Principal investigator: Angelo Cruz do Nascimento Varella
Article title: EXPOSING CORRUPT POLITICIANS: THE EFFECTS OF BRAZIL'S PUBLICLY RELEASED AUDITS ON ELECTORAL OUTCOMES
Author of the article: Claudio Ferraz and Frederico Finan
Location of the intervention: Brazil
Sample size: 373 Brazilian municipalities
Main theme: Economic Policy and Governance
Type of Intervention: Reelection of mayors in
audited municipalities
Primary variable of interest: Information from random public audits in municipalities with mayors running for reelection.
Evaluation method: Experimental Evaluation (RCT)
Evaluation Context
Information and transparency are necessary for the proper functioning of a democracy. To that end, citizens must be informed about their leaders in order to hold them accountable for their actions and determine who should be chosen in electoral processes. However, in systems where citizens do not receive sufficient information about politicians in office, it is not possible to monitor or penalize improper acts.
In this sense, efforts to increase transparency and the availability of information useful to society, such as public audits, are fundamental elements for voters to identify corrupt politicians. Consequently, it becomes possible to punish corruption, forcing such politicians to act in the public interest, or even strengthening the candidacy of honest individuals.
With the aim of strengthening transparency and making relevant information available regarding the use of federal funds in Brazilian municipalities, the Federal Government initiated, in 2003, one of the main initiatives to combat political corruption in Brazil, through the Comptroller General of the Union (CGU). This program was used by researchers to evaluate the impact that the disclosure of such information had on the Brazilian municipal elections of 2004.
Intervention Details
The CGU's Public Lottery Audit Program consists of a lottery system that randomly selects Brazilian municipalities to have their accounts audited. Municipalities with fewer than 450 inhabitants can be selected, encompassing approximately 73% of the country's population, excluding capital cities and large urban centers. The program's aim is to ensure the appropriate use of federal funds allocated to Brazilian municipalities, guaranteeing public participation in monitoring public spending.
Once selected by lottery, the CGU (Brazilian Federal Comptroller General) collected all available information on federal funds received by the municipality between 2001 and 2003. Subsequently, a group of 10 to 15 auditors were sent to the location to examine the accounts and official documents, inspecting whether there was anything improper or inadequate in the municipality's budget execution, as well as the quality of the public goods and services offered. It should be noted that the auditors underwent a rigorous verification system and were well-trained, remunerated, and supervised.
After approximately 10 days of inspection, the auditors produced a formal report that was sent to the CGU headquarters in Brasília. Subsequently, the reports were forwarded to the Federal Court of Accounts (TCU), public prosecutors, and the respective municipal legislative branch. The reports were also summarized for publication on the website. online from the CGU and subsequently sent to major media outlets.
Methodology Details
By July 2005, the program had conducted 676 audits in 669 different municipalities. To determine the impact of publicizing the results of the audits on the elections, it was necessary to select only the municipalities where the mayors were eligible for reelection, totaling 373 municipalities, which constitute the database used in the research.
Based on the formation of this database, the researchers divided the municipalities included into two groups, according to the date the public audits were carried out:
- Before from the 2004 municipal elections – 205 municipalities;
- After from the 2004 municipal elections (control group) – 168 municipalities.
The purpose of this division is to investigate whether the occurrence of a public audit has the potential to influence voters' decisions. Since the database was randomly generated by lotteries, it can be used to measure whether municipal audits have impacted mayoral reelection and vote margins.
With regard to acts of political corruption, the authors differentiate levels of corruption according to the frequency with which corrupt practices or acts are identified. In other words, how many times the reports identified acts of fraudulent procedures, misappropriation of public funds, or overpricing during the term of the mayor eligible for reelection.
To complement the analysis, information on the municipal electoral processes in 2000 and 2004 was also used, as well as data on mayors from the Superior Electoral Court (TSE). Additional information on the municipalities was extracted from the 2000 Demographic Census and the research "Profile of Brazilian Municipalities: Public Management," both conducted by the Brazilian Institute of Geography and Statistics (IBGE).
One of the most important complementary variables for the analysis is the presence of local media, mainly represented by AM radio stations. The authors highlight that although newspapers and television media cover the results of public audits, the absorption of these media does not compare to the presence of regional radio stations, which have significant importance in disseminating political information in Brazilian municipalities, since approximately 79% of households in these localities own a radio.
Results
The results obtained by the researchers indicate that public audits and the disclosure of information about acts of political corruption before elections decrease the likelihood of reelection for mayors identified as corrupt. In municipalities with two violations identified by the audits, the probability of mayoral reelection was 17% lower when the audits occurred before the elections, compared to municipalities audited after the elections. Even in cases where mayors identified as corrupt managed to get reelected, the observed margin of votes was smaller.
Another finding by the researchers points to the fact that well-informed voters penalize politicians with higher levels of corruption. In the municipalities audited before the elections, the probability of the mayor's reelection was reduced by 7 percentage points for each act of corruption identified. In the control group, when the audits took place after the elections, there were no observable changes in the probabilities of reelection of corrupt mayors. Obviously, this information supports the argument that voters would like to penalize acts of corruption, but do not have the necessary information to implement such actions.
The study also confirms that the presence of local media favors the dissemination of useful information to society. Researchers identified that in municipalities where mayors had three corrupt acts identified by audits before the elections, AM radio coverage reduced the probability of reelection by 16 percentage points. In locations without regional radio, the observed reduction was only 3,7 percentage points.
In addition to penalizing corrupt politicians, the presence of local media also favors the reelection of honest politicians. Under the coverage of regional radio stations, politicians who have not had corrupt acts identified increased their chances of reelection by 17 percentage points. This occurs because credible information shapes voters' decisions, causing corrupt politicians to be penalized and honest politicians to be benefited, which can also impact the performance of the political class as a whole.
Lessons in Public Policy
Research demonstrates that the dissemination of credible information is fundamental to the full exercise of democracy. Public audits have the capacity to inform the population in a way that encourages good practices in the fight against corruption, provided that this data is conveyed in an intelligible and accessible manner. Once relevant information is absorbed by voters, corrupt politicians tend to be penalized and honest opponents tend to benefit.
Therefore, the presence of local media is relevant, as is transparency and the production of useful information. The presence of local radio stations demonstrates that communication channels with the public are important because they disseminate necessary information and adapt to voters' expectations, influencing election results and implementing accountability for acts of political corruption. It is not enough to produce the information; it must be properly absorbed by society.
Reference
FERRAZ, Claudio; FINAN, Frederico. Exposing corrupt politicians: the effects of Brazil's publicly released audits on electoral outcomes. The Quarterly journal of economics, vol. 123, no. 2, p. 703-745, 2008.
Normal 0 21 false false false EN-US X-NONE X-NONE
