It’s time to take on politics

A remarkable new World Bank report makes the case for a radical change in the Bank’s approach to political systems. For years the Bank and other international agencies have sought to give the poor a voice in health, education, and infrastructure decisions through channels unrelated to politics. They have set up school committees, clinic committees, water and sanitation committees on which sit members of the local community. These members are then asked to “oversee” the work of teachers, health workers, and others. But a body of research suggests that this approach has produced disappointing results. More recently, researchers have tested ways to improve accountability of government services by strengthening, rather than ignoring, the political route to oversight. There are a number of quite promising results from a range of countries which have prompted this welcome proposal to change the emphasis of accountability programs at the World Bank.

There is little doubt that in many countries political systems have failed to deliver accountable government. In most poor and middle income countries many teachers and health workers fail to turn up to work, theft of essential drugs is common, as are kickbacks on construction projects. Informal fees are charged for services that by law are meant to be free. A few examples illustrate the point:

  • India: Only half of teachers were teaching during unannounced visits
  • Aceh, Indonesia: Truck drivers spent approximately US$40 on bribes per trip, 13% of the trip cost
  • Sierra Leone: Over 50% of parents reported paying for immunizations which are meant to be free. (Note this was before a big push to make all services for under 5s free)

If politics is not delivering accountable government, a natural response is to attempt to circumvent it: give power directly to consumers who care most about service quality. The result has been a plethora of committees at the very local level. But rarely are these given any real money and power. Teachers and health workers (by far the biggest expenditure in health and education) continue for the most part to be paid by governments.  When researchers have investigated how these committees function on the ground they often look very different from what they appear like on paper. In India, most Village Education Committees that were mandated by the government were not operating: 25% of people whose names were listed as being on the committee when surveyed did not know they had this role. Even the efforts of the otherwise highly effective NGO Pratham failed to achieve effective village engagement through the committee or improve educational outcomes.

There have been some successes, particularly in health. In Uganda, researchers found that most, legally prescribed, local health committees were not active. However, support from an outside NGO and information from a detailed survey of health care quality led to higher attendance of nurses and improvements in health. But this involved very expensive intensive data collection efforts and when the researchers tested a cheaper version without village-specific data there was no positive effect. In Sierra Leone, NGO support to community health committees (with an intensity between that of the two Uganda approaches) was successful in improving health outcomes, cutting health worker absenteeism (paper in progress).

Another type of project that promotes community-level participation, not through the political process but through community action, is Community Driven Development. Grants are given to communities on the condition that there is strong community participation in the decision-making about how the grants are spent. Again the results of rigorous evaluations are rather disappointing: the grants themselves have produced real outputs on the ground, but there is little evidence community participation persists once the grant is finished from RCTs in Sierra Leone, DRC, and Liberia (the last shows some gains though arguably small and confined to specific groups).  

Partly in response to these underwhelming results, attention switched back to enhancing engagement and transparency of the political process. Results from a series of RCTs have been more encouraging. Voters in Brazil respond to the release of audit reports on corruption in municipalities (more corruption, the lower the vote share). Voters in Delhi respond with considerable sophistication to information about how members of parliament spend funds made available for their constituency (increased vote share for those who spend more in the voters’ area or attend committees for issues of importance to voters, and lower vote shares for those who spend money on goods that are not seen as useful such as fountains).

There are some negative effects: publicizing corruption levels when these levels are very high led to lower voter turnout. In Kenya, providing messages about the integrity and efficiency of the electoral commission backfired when there were significant problems with voting systems.

In my view the most promising results are from RCTs of programs that seek to engage voters rather than just provide them with information. In Benin and the Philippines, when political parties held town hall meetings, rather than the normal rallies with little policy content, challengers were able to combat clientelism of the dominant party and increase support. In Sierra Leone, filming and then screening debates between candidates for MP led to increased knowledge about political candidates, their policy positions, and the political process, better alignment on policies between voters and the candidate they voted for, and a shift in votes towards the candidate considered most effective in the debate. MPs elected from constituencies where debates were screened visited their constituency more often, held more meetings with constituents, and were considered to be doing a better job by neutral parties. It was also possible to track debate MPs’ spending to real projects on the ground.  

The accumulated evidence in the last 15 years has changed my view on the best way to improve accountability and it is great to see this shift in this World Bank document, written by Stuti Khemani, with whom I worked to evaluate village-level accountability in India, as well as J-PAL affiliates Claudio Ferraz and Fred Finan. There is likely to be some push back to what some will portray as “the Bank getting involved in elections.”  But as the report sets out, there are many ways the Bank and others can support citizen engagement in ways that do not compromise the Bank’s neutrality.