Do Elections Shape India’s Rural Infrastructure Programmes?

Even programmes designed to be rules-based and insulated from political discretion can still bend subtly to electoral incentives. New research on India’s rural roads programme suggests that elections shape more than how much infrastructure gets built. They also influence when projects are approved and which ones move up the queue.

The literature on political business cycles, particularly the work of William Nordhaus, argues that politicians manipulate economic policies or public spending to improve their chances of re-election. Earlier studies largely examined macroeconomic indicators or fiscal variables. New research instead focuses on a specific form of public infrastructure: rural roads.

The study examines the Pradhan Mantri Gram Sadak Yojana (PMGSY), one of India’s largest rural infrastructure programmes. In theory, the scheme is designed to limit political discretion. It is centrally funded, rule-based and closely monitored.

In practice, however, politics still intrudes. State legislators cannot directly control the scheme, but they can influence decisions through lobbying, consultations and planning committees. That makes the PMGSY an ideal setting to test a familiar question in political economy: do elections shape public spending even when institutional rules appear tight?

Drawing on administrative data covering more than 150,000 rural roads across 18 states between 2000 and 2013, the study finds a striking pattern. Road approvals rise noticeably as elections approach. Although this surge does not appear to increase costs or reduce construction quality, it does affect the mix of projects that get prioritised.

The authors compile a detailed dataset linking individual roads to villages and legislative assembly constituencies using geographic information systems. They combine these data with records from the Election Commission of India and demographic information from the national census. Their analysis takes advantage of the fact that state legislative elections occur at regular five-year intervals, making election timing largely exogenous.

The results reveal a clear electoral cycle. The first stage of the construction process—sanctioning roads—rises sharply in the fourth year of a legislator’s term, roughly a year before voters return to the polls.

On average, constituencies see about two additional roads sanctioned during this period. Because construction takes time, the surge in approvals feeds into more contract awards and completed projects in the months leading up to elections. Construction activity therefore peaks when voters are most likely to notice it.

Information Gap
Interestingly, the study finds little evidence that these electoral cycles reduce efficiency. Construction costs, delays and quality ratings do not systematically worsen during election periods. The PMGSY’s monitoring systems, including centralised data tracking and independent quality inspections, appear to limit opportunities for misuse of funds.

Yet the absence of cost overruns does not mean electoral incentives are harmless. Rather than affecting efficiency, elections appear to influence the composition of projects.

Roads approved close to elections tend to have shorter construction timelines, making them easier to complete quickly. Politicians therefore prioritise projects that can deliver visible results before voters head to the ballot box. Areas with more difficult terrain, where construction takes longer, may receive fewer projects during election periods.

To explain these patterns, the researchers develop a model of electoral competition in which politicians attempt to signal competence by expanding visible public goods before elections. Even when voters anticipate such behaviour, competition between parties still creates incentives for pre-election spending.

A key element of the model is the distinction between informed and uninformed voters. Politicians are more likely to manipulate spending in areas where voters have less information about government performance.

Empirical evidence supports this prediction. Electoral cycles are stronger in constituencies with higher levels of illiteracy or weaker access to media, both of which serve as proxies for lower voter awareness. In these areas, road approvals rise more sharply before elections. By contrast, constituencies with greater access to newspapers and magazines experience smaller electoral swings in project approvals. This pattern suggests that information asymmetries permit politicians to manipulate infrastructure provision.

The researchers also test alternative explanations. One possibility is that legislators simply learn to manage the programme more effectively over time. Yet if learning were the main driver, road approvals would rise gradually across the five-year term rather than spike shortly before elections. The data show no such pattern. Nor do corruption or heightened political competition appear to explain the timing of approvals.

The broader lesson is clear. Even programmes with strong institutional safeguards are not entirely insulated from electoral incentives. Elections may not inflate costs or compromise construction quality, but they can still influence when projects are approved and which ones move forward.

In rural road building, this means timing and visibility can sometimes take precedence over longer-term development priorities. The research therefore underscores the importance of transparent information flows, strong media access and informed voters in limiting political distortions in public investment.

(*Batabyal is a Distinguished Professor, the Arthur J. Gosnell professor of economics, and the Head of the Sustainability Department, all at the Rochester Institute of Technology in Rochester, New York. These views are his own.)