The Cost of State Officials' Manager's Demand for Bribes

How to cite

sierra, Raul sanchez de la; Titeca, Kristof; Nkuku, Albert Malukisa; Lameke, Aimable A.; Haoyang. 2022. "The Cost of State Officials' Manager's Demand for Bribes." AEA RCT Registry. AEARCTR-0009691. https://doi.org/10.1257/rct.9691-1.1

Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.

Raul sanchez de la sierra* , Kristof Titeca , Albert Malukisa Nkuku , Aimable A. Lameke , Haoyang

* Primary investigator

Registry ID
AEARCTR-0009691
DOI
https://doi.org/10.1257/rct.9691-1.1
First registered
Last updated
Status
Completed
Trial dates
2015-06-19 – 2015-07-30
Countries
Congo (Democratic Republic of the) (Africa)
JEL codes
D23, O1, D73
Registry record
http://www.socialscienceregistry.org/trials/9691

Abstract

We analyze the organization of corruption in a state agency. The dual mandate of the Democratic Republic of the Congo's traffic police agency is to manage traffic and to enforce the traffic code. We first document that, in the capital's branch, Kinshasa, traffic fines account for only 22% of the revenue generated by the branch. The remaining 78% of revenue comes from bribes paid by drivers. 63% of revenue from bribes is generated through a "quota scheme:'' managers at police stations ask agents posted at street intersections to escort drivers to the stations, where the managers extract a bribe from the drivers themselves. Experimentally decreasing the quota level, hence mitigating its effect, we find that the quota scheme worsens the agency's ability to fulfill its first mandate, while not improving its ability to fulfill the second. First, the quota causes 65% of all traffic jams and almost all accidents at the branch's intersections. Second, we find evidence suggesting that it fuels allegations against drivers that are more often false-extortion-than true, consistent with the scheme not creating incentives to comply with the code. The findings emphasize that the manager's demand for bribes is significant and creates profits and distortions beyond those that would be made possible via corruption by individual state officials.

Intervention

We provided a randomized encouragement to the police station managers to reduce the quotas of some intersections in some days in the battalion.

This generates random variation in the manager's demand for bribes through its main instrument, the quota.

In reducing the quota, we can compare public service under current system of corruption vs. with agent corruption alone. We use this experiment to analyze the social cost of manager organized corruption.

Experimental design

The design leverages that each manager commands multiple teams of agents, and deploys each of his teams permanently to a specific intersection. While agents have some freedom to move between intersections during the day, re-deployment of agents across intersections is rare. The normal channel through which managers communicate the quota to the corresponding teams is a routine phone call placed by the commander to the team leader each morning. In the experimental sample, the mean quota is 6.5 among the untreated team-days. While teams' quotas vary day by day, the typical quota level of each team is well-known (henceforth, natural quota).

Design details

Randomization method
Randomization done in office by a computer.
Randomization unit
Assignment of the quota reduction encouragement to days was random within each of the 18 teams, between June 20th and July 20th 2015 (excluding Sundays). That is, the assignment to the quota reduction treatment was random, with teams as randomization strata and team X day as randomization clusters. Since we randomized within team strata, this allowed the number of treated and control days to be the same for each team.
Clusters
450 team of agent/days
Observations
450 team of agent/days, 900 observations of agents/days
Arms
Planned: Quota reduction encouragement: 195 team of agents/days No quota reduction encouragement: 195 team of agents/days
Primary outcomes
Occurrence of Traffic Accidents Occurrence of Traffic Jams
Secondary outcomes
Number of drivers escorted to the police station, by whether they are high-bribe or low-bribe, bribe revenue and fine generated inside the police station and in the street.

Cited by

No citation data available for this registration yet.

Keywords