Lang : English
  • English
  • Français
  • हिन्दी
  • Português
  • Español
close× Call Us

Brief 67: The Oversight of Corruption and Inefficiency in Local Public Works in Peru

In the construction of public works, does notifying governments about civil society oversight reduce costs due to corruption, or does it introduce inefficiencies by demoralizing and distracting officials? This study approaches this question in the context of a national program in Peru wherein 200 urban and peri-urban district governments were given funds for public works programs. Half of the districts were randomized into a treatment group and received letters indicating that specific public works under their charge were being monitored by a civil-society organization with the support of the country’s leading anticorruption agency. Over a two-year period, the intervention lowered the cost of public works in the treatment group without reducing rates of completion.

Read full study

Category: Corruption

Tags: Oversight, Public Works, Civil Society

Date of Publication: Tuesday, December 10, 2019

EGAP Researcher: Paul Lagunes

Geographical Region: South America

Research Question:

In the construction of public works, does notifying governments about civil society oversight reduce corruption, or does it introduce inefficiencies by demoralizing and distracting officials?

Preparer: Tanu Kumar


In Peru, decentralization is thought to have spread corruption at the district level, with the majority of national corruption claims about public works involving district or municipal governments.  In 2012, the Office of the Comptroller General increased its capacity to monitor corruption by launching an online platform called INFObras that collects and publishes information about the costs and completion of public infrastructure projects across the country. This platform is also used by a civil society organization called Proética to conduct corruption investigations that it makes available to the central government. This study entailed a collaboration between the Office of the Comptroller and Proética in using INFObras to monitor the use of funds allocated through a public works program run by the Housing Ministry known as the Programa de Mejoramiento Integral de Barrios in 2015. In this program, 200 urban or peri-urban districts received funding for roads, sidewalks, exercise courts, public markets, and other projects.

Research Design:
These 200 districts were randomly assigned to either control or treatment, with half in each category. The districts in the treatment group received a total of four letters over the course of fourteen months. In October 2015, Proética sent the first letters warning that it would be using INFObras to monitor the execution of local public works. Soon after, the Office of the Comptroller General mailed letters of its own to let the mayors know that the anticorruption agency was actively coordinating with Proética on the monitoring of public works. In October 2016, Proética sent another letter to remind the mayors of the ongoing monitoring. In December 2016, the Office of the Comptroller General concluded the intervention with a final letter noting that the agency would continue to coordinate with the CSO on the monitoring of public works. The treatment’s impact was measured by looking at the amount of time and funds required to complete the public works.
Roughly two years after the first letter was issued the treatment reduced the average cost of public works by 455,370 Peruvian soles (or approx. 140,300 USD) when estimated without controls, or about 297,900 soles (approx. 91,000 USD) when estimated with controls (Figure 1). These estimates reflect a roughly 15% decrease in costs, and both are statistically significant at the five percent level.
Figure 1
Hypothetically speaking, it was possible that the treatment could displace corruption rather than contain it, as monitoring over one public work can lead officials to shift the corruption to an unmonitored public work. The author therefore examined districts in the treatment group that had more than one public work being built around the same time and found a total of 52 untreated public works. If corruption was displaced in these districts, then the untreated public works might be expected to cost more than their treated counterparts. Figure 2 shows that both sets of projects had similar average costs. Finally, 87.23% of public works in the control group and 85.41% in the treatment group had reached completion. These similar rates of completion suggest that the monitoring generated reductions in cost without creating inefficiencies or delays.
Figure 2

Policy Implications:
While oversight mechanisms are important tools for controlling corruption, some have warned that monitoring may paralyze the work of a bureaucracy. The results of this study, however, show that districts made aware of monitoring completed public works at a similar rate and at a lower cost than those not explicitly notified about the monitoring. Alerting governments of exiting monitoring programs seems to be a cost-effective tool for controlling corruption given that field experiment cost no more than $130,000 and appears to have saved the Peruvian treasury nearly $9 million. The results suggest that civil society oversight backed by government authorities can result in the timely and cost-effective execution of public works.