In Brazil, anti-corruption audits substantially reduced vote buying and citizens’ demands for private favours – demonstrating how transparency initiatives can weaken clientelism and strengthen democratic accountability.
Across many democracies, corruption and clientelism often reinforce each other. Corrupt officials frequently steal public funds in order to buy political support through clientelist exchanges: handing out private benefits to voters in return for electoral support during elections (Bobonis et al. 2022). Clientelism erodes accountability, distorts policy priorities, and contributes to the under-provision of public goods (Keefer 2007, Hicken 2011). Yet while governments and international organisations have invested heavily in anti-corruption campaigns, much less is known about whether these efforts also weaken clientelism.
Auditing corruption in Brazil
Our research (Bobonis et al. 2025) examines whether fighting corruption can, in turn, reduce clientelism. We study Brazil’s prominent anti-corruption programme that randomly audited more than 2,200 municipalities between 2003 and 2015. Each audit scrutinised how local governments spent federal funds, with publicly released findings. Earlier research has shown that these audits curb corruption and improve accountability (Ferraz and Finan 2008, Avis et al. 2018).
We combine this random assignment of audits with a unique panel survey that we conducted among rural households in the semi-arid zone of Northeast Brazil. This zone’s 28 million residents are disproportionately poor and have long been subjected to clientelist politics. Local mayors and councillors often distribute cash, water, or medicine in exchange for political support, and citizens frequently request such help, especially during droughts or health crises (Nichter 2018).
Because audits were randomly allocated, they allow us to estimate causal effects on clientelism. Roughly 40% of the municipalities in our survey were audited before Brazil’s 2012 municipal elections, providing a natural experiment on how anti-corruption efforts affect participation in clientelist exchanges by both politicians and citizens.
Anti-corruption audits reduced vote buying and citizen requests
We find that anti-corruption audits substantially reduce key forms of clientelism. In randomly audited municipalities, the share of citizens reporting campaign handouts during the 2012 election campaign fell by 3 percentage points – a striking 51% decline relative to unaudited areas (Figure 1). These handouts, which often take the form of medicine, construction materials, and cash, are illegal under Brazilian law and frequently result in the ouster of mayors and city councillors (Nichter 2021).
Figure 1: Citizens receiving campaign handouts
Audits also changed citizens’ behaviour. Residents of audited municipalities were 3 percentage points less likely to request private goods from politicians, a 21% reduction. And the share of fulfilled requests (cases where politicians actually delivered those goods) fell even more sharply, by 4 percentage points, equivalent to a 44% drop.
These effects were not limited to the election period. Even during the following non-election year, citizens in audited municipalities remained less likely to ask for or receive private benefits, suggesting that anti-corruption efforts can have persistent effects.
How anti-corruption audits change incentives
Why do anti-corruption audits weaken clientelism? One possibility is that they might make citizens less willing to engage in such exchanges. Audits reveal information about wrongdoing, potentially undermining voters’ trust in local politicians and altering their willingness to trade votes for favours. Consistent with this interpretation, we find that in audited municipalities citizens were less willing to sell their votes in hypothetical scenarios, rated politicians as less honest and less competent, and perceived their local councillors as less reciprocal.
These perception shifts also appear to change the ‘price’ of clientelism. Among voters who did receive campaign handouts, the reported value of these handouts increased by nearly 40% in audited municipalities, implying that politicians had to offer larger transfers to secure the same support. In short, citizens demanded more in exchange for their votes, if they were willing to engage in clientelism at all.
Audits also reduced the frequency of contact between voters and politicians. Citizens in audited municipalities were 8 percentage points less likely to interact regularly with local officials, a behaviour which often helps to sustain clientelist networks.
Together, these findings show how audits reduce the supply of clientelist votes. When citizens lose trust in politicians’ honesty and reciprocity, clientelist relationships become harder to sustain.
Short-term versus lasting impacts
We also explore whether these effects persist over time. We compare municipalities audited in the previous mayoral term (2009–2012) with those audited in earlier years, and find the strongest declines in clientelism where audits occurred most recently. For example, recent audits reduced citizen requests for private goods by nearly 5 percentage points, while older audits had smaller, statistically insignificant effects.
These patterns suggest that anti-corruption campaigns can deliver substantial but possibly short-lived reductions in clientelism. The behavioural effects – diminished trust, fewer exchanges, and less interaction – may fade unless information about corruption remains salient or new audits continue to be publicised. This points to the importance of sustained transparency efforts rather than episodic interventions.
Implications for governance and welfare
Our findings highlight a largely overlooked channel through which anti-corruption initiatives can strengthen democratic accountability. By curbing both corruption and clientelism, Brazil’s audit programme helped to reduce voters’ use of personal favours and increased the distance between politicians and citizens.
This shift is normatively significant: in environments where clientelism has long undermined formal channels of social protection, breaking that link creates space for programmatic politics and better public service delivery. It also underscores that anti-corruption policies may generate positive externalities beyond their immediate goal of deterring graft.
While fighting corruption may well reduce citizens’ receipt of clientelist benefits, a back-of-the-envelope calculation emphasises the net benefits of anti-corruption audits. Drawing on estimates from Avis et al. (2018), audits reduced corruption by roughly BRL 355,000 per municipality per year, while our data suggests a reduction in clientelist transfers worth around BRL 85,000. Thus, the monetary value of reduced corruption is several times greater than the value of reduced vote-buying, implying that the bulk of corruption proceeds are not simply directed to voters through clientelism.
Nevertheless, the welfare implications remain nuanced. In the short term, some citizens may lose access to material assistance they previously relied on, especially in poor or drought-affected regions. However, in the longer run, shifting from clientelist exchanges to transparent, programmatic governance is likely to yield substantial welfare gains.
References
Avis, E, C Ferraz, and F Finan (2018), “Do government audits reduce corruption? Estimating the impacts of exposing corrupt politicians,” Journal of Political Economy 126(5): 1912–1964.
Bobonis, G J, P Gertler, M Gonzalez-Navarro, and S Nichter (2022), “Vulnerability and clientelism,” American Economic Review 112(11): 3627–3659.
Bobonis, G, P Gertler, M Gonzalez-Navarro, and S Nichter (2025), “Does combating corruption reduce clientelism?,” The Economic Journal, ueaf082.
Ferraz, C and F Finan (2008), “Exposing corrupt politicians: The effects of Brazil’s publicly released audits on electoral outcomes,” Quarterly Journal of Economics 123(2): 703–745.
Hicken, A (2011), “Clientelism,” Annual Review of Political Science 14(1): 289–310.
Keefer, P (2007), “Clientelism, credibility, and the policy choices of young democracies,” American Journal of Political Science 51(4): 804–821.
Nichter, S (2018), Votes for survival: Relational clientelism in Latin America, Cambridge University Press.
Nichter, S (2021), “Vote buying in Brazil: From impunity to prosecution,” Latin American Research Review 56(1): 3–19.