Why political competition matters when inequality is high

In a high-inequality setting, local politicians with secure positions may favour the rich by diverting resources towards them, at the cost of the poor. To test this hypothesis, this article analyses data from rural India, and demonstrates that lower political competition worsens the impact of inequality on public provisioning as well as developmental outcomes such as infant mortality.

Political and civil rights are essential not only in themselves but also because they can ensure government accountability. Amartya Sen has long argued that democratic governance prevents catastrophes like famines – while emphasising that democratic rule is less excellent in preventing chronic diseases (such as endemic hunger) than acute food crises (see, for instance, Sen 1983, 1999).

Under what circumstances can democracy bring an end to chronic issues? We explore the role of political competition in addressing prolonged health crises. Even outside general emergencies, survival depends heavily on access to healthcare and other social services. We examine the impact of economic inequality and political competition on development and social policies. High inequality and low political competition can distort the provision of essential services to the poor (Banerjee 1997, Bardhan and Mookherjee 2000, Esteban and Ray 2006). In particular, when inequality is high, local politicians with secure positions – captured by their likelihood of winning elections – may neglect the poor and instead favour the rich.

In other words, as the rich get richer, incumbent politicians may have more to gain by diverting resources towards them. This implies that social provision to the wider population, including healthcare, diminishes. In a tight election, however, the incumbent seeking re-election cannot entirely neglect the needs of the poor. Therefore, a high level of political competition can fuel spending on public provision and mitigate the adverse impact of economic inequality.

Empirical study of rural India

In our research (Kjelsrud, Moene and Vandewalle 2024), we test this hypothesis using recent fine-grained data from rural India.¹ We measure the impact on infant mortality because it is affected by public provision at the local level. Over the last three decades, there has been a clear improvement in child survival worldwide, and India is no exception – however, infants still die from preventable diseases and injuries, highlighting the need for social provision. Further, there is a substantial variation across states. The worst-performing state’s (Madhya Pradesh) estimated infant mortality rate is 14 times higher than that of the best-performing state (Mizoram) (Duggal and Dilip 2022).

To test the hypothesis more directly, we also study two types of social provision for the poor: basic healthcare and employment. For healthcare, we focus on Primary Health Centres (PHCs), which provide basic health products and services and constitute the cornerstone of the health system in rural India. For employment, we focus on the Mahatma Gandhi National Rural Employment Guarantee Act (MNREGA), the world’s largest employment programme, which guarantees 100 days of minimum-wage employment per year for each rural household.

To provide causal evidence, we exploit the large re-districting of parliamentary constituencies in 2008.² We compare locations that were in the same constituency before the delimitation but fell in different constituencies after the delimitation. This acts as a ‘natural experiment’ where locations that were characterised by similar levels of political competition and income inequality before the delimitation were reallocated into new constituencies with now different levels of political competition and income inequality compared to before the delimitation. Overall, the re-districting shifted about one-fourth of the rural population into a new constituency. Given how the exercise was organised, we argue (and test) that the reallocation was as-good-as-random in rural areas.

Democracy as insurance against bad policies

Our results suggest a clear pattern. Economic inequality leads to more post-neonatal infant deaths, less functioning PHCs, and lower MNREGA employment – but only in situations with weak political competition. In constituencies with average levels of political competition (as measured by one minus the Herfindahl-Hirschman index³, inequality has little impact on most of our outcomes.

The pattern we establish is worrisome for everyone who fears that political competition is declining in India. We demonstrate that less political competition worsens the impact of inequality on health outcomes and social policies. For instance, if the level of political competition is one ‘standard deviation’⁴ below the sample mean, then a one standard deviation increase in inequality raises post-neonatal infant mortality by 13% of the sample mean. Similarly, with this low level of political competition, a one standard deviation increase in inequality reduces the healthcare index that we constructed (capturing key services and health personnel) by 0.15 standard deviations and reduces the wage amounts disbursed through MNREGA by 6% of the sample mean.

Our three main outcome variables (infant mortality, healthcare provision, and MNREGA employment) are extracted from three independent datasets. In our view, this strengthens the plausibility of our conclusions, making a solid case for the general importance of democratic accountability in the face of high inequality. Political competition matters for local provision of public goods and, more generally, for policies that improve the well-being of the poor.

Key lessons

The arguments we use and the empirical results we derive support Sen’s general mechanisms but in a somewhat different and less acute setting. We also highlight a different aspect of democracy – the role of political competition rather than a free press – but the two may be closely related since a free press may eliminate information biases that favour one side more than the other and thus level the playing field. In addition, we can specify conditions for democracy to protect the poor: without sufficient electoral pressure, politicians in areas with high inequality are likely to offer inadequate social provision, the consequences of which can be fatal for many. With sufficient electoral pressure, local politicians can be induced to increase social provisions not only in with the event of an abrupt crisis, but also under circumstances where the policy mainly helps those that are most in need of support.

Note:

We use the 2015-16 National Family and Health Survey (NFHS-4) as our data source on mortality. For public health centers we obtain the data from the District Level Household and Facility Survey for 2007-08 and 2012-13. Finally, we extract GP level data from the MNREGA Public Data Portal for the financial years 2011-12, 2012-13 and 2013-14
When the delimitation of Indian electoral constituencies was implemented, new constituency borders were drawn by the Delimitation Commission. This had an impact on the electoral importance of Indian administrative district, as some districts lost one or several constituencies and others gained some.
The Herfindahl Hirschman Index is a measure of market concentration – in this case, when there is no political competition, the constituency will have an HHI of 1. On the other extreme, when there is a high level of political competition, HHI approaches 0.
Standard deviation is a measure used to quantify the amount of variation or dispersion of a set of values from the mean (average) value of that set.