Simple economic models of corruption treat bribes as monetary transfers from citizen to official: unpleasant, but not inherently more inefficient than regular taxes. Yet, the real-world data tells a different story – there’s a strong correlation between corruption and slow public services.

However, not all public services suffer the same. A new NUS study published in the Journal of Public Economics showed that some public services are slowed by corruption more than others. In particular, services where it is hard for bureaucrats to find out how urgently the citizen needs the service and how much they are willing to pay in bribes to get it faster.

The study’s author, NUS Economics Assistant Professor Martin Mattsson, introduced a new economic model adapted from monopolistic price discrimination theories, with a corrupt bureaucrat in place of a product-selling monopolist. With this model, he developed predictions for scenarios where officials have a high or low level of information about the maximum bribe an applicant is prepared to give to get the public service faster – then verified the model against real-world evidence.

Mattsson theorised that in the former, high information scenario, the bureaucrat can maximise his profits by charging each applicant their highest possible fee, in a take-it-or-leave-it deal, without affecting service delivery time.

In the latter, low information scenario, however, a corrupt bureaucrat is incentivised to offer tiered services in exchange for bribes. Even if it were easy to process everyone’s applications speedily, the bureaucrat can extract more bribes by intentionally slowing the process down for people who pay less, so that he can charge higher bribes from those willing to pay more. Hence, corruption produces more than just transfers from those in need of services to government officials; it can lead to intentional red tape and inefficiencies.

Mattsson tested his model’s predictions against data from the World Bank Enterprise Survey (ES), which covers a comprehensive range of government-firm interactions over eighteen years, across 158 countries. His predictions were broadly consistent with the data patterns, but one surprise result did emerge.

Does corruption grease or sand the wheels?

Mattsson had predicted that corruption would slow all services. But from the ES data, he instead found that there is, in fact, no relationship between corruption and slow service delivery in high information contexts.

If corrupt bureaucrats have instantaneous, complete information about how much every applicant is willing to pay in bribes, the model shows that they could theoretically be more economically efficient than non-corrupt bureaucrats who process all applications equally. “But in practice, we rarely see this,” cautioned Mattsson. On average, countries with corruption problems also tend to provide slower public services, he said.

How to speed up public services

The study has practical policy advice for how to improve corrupt bureaucracies: In high-information contexts, the speed of public service delivery is limited by the bureaucrats’ ability to provide services in a timely manner. Therefore, things like adding more personnel and technological solutions can improve service delivery speed. Anti-corruption efforts in these situations are unlikely to substantially improve the speed of service delivery.

However, in low-information contexts, where bureaucrats are intentionally slowing down service delivery, there is no point in hiring more personnel or installing tech solutions. Instead, working to root out corruption may have larger positive effects, as removing corruption would also remove the bureaucrats' incentive to slow down service delivery.

The model also makes a case for tiered public services under the rule of law. Mattsson suggested that honest bureaucrats can raise their efficiency by allowing people to pay an extra fee – that will go to the government – to speed up their application. “But it's surprising how rare it is for those services to exist,” he mused, comparing it to how common it is among private businesses. He speculated that this may be due to social norms concerning equality or fears of a slippery slope, where the government charges higher and higher fees from those who are willing to pay more.

Mattsson noted that talking to policymakers about corruption is always hard to do, given the topic’s sensitivity, and good data on corruption is hard to come by, so conversations often stray into speculation. He hopes that this study, which is grounded in real-world data, will help guide anti-corruption work as well as broader public service reforms.