Extractive institutions
Extractive institutions is a barrier to human progress within the societal factors. The evidence for it, and the methods proposed to address it, follow.
Acemoglu and Robinson (2012) argue that nations prosper when institutions are inclusive and decline when they are extractive. Corruption, weak rule of law, and instability belong here.
Evidence
- Corruption. In the 2025 Corruption Perceptions Index, the global average fell to 42 out of 100. It was the first drop in more than a decade. Of 182 countries, 122 (67%) scored below 50. Only 31 have improved their scores since 2012 (Transparency International).
- Rule of law. Rule of law declined in 68% of 143 countries in 2025, up from 57% the previous year. It was the eighth straight year in which more countries declined than improved. Freedom of opinion and expression fell in 73% (World Justice Project).
- Peru's mita. Districts just inside the boundary of a colonial forced mining labor system (1573 to 1812) have about 25% lower household consumption today than districts just outside it. They also have about 6 percentage points more child stunting. The boundary was drawn centuries ago. Because of this, the comparison isolates an extractive institution's effect better than cross-country studies (Dell, 2010).
What can be done
Actions rated on the strength-rating scale (A strong and replicated, B solid but limited, C weak or debated, D contested or failed), applied to the specific claim made.
- Top-down audits of public spending (B). A randomized experiment covered over 600 Indonesian village road projects. In that experiment, raising the chance of a government audit from 4% to 100% cut missing expenditures by eight percentage points, measured against independent engineers' cost estimates (Olken, 2007). Missing expenditures fell from 27.7% to 19.2% of spending, and each $500 audit improved roads by an average of $1,165 (J-PAL, n.d.). Village elites shifted toward hiring family members, which audits detect less easily (J-PAL, n.d.). See also the random audits in Brazil in the proposed methods below.
- Publishing what money should reach local services (B for less theft, not shown for learning). In the mid-1990s Ugandan primary schools received only one-fifth of the government grants meant for them (Kuecken and Valfort, 2019). A newspaper campaign told schools and parents how much each school should get. The campaign sharply reduced capture of the grants and raised enrollment (Reinikka and Svensson, 2005). An independent replication confirmed the fall in corruption. The replication found the enrollment gain only with an improved enrollment measure. The replication could not support the claim that learning improved (Kuecken and Valfort, 2019). The design relies on distance to newspaper outlets, not random assignment.
- Higher advertised pay to recruit better civil servants (B). In a Mexican public sector recruitment drive, salaries were randomly set across sites. Higher wages attracted more able applicants, measured by IQ, personality, and interest in public sector work. There was no sign of weaker public service motivation. Higher wages also raised acceptance rates, including in less attractive municipalities (Dal Bó et al., 2013). This is one country's program for one type of post.
- Not shown to work: community meetings as the main check on corruption (B). In the same Indonesian experiment, more grassroots participation in monitoring had little average impact on missing expenditures, and helped only where free-rider problems and elite capture were limited (Olken, 2007).
- Not shown to work: voter information campaigns about politicians' performance (A for no average effect). Seven coordinated randomized trials ran in six countries, with a meta-analysis registered before the data were seen. The trials found no evidence overall that typical nonpartisan information campaigns changed voter behavior (Dunning et al., 2019).
Proposed and experimental methods
Methods that are proposed, under trial, approved in some places, or tried and then failed. Each shows a stage label and an evidence rating. A stage label shows how far a method has progressed, not whether it works. The stage labels are explained on the societal factor page.
- Digital payments and e-invoicing to cut leakage from public programs (Large trial, B). Paying benefits electronically, and releasing funds only against electronic records, is proposed to reduce theft by officials. A randomized rollout of biometric smartcard payments covered 157 subdistricts and 19 million people in Andhra Pradesh, India. It made employment-scheme payments faster, more predictable, and less corrupt without reducing access. It also reduced leakage in both the employment and pension programs (Muralidharan et al., 2016). A field experiment in Bihar replaced advance payments with "just-in-time" payments triggered by e-invoices. This cut program spending by 24% with no fall in employment. It also reduced fake households in the database and lowered officials' personal wealth by 10%. The national scale-up gave a lasting 19% fall in spending, although payments became slower (Banerjee et al., 2020). A later study in Jharkhand used randomized and natural experiments. It found that reforms adding stricter biometric identity requirements for food rations reduced corruption. The study also found that the reforms led about 1.5 to 2 million legitimate beneficiaries to lose access at some point, mainly because of how the transition was managed. This shows that stricter verification can exclude eligible people (Muralidharan et al., 2025).
- Random government audits made public (Approved but not scaled, B). Brazil's anticorruption program randomly chose municipalities for audits of federal funds. A study found that being audited in the past reduced future corruption by 8% and raised the chance of later legal action by 20%. This worked mainly by raising the perceived non-electoral costs of corruption (Avis et al., 2018). This builds on "Antisocial and self-serving traits" in the psychological factors, which covers machine learning to target these audits. Evidence comes from one country's program.
- Electronic public procurement (In wide use, B). E-procurement moves tendering online. This makes information easier to get and cuts face-to-face contact with officials. A study used differences in adoption across regions of India and Indonesia. It found no evidence of lower prices but found better quality, with better road quality in India and fewer delays in Indonesia. It also found more contracts won by firms from outside the region (Lewis-Faupel et al., 2016). The design is not randomized.
- Reversed: International anti-impunity commission in Guatemala, CICIG (Failed or reversed, C). The UN-backed International Commission against Impunity in Guatemala began work in 2007 to investigate criminal networks inside the state. The International Crisis Group made a synthetic-control estimate of the commission's contribution. The estimate put it at 2.77 fewer homicides per 100,000 people per year, or more than 4,500 homicides avoided from 2007 to 2017 (International Crisis Group, 2018). The commission opened an investigation into President Jimmy Morales, his brother, and his son. After that, Morales declined to renew its mandate. The commission closed on September 3, 2019, and its remaining cases passed to a special prosecutor's office (Abbott, 2019). The homicide estimate is a model from an advocacy organization, not peer reviewed.
- Reversed: Charter cities in Honduras, Próspera (Failed or reversed, B). Charter cities are new zones with their own laws and governance. They are proposed as a way to escape weak national institutions. Honduras' Congress repealed its law on Zones for Employment and Economic Development (ZEDEs) in 2022. On September 20, 2024, the Supreme Court declared the ZEDE framework unconstitutional and applied the ruling back to 2013 (Escalante, 2024). The Próspera zone kept operating. Its investors filed a claim against Honduras under the CAFTA-DR trade agreement in December 2022. The investors stated damages as high as $10.775 billion if their entire investment were lost (Martinez, 2026, Próspera, 2022). In October 2025 Honduras' Attorney General's office said the claim had been reduced to $1.63 billion (Estrategia & Negocios, 2025). The case (ICSID Case No. ARB/23/2) was still pending in September 2026 (ICSID, n.d.). No evaluation of the zone's effects on residents was found.
Sources cited on this page
- Dell, M. (2010). The persistent effects of Peru's mining mita. Econometrica, 78(6), 1863-1903. DOI B Moderate
- Acemoglu, D., & Robinson, J. A. (2012). Why Nations Fail. New York: Crown. C Limited
- Acemoglu, D., Naidu, S., Restrepo, P., & Robinson, J. A. (2019). Democracy does cause growth. Journal of Political Economy, 127(1), 47-100. DOI B Moderate
- World Justice Project (2019). Measuring the justice gap: a people-centered assessment of unmet justice needs around the world C Limited
- Abbott, J. (2019, September 3). Guatemala's CICIG: UN-backed anti-corruption body shuts its doors. Al Jazeera. link B Moderate
- Avis, E., Ferraz, C., & Finan, F. (2018). Do government audits reduce corruption? Estimating the impacts of exposing corrupt politicians. Journal of Political Economy, 126(5), 1912-1964. link B Moderate
- Banerjee, A., Duflo, E., Imbert, C., Mathew, S., & Pande, R. (2020). E-governance, accountability, and leakage in public programs: Experimental evidence from a financial management reform in India. American Economic Journal: Applied Economics, 12(4), 39-72. link B Moderate
- Escalante, L. (2024, September 28). Honduran government praises repeal of the ZEDE law as investors denounce "lies" and "abuse." Contra Corriente. link B Moderate
- Lewis-Faupel, S., Neggers, Y., Olken, B. A., & Pande, R. (2016). Can electronic procurement improve infrastructure provision? Evidence from public works in India and Indonesia. American Economic Journal: Economic Policy, 8(3), 258-283. link B Moderate
- Martinez, J. (2026, September 19). Honduran critics call Próspera's new officials a "modern coup." The Rio Times. link B Moderate
- Muralidharan, K., Niehaus, P., & Sukhtankar, S. (2016). Building state capacity: Evidence from biometric smartcards in India. American Economic Review, 106(10), 2895-2929. link B Moderate
- Muralidharan, K., Niehaus, P., & Sukhtankar, S. (2025). Identity verification standards in welfare programs: Experimental evidence from India. Review of Economics and Statistics, 107(2), 372-392. link B Moderate
- Dal Bó, E., Finan, F., & Rossi, M. A. (2013). Strengthening state capabilities: The role of financial incentives in the call to public service. Quarterly Journal of Economics, 128(3), 1169-1218. link B Moderate
- Dunning, T., Grossman, G., Humphreys, M., Hyde, S. D., McIntosh, C., Nellis, G., et al. (2019). Voter information campaigns and political accountability: Cumulative findings from a preregistered meta-analysis of coordinated trials. Science Advances, 5(7), eaaw2612. link A Strong
- Kuecken, M., & Valfort, M.-A. (2019). Information reduces corruption and improves enrolment (but not schooling): A replication study of a newspaper campaign in Uganda. Journal of Development Studies, 55(5), 1007-1029. link B Moderate
- Olken, B. A. (2007). Monitoring corruption: Evidence from a field experiment in Indonesia. Journal of Political Economy, 115(2), 200-249. link B Moderate
- Reinikka, R., & Svensson, J. (2005). Fighting corruption to improve schooling: Evidence from a newspaper campaign in Uganda. Journal of the European Economic Association, 3(2-3), 259-267. link C Limited
Every source for this factor is listed on the societal factor page.