India assumed corrupt officials were stalling workfare wages, so a dashboard tested that diagnosis before adding a single auditor

the problem

Wage payments under India's MGNREGA, the world's largest workfare program, were chronically delayed, and the default explanation was corruption — officials skimming or slow-walking payments for a bribe

background

what everyone would do

The standard response to chronically delayed payments in a large bureaucracy is to assume dishonesty and respond with more oversight — additional audits, stricter penalties, closer monitoring of the officials in the processing chain. That diagnosis treats every case of delay as evidence of corruption waiting to be caught, and the fix it implies (watch officials more closely, punish the ones found responsible) does nothing for delays that are actually caused by something else entirely.

what they saw

Dodge, Neggers, Pande and Troyer Moore saw that the corruption diagnosis had never actually been tested against the alternative explanation — that officials weren't hiding payments for personal gain, they simply had no visibility into where in a long processing chain a given payment had stalled, so nothing got chased up until it was already weeks overdue. Rather than build another audit or enforcement tool on the assumption of dishonesty, they built a dashboard that gave officials the same information a corruption investigation would eventually surface anyway — instantly and routinely, rather than only after the fact — and used random assignment to isolate whether visibility alone, with no added punishment, actually closed the gap.

the move

Rather than add auditors or penalties on the assumption that dishonest officials needed to be caught and punished, the researchers gave a randomized set of already-responsible bureaucrats the same visibility a corruption investigation would eventually surface — but instantly and routinely, before a payment had already gone stale for weeks — and measured whether that alone changed anything.

why it works

By randomizing access to the same real-time visibility a corruption audit would eventually produce, the study isolated information access as the actual lever, rather than assuming the fix required catching and punishing dishonest actors — and the fact that giving the dashboard to frontline managers alone produced the full effect, with no additional gain from also giving it to their supervisors, is itself the evidence that the bottleneck was overburdened officials lacking visibility, not officials who needed to be watched more closely to behave honestly. Because PayDash simply reorganized information the government's own system already collected, rather than requiring new infrastructure, new staff, or new punitive authority, the fix was cheap relative to hiring additional auditors or oversight staff, which is likely why it was viable to test at multi-state scale and why the Ministry of Rural Development pursued expanding it nationally rather than treating it as an isolated pilot. Giving supervisors the dashboard also reduced punitive officer transfers by 24%, suggesting that once real information about where delays actually originated was available, blunt punishment aimed at the wrong target became less necessary — supervisors could see the real cause of a delay instead of defaulting to reassigning whichever officer was nominally responsible.

the payoff

PayDash access sped up bureaucrat processing of workfare payments by 17%, and increased available worksites and participating household work-days by 23% and 10% respectively, with the largest gains concentrated in the agricultural lean season — exactly when the safety net matters most. The effect held whether the dashboard was given to frontline managers or their supervisors, with no additional gain from giving it to both, indicating the constraint really was information access rather than a need for more supervisory pressure; giving supervisors the dashboard also cut punitive officer transfers by 24%, a blunt performance-management tool the added visibility made less necessary. Per University of Michigan reporting on the study, the research team has partnered with India's Ministry of Rural Development to expand PayDash availability to users across the country, building on the multi-state field trial.

where it breaks

This mechanism depends on the underlying diagnosis actually being wrong in the specific way information-visibility fixes address — a bureaucracy where officials genuinely are engaged in deliberate rent-seeking would not respond the same way to a dashboard, since the study's own framing notes that increasing a manager's information in the presence of real corruption can backfire (better information can help a corrupt actor hide more effectively, not less). It also depends on officials being willing and able to act on the new visibility once it's provided — a dashboard revealing exactly where a payment is stuck accomplishes nothing if the officials shown that information have no actual capacity, authority, or incentive to intervene once they see it. And this case is a caution against assuming a diagnosis at all: the study's real methodological lesson is that the corruption explanation felt obviously true to observers before being tested, and randomizing access to test it directly, rather than defaulting to the standard enforcement playbook, is what revealed a completely different and cheaper fix — a lesson that applies well beyond this specific program to any institution reaching for punishment before checking whether the actual constraint is something else.

what came after

Published as Dodge, Neggers, Pande & Troyer Moore, 'From Delay to PayDay: Easing Bureaucrat Access to Implementation Information Strengthens Social Protection Delivery' (NBER Working Paper 33756, 2025), the study is cited in development-economics and public-administration circles as evidence that a program's own diagnosis of its failure mode (corruption vs. overload) should be tested before its intervention is chosen, since the two failure modes call for opposite fixes — one wants more scrutiny of officials, the other wants to reduce the burden on them.

references

1. [1]From Delay to PayDay: Easing Bureaucrat Access to Implementation Information Strengthens Social Protection DeliveryNational Bureau of Economic Research (Working Paper 33756), 2025nber.org
2. [2]Research: Overloaded Officials, Not Corruption, Delay BenefitsMirage News, via University of Michigan, 2026miragenews.com

keep it

same kind of clever

• 1991 · MarriottMarriott took the airlines' playbook and priced every room by the night it would die
• 2016 · NHS clinicsNHS clinics clawed back millions of missed appointments with reminder texts
• 1609 · Amsterdam Wisselbank (Bank of Amsterdam)Amsterdam stopped trying to police 800 coin types and just declared what a guilder meant instead