Governance & participation
Participation, budgets and civic engagement. Real interventions, each with its sources: what worked, what didn't, and why.
14 worked7 mixed3 didn't work
What worked
- WorkedGovernance & participation
Brazil's random audits of city governments
ChallengeCorruption in local governments' use of federal funds was widespread and rarely exposed.
Picking municipalities by lottery for federal audits, and publishing the results.
−7 ppre-election chance, mayors with two corruption violations exposed pre-electionQuasi-experimentalRead case - WorkedGovernance & participation
Publishing school grant amounts in Uganda
ChallengeMost school grant money sent by the government never reached schools.
Printing monthly school grant transfers in newspapers so communities could check them.
24% → 80%share of school grants reaching schools, 1995 to 2001Quasi-experimentalRead case - WorkedGovernance & participation
Ireland's Citizens' Assembly
ChallengePoliticians had avoided divisive constitutional questions like abortion for decades.
A randomly selected group of citizens deliberating on divisive constitutional questions.
66.4%voted yes in the referendum that followedObservationalRead case - WorkedGovernance & participation
Guadalajara's Visor Urbano licensing platform
ChallengeGetting business and building permits invited officials to demand bribes.
Moving business and building permits online to take away chances for bribes.
−74%bribe requests from municipal agentsQuasi-experimentalRead case - WorkedGovernance & participation
Biometric Smartcard payments for welfare in Andhra Pradesh, India
ChallengeOfficials skimmed welfare money through fake work claims and slow, opaque cash payments.
Paid rural job-guarantee wages and pensions through biometric Smartcards and village bank agents.
−41%leakage of job-scheme wage fundsRCTRead case - WorkedGovernance & participation
Small-donor matching funds in New York City elections
ChallengeCity campaigns depended on large donors from a few wealthy neighbourhoods, inviting undue influence.
Multiplied small donations from city residents with public money for candidates who accept limits.
89% vs 30%neighbourhoods with small donors, city vs state racesObservationalRead case - WorkedGovernance & participation
Cash rewards for whistleblowers at the US Securities and Exchange Commission
ChallengeSecurities fraud and corporate bribery are hidden inside firms and hard for regulators to detect.
Pays tipsters 10–30% of sanctions over $1 million collected in cases their information helped win.
$2.2bnpaid to 444 whistleblowers, 2011–2024ObservationalRead case - WorkedGovernance & participation
Connecticut’s Citizens’ Election Program after the Rowland scandal
ChallengeA governor resigned over gifts from state contractors, exposing how donors bought access.
Offered public grants to candidates who raise small $5–$100 gifts and refuse lobbyist and contractor money.
97%of legislative campaign gifts from individuals (49% before)ObservationalRead case - WorkedGovernance & participation
Georgia’s overhaul of its traffic and patrol police
ChallengeGeorgia’s police were seen as the country’s most corrupt institution, extorting bribes daily.
Fired the whole traffic police in 2004 and rebuilt a smaller, better-paid patrol force.
69%viewed police favourably in 2009ObservationalRead case - WorkedGovernance & participation
Positive police door-knock visits: New Haven, Connecticut
ChallengeLow trust in police, especially among Black residents, undermines cooperation and public safety.
Uniformed officers made brief, friendly, non-enforcement visits to randomly chosen homes.
+11 ptsattitude gain among Black residents visited (0–100 scale)RCTRead case - WorkedGovernance & participation
Procedural justice training at crime hot spots: Tucson, Cambridge, Houston
ChallengeAggressive hot-spot policing can cut crime but erode residents' trust in police.
Officers patrolling crime hot spots got intensive training in fair, respectful, transparent policing.
−14%crime incidents in hot spots with trained officersRCTRead case - WorkedGovernance & participation
Showing residents their fixed potholes: Boston 311 app
ChallengeResidents rarely see the work government does, which feeds low trust and disengagement.
Boston's 311 app sent residents photos of city crews fixing the problems they reported.
+60%more service requests from users sent photos of fixesQuasi-experimentalRead case - WorkedGovernance & participation
Participatory budgeting: New York City Council
ChallengeLow turnout and little say for residents, especially poorer and younger ones, in city spending.
City Council members let residents propose and vote on how to spend part of their capital budgets.
+8.4 ptsprobability of voting in regular elections among PB participantsQuasi-experimentalRead case - WorkedGovernance & participation
IDNYC free municipal ID card: New York City
ChallengeMany undocumented, homeless and low-income residents lacked ID to use city services or deal with police.
New York City offered a free photo ID to all residents regardless of immigration status.
77%of immigrant cardholders surveyed felt a greater sense of belongingObservationalRead case
Mixed results
- MixedGovernance & participation
Participatory budgeting in Porto Alegre
ChallengeCity investment followed patronage, and poor neighborhoods lacked basic services.
Residents vote on how part of the city's budget is spent.
75% → 98%households with water and sewer connections, 1988–1997ObservationalRead case - MixedGovernance & participation
Audits versus community monitoring in Indonesia
ChallengeFunds for village road projects were being lost to corruption.
Comparing top-down audits with grassroots oversight of village road projects.
−8 ppmissing funds with audits; little change from monitoringRCTRead case - MixedGovernance & participation
Democracy Vouchers for city elections in Seattle
ChallengeA small, unrepresentative group of wealthy donors funded city campaigns and shaped who could run.
Gives every voter four $25 vouchers to donate to city candidates who accept spending limits.
48,071residents used vouchers in 2021 (20,727 in 2017)Quasi-experimentalRead case - MixedGovernance & participation
ProZorro open e-procurement system in Ukraine
ChallengePublic contracts worth about 15% of GDP were often awarded without real competition.
Moved all public purchasing online with open data and auctions as the default method.
24.5% → 70.3%procurement value awarded competitively, 2015–2017ObservationalRead case - MixedGovernance & participation
Shakman court decrees against patronage hiring in Chicago
ChallengeCity jobs were handed out as rewards for political work, fuelling machine politics and fraud.
Federal consent decrees banned political hiring and firing in Chicago city government.
1972 -> 2014from first consent decree to court finding of substantial complianceObservationalRead case - MixedGovernance & participation
Procedural justice training for police: Chicago
ChallengeDeep distrust of the Chicago Police Department and high levels of complaints and force.
Chicago police put 8,480 officers through a one-day course on fair and respectful policing.
8,480officers trained; corrected analysis finds no clear drop in complaintsQuasi-experimentalRead case - MixedGovernance & participation
Buffalo's Clean Sweep and resident reporting of crime and blight
ChallengeResidents in many US cities rarely report crimes or service problems to 911 and 311.
Weekly city 'sweeps' of high-need blocks: door-to-door outreach, rapid clean-up and community policing.
+32%likelihood of reporting a drug crime within 3 months, treated vs matched householdsQuasi-experimentalRead case
What didn't work
- Didn't workGovernance & participation
Iceland's crowdsourced constitution
ChallengeAfter the 2008 crash, Icelanders had lost trust in their political system.
A new constitution drafted with public input online, that parliament never adopted.
64.2%of votes cast backed the draft, 2012 advisory referendumObservationalRead case - Didn't workGovernance & participation
Mexico’s National Anticorruption System
ChallengeHigh-level scandals went unpunished as agencies worked in isolation and lacked independence.
Created a body to coordinate auditors, prosecutors, courts and a citizen committee against graft.
35 → 26Corruption Perceptions Index score, 2015–2024ObservationalRead case - Didn't workGovernance & participation
Police body-worn cameras: Washington, DC
ChallengeCities hoped cameras would curb police use of force and complaints and rebuild public trust.
DC randomly gave body-worn cameras to about half of its eligible police officers.
2,224officers randomized; no significant effect on force or complaintsRCTRead case