Politics of international development · Field lab · Week 7

The Epicenter Bet

An NGO built community centres (“epicenters”) in Eastern Ghana's villages, chosen by lottery with no regard to party. Five years later, households that backed the governing party were worse off than similar households without the programme. How does a neutral programme end up with partisan losers?

Built on Baldwin, K., Karlan, D., Udry, C. & Appiah, E. “How political insiders lose out when international aid underperforms: Evidence from a participatory development experiment in Ghana.” World Development 169, 106296 (2023).

Follow the programme to well-being · Figure 1, redrawnchange per household vs control villages

Swipe sideways to see the whole chain. The table below has the same numbers.

NDC households (the president's party) All other households helps hurts

Who ends up better off?—

—

NDC

—

Others

—

Start here

Then try it yourself

How good is the epicenter?

Value per cedi, compared with the chiefs' self-help projects it replaced. At 100% it's just as good. The paper finds it was worse.

Cedis a year moved from self-help projects into the epicenter

NDC Others

The paper: NDC households gave about 27 cedis less to other local projects, others about 7 less.

The other stages stay at the paper's estimates: equal participation in the programme, more political engagement for NDC households only, and fewer district projects in NDC villages.

01What the experiment found

Effects in villages that took up the programme, with 95% confidence intervals. Each row has its own scale. The pattern: NDC supporters got more involved, gave up more, and ended up worse off.

Treatment effects by partisan group (TOT)Tables 2–5

NDC others + p<0.10 · * p<0.05 · ** p<0.01

02The argument

In three steps

1. A neutral start. A lottery picked 51 of 97 village groupings for The Hunger Project, and households of every party were equally exposed to it. Donors use NGOs and community programmes like this to avoid the partisan targeting in the Kenya Lab.

2. Insiders bet bigger. Supporters of the governing NDC got more politically engaged, ran for local office, and moved more of their own money and labour out of the chiefs' self-help projects and into the new epicenter.

3. The bet didn't pay. The epicenters delivered less than what they replaced, and engaging local officials brought little because budgets are controlled from the capital. The people who invested most lost most: about −0.43 SD in well-being.

For discussion

Find the break-even. Starting from “What happened”, how good would the epicenter need to be for NDC households to come out even?
Close the gap. Set both groups' giving to the same number. What's left of the partisan gap, and why?
Why insiders? Why might government supporters have expected more from this programme than opposition supporters did?
Equal chances, unequal results. Kenya's problem was biased inputs. Here the inputs were neutral and the outcomes weren't. Which should donors be responsible for?