensenius@stv.uio.no

A sizable literature on distributive politics examines how politicians allocate

the public resources under their control. There is a robust debate about

whether politicians are more likely to direct resources towards their core

supporters, or to competitive areas where they may attract new voters.

1

It may

seem more strategic for politicians to focus on persuading “swing”voters to

support them, but when coordination and voter mobilization are taken into

account, the core-voter argument is strengthened (Cox, 2010). A review of

empirical studies shows that most, though not all, parties and politicians seem

to be allocating disproportionately to areas that vote for them in higher

numbers (Stokes et al., 2013). However, we still have limited micro-level

evidence for why and under what circumstances politicians choose different

allocation strategies.

The emergence of constituency development funds over which politicians

have considerable discretionary power makes it possible to study distribu-

tional choices more directly, and at a granular level (Tshangana, 2010). In this

article, we examine the spending choices of 232 Members of Parliament

(MPs) in India under the Member of Parliament Local Area Development

Scheme (MPLADS). Under this scheme, MPs have access to yearly funds that

they can allocate to any development activities they deem appropriate within

their respective electoral districts (“constituencies”in India).

2

As MPs have

total discretion over the allocation of funds under this scheme, it is ideal for

studying the choices of individual politicians.

Drawing on original datasets manually linking both polling-station-level

election returns and MPLADS projects to census villages, we identify three

main patterns. First, across 227,507 villages we f‌ind a strong positive as-

sociation between support for the winning candidate in the 2009 general

elections and MPLADS allocations 2009–14. Although in the raw data there is

a curvilinear relationship between support for the MP and funding

allocation—a pattern indicating that MPs give more to competitive areas—

when we examine the distribution of projects within constituencies and control

for obvious confounders, such as village population size, the curvilinear

pattern is weakened substantially. This contradicts spending patterns under a

similar scheme in Kenya (see Harris & Posner, 2019), but is consistent with

work on India by Bussell (2019), as well as the general expectation in India

that politicians will provide resources to “their own”when they win an

election.

Second, to shed light on why we f‌ind this positive association between

village-level voting patterns and MPLADS spending, we turn to a natural

experiment created by the delimitation (redistricting) of electoral boundaries

in 2008. The redistricting resulted in many villages being moved from one

constituency to another one. These changes were announced several years

prior to the 2009 elections, in many cases rendering the affected villages

electorally unimportant for the sitting MP. If politicians were distributing their

504 Comparative Political Studies 56(4)