The Political Economy of Government Responsiveness:
Theory and Evidence from India
by
Timothy Besley
and
Robin Burgess
The Suntory Centre
Suntory and Toyota International Centres for
Economics and Related Disciplines
London School of Economics and Political Science
Houghton Street
London WC2A 2AE
DEDPS 28 Tel: (020) 7955 6674
December 2000
We have received useful comments from Oriana Bandiera, Steve Coate, Torsten Persson and a number
of seminar participants. We are grateful to STICERD for financial support. Grace Wong and Berta
Esteve-Volart provided some excellent research assistance.
Abstract
The determinants of government responsiveness to its citizens is a
key issue in political economy. Here we develop a model based on
the solution of political agency problems. Having a more
informed an politically active electorate strengthens incentives for
governments to be responsive. This suggests that there is a role
both for democratic institutions and the mass media in ensuring
that the preferences of citizens are reflected in policy. The ideas
behind the model are tested on panel data from India. We show
that public food distribution and calamity relief expenditure are
greater, controlling for shocks, where governments face greater
electoral accountability and where newspaper circulation is
highest.
JEL Classification: D72, H11, H41, I38, 012, P26.
Key Words: government responsiveness, mass media,
newspapers, political economy, political agency, political
accountability, democracy, social protection, disaster relief,
public food distribution
© The authors. All rights reserved. Short sections of text, not to
exceed two paragraphs, may be quoted without explicit
permission provided that full credit, including notice, is given to
the source.
1 Introduction
Understanding what makes government responsive to citizens’ needs is a key issue in
political economy. It is particularly poignant in low-income countries where, in the
absence of market opportunities, vulnerable populations rely on state action for their
survival. A key issue is then what institutions — economic, social and political — can
be built to enhance the effectiveness of the state in social protection. This paper lays
out a framework for thinking about the issues and explores its empirical implications
in an Indian context. Among other things, the approach highlights the importance
of information ßows about policy actions in increasing government responsiveness,
particularly the role of mass media in creating an incentive for governments to respond
to citizens’ needs.
There is a large literature that emphasizes why the poor and vulnerable may
not obtain the full attention of politicians even in a democracy where they have
numerical strength. These groups are typically poorly informed and are generally
less inclined to vote than richer and better educated citizens. A key question then is
what institutions and mechanisms enable vulnerable citizens to have their preferences
represented in policy. The approach laid out here does not rely on the vulnerable
being in a majority for them to obtain political power. What is key is that they have
enough electoral power to “swing” outcomes, otherwise politicians will not have an
incentive to be responsive to their demands. However, this is not enough. In order to
exercise their electoral po wer they must also perceive differences between candidates
or parties as regards responsiveness. Mass media can play a key role here by enabling
vulnerable citizens to monitor the actions of incumbents. This should give them some
sense of about which politicians will protect them in the future and can be used as
the basis of their voting decisions. Media development can therefore considerably
strengthen incentives for incumbents to build reputations for being responsive as an
informed electorate will have much greater power to punish unresponsive candidates
than an uninformed electorate. Mass media thus capitalizes upon shared vulnerability
among poor voters to ensure that they get receive greater policy attention.
It has long been recognized that the quality of government policy requires the de-
velopment of k ey institutions. There is long tradition among social thinkers including
Rousseau, Smith, Hobbes, Lock e, Madison, Jefferson and Mill who have argued that
press freedom is essential for citizens to make in telligent and well-grounded decisions
about public affairs. The idea that a key role of the press is to inform the electorate
is central to the political science literature on the role of mass media.
1
Emphasis
on the role of the media is also in tune with the idea that a strong civil society is
an important aspect of economic development. The recent resurgence of interest in
political economy, however, has paid little attention to the role of news media in
inßuencing policy.
2
There is also no Þrm empirical evidence in the literature of a link
1
See, for example, Brians and Watten berg [1996] and Mondak [1995].
2
Stromberg [2000] being a key exception.
2
between media development and public policy.
In this paper we develop a model where mass media has a role in providing
information that mitigates political agency problems. Having a more informed and
politically active electorate strengthens incentives for governments to be responsive to
the needs of vulnerable citizens. Using panel data for Indian states over the 1958-1992
period we then demonstrate that there is a robust link between media development
and government responsiveness.
India is an interesting and important context for analyzing the political economy of
government responsiveness. First, there are measurable shocks, such as droughts and
ßoods, which hit vulnerable populations and vary over space and time. We exploit
this variation in a panel of states between 1958 to 1992 to look at the factors that affect
how responsive state governments are to these shocks. Second, state level programs
allocate relief to populations that experience shocks. The relief measures we study
include public food distribution and calamity relief expenditure. The institutions
that administer relief are fairly similar across states — what differs is the political will
to exploit them. As India is a federal democracy this allows us to study how political
variables affect responsiveness. Third, India has a relatively free and independent
press among countries at a similar level of economic development. Nonetheless, there
are differences in newspaper circulation across the states. This variation will allo w
us to consider the role of news media in promoting government responsiveness.
We begin by showing that differences in government responsiveness are only
weakly related to economic development of Indian states. However, politics does
seem to matter. States that have higher historical turnout exhibit greater respon-
siveness to droughts and ßoo ds. We also Þnd that the level of political competition
and timing of elections exerts some inßuence on responsiveness. Finally we show
that newspaper circulation is strongly and positively correlated with government re-
sponsiveness. Moreover, this comes primarily from publications that are not in the
dominant languages — Hindi or English. This is consistent with the idea that regional
presses play a key role in forcing state governments to respond to shocks.
The remainder of the paper is organized as follows. The next section lays out
a theoretical structure through which to view the results. Section three describes
the institutional context for the empirical test. S ection four describes the data and
methodology we employ to test the main ideas behind the model and presents the
results. Section Þve concludes.
3
2TheModel
2.1 The Environment
There is a continuum of citizens of size one who care about two issues: ideology
and s ocial protection.
3
There are two time periods and no discounting. Citizens
have a preferred ideology which we label as left and right: k ∈ {L, R}.Acitizen
of type k receives a payoff φ
kl
from a policy maker of ideology l. Weassumethat
φ
LL
− φ
LR
= φ
RR
− φ
RL
= λ > 0, where λ is a measure of ideological polarization.
A fraction of the citizens is vulnerable to sho cks and care about social protection.
Let j ∈ {v, n} denote a citizen’s vulnerability where v stands for “vulnerable” and n
for “non-vulnerable”. In each period, a fraction β ∈
h
β, 1
i
of the vulnerable citizens
get hit by a shock to their consumption. We assume that each vulnerable citizen is
equally lik ely to experience a shock so that
1+β
2
is the probability that any citizen is
hit. The per-period utility function of a citizen of preference type (k, j )isy−fδ
jt
+φ
kl
with δ
jt
=1ifj experiences a shock and f>0 denotes the size of the shoc k.
The fraction of citizens of type (k, j )isdenotedγ
k
j
.Weletγ
k
= γ
k
n
+γ
k
v
denote the
fraction of the population with ideology k and γ
j
= γ
L
j
+ γ
R
j
the fraction with shock
vulnerability j. We assume throughout that there are enough vulnerable citizens so
that γ
k
v
>
¯
¯
¯
γ
L
− γ
R
¯
¯
¯
for k ∈ {L, R}. This also requires that there be real political
competition with the relative fractions of left- and right-wing citizens not being too
far apart.
The governmen t can act to protect the vulnerable citizens from the endowment
shock. For simplicity, we assume that there are only two policies: protection (χ =1)
and non-protection (χ = 0). If the government chooses protection, this completely
neutralizes the effect of the endowment shock for all of the vulnerable citizens. The
entire cost of protecting the citizens is assumed to be borne by the policy maker.
4
There are two possible (per capita) costs of protection: high (c
H
)andlow(c
L
)with
c
L
<c
H
. Theshockisrealizedatthesametimeasβ is revealed and is low with proba-
bility ρ. We assume that the cost realization is private information to the government
and that βc
H
> λ > βc
L
.
2.2 P olicy Determination
There are two parties, denoted A and B, that put up candidates for public office. Each
party is comprised of a subset of non-vulnerable citizens with identical ideological
preferences — party A comprises left-wingers and party B right-wingers. Candidates
are party members and cannot commit t o policies ahead of time. Each candidate is
3
Ideology should be thought of as a catch-all for all government activity other than social pro-
tection.
4
This is best thought of as being an effort rather than a Þnancial cost — the latter falling on all
citizens. If the cost affects taxes, then it would be possible for citizens to infer something about it.
The analysis would go through providing taxes wer e not fully revealing of the cost shock.
4
