Syndicate content

PacDev 2014: So Many Evaluations Packed Into So Little Time

David Evans's picture
I returned last week from the 2014 Pacific Conference for Development Economics, with 80 presentations over the course of a very full day. These conferences are a great way to see what research is coming down the line across a whole array of topics (see below the word cloud generated from the presentation titles).
Here is a run-down of some of the evaluations presented, each in one tweet or less!  (The papers are grouped into a single PDF per session, so search or scroll within the link.)


Cash transfers!
  • Cash transfers to moms (vs household head) in Macedonia improves secondary school enrollment only for kids already expected to do well. (Armand)
  • Conditional cash transfers in India increased supply of health care but not quality! (Triyana)
  • Cash transfers in Mexico hurt ineligible households in the same neighborhood: Higher prices outweigh higher wages (Lehmann)
Education!
  • Providing accurate info about children’s academic abilities to parents in Malawi significantly affects investments (Dizon-Ross)
  • Teacher hardship allowance in the Gambia gets more teachers to hardship schools but doesn’t seem to improve education for the average student. (Pugatch & Schroeder)
  • Armed conflict in Punjab was bad for education (?!). Particularly bad for girls. (Singh & Shemyakina)
  • Want to improve educational outcomes and spending on children of HIV+ individuals in Malawi? Treat their parents! (Baranov & Kohler)
  • Giving textbooks to schools in Sierra Leone doesn’t help students learn if students never get them (Sabarwal et al)
  • DVD lectures used for college prep in Bangladesh significantly increased university admission (Kono et al)
  • Nationally representative panel data in India shows that absence hasn’t gotten better, but inspections really reduce them (Muralidharan et al)
  • The switch from English to Bahasa Malaysia as the language of instruction in Malaysia in 1970s improved schooling and wages (but not literacy) for some (del Carpio et al)
Health!
  • Smart phone monitoring improved health inspectors’ performance most acutely for inspectors with great personalities (Big 5 personality traits) (Callen et al)
  • High absolute incentives for reducing malnutrition in daycare centers in India worked, but mostly for boys and for kids whose religion is the same as the worker (Singh & Mitra)
  • Rapid diagnostic tests for malaria, even after patients had bought medicine, dramatically improved compliance. A text msg the next day boosted compliance even more! (Modrek et al)
Jobs!
  • Giving a transport subsidy to job seekers for up to 11 weeks in Ethiopia increases the likelihood of finding steady employment by more than a third! (Franklin)
  • Vocational training in Turkey had small effects in the first year, but those dissipated by the end of 3 years. (Hirshleifer et al) – co-authored by our own David McKenzie
Corruption, conflict, and violence!
  • Drug violence in Mexico is bad for labor market participation. (Calderón et al)
  • Falling maize prices in Mexico meant more drugs and more violence (Dube et al)
  • Poll observers in Ghana reduced fraud reduce over-voting at their stations but it gets displaced to unobserved stations (Asunka et al)
  • Legal aid in rural Liberia improved satisfaction among plaintiffs, reduced bribes, and improved food security. Get me a quasi-lawyer! (Sandefur & Siddiqi)
Consumption!
  • India’s quotas for subsidized cereals improve intake across food groups! (Kaul)
  • Rural roads in India increased adoption of agricultural technologies J but also led to school drop-outs to join the labor force L (Aggarwal)
  • Mobile money in Uganda has dramatically increased consumption through remittance income. (Munyegera & Matsumoto)
  • A one-time fertilizer subsidy in Mozambique leads to persistent impacts on fertilizer use as well as consumption (Carter et al)
Savings and credit!
  • For a commitment savings product in the Philippines, the clients who were more sophisticated about their hyperbolic discounting just avoided the product. (Hofmann)
  • Flexible microcredit design in Bangladesh did not result in higher default (as feared), and after the first year seems to increase consumption in the lean season. (Shonchoy & Kurosaki)
More on kids!
  • Child sponsorship programs improve not only schooling outcomes, but also educational & vocational aspirations! (Glewwe et al)
  • China’s economic liberalization post-Mao doubles rate sex selection toward boys (Almond et al)
Lots of other good stuff, some of which wasn’t posted on-line or wasn’t evaluations.  Here are two with cool measurement innovations:
  • Simons et al use temperature sensors to check for the Hawthorne effect in stove use (per my earlier post). (Big Hawthorne effect uncovered!)
  • Dizon-Ross et al use undercover enumerators to hunt for corruption in bednet distribution in Kenya, Uganda, and Ghana. (Little corruption found: yay!)
As always with conference presentations, not everything is circulate-able or cite-able. But it’s exciting to see how much good work is (hopefully) coming out in the coming months and years.

Note on the word cloud: Someone could (or has) written a thesis on word choices in titles. Across these titles, “evidence” wins, followed by “rural,” “India,” “impact,” “health,” “experiment,” and “experimental.” (I used textalyser.net for the analysis, and WordItOut.com to generate the word cloud.)
 

Comments

Submitted by Hans Jansen on

Hi David, thanks for your posting - since I am an ex-IFPRI researcher, this kind of stuff always is of interest to me.
I tried to access the Carter paper on fertilizer subsidy in Mozambique - however, link gives me paper about rural employment guarantee scheme in India! Would you please alert the IT guys and/or send me the Carter piece?
Many thanks and best wishes, Hans

For some reason the Berkeley website puts all the papers from one session into a single file - so scroll through the pdf that the link goes to and you should find the right paper.

Submitted by JohnGibson on

David,

Interesting word cloud. I didn't see the word "poverty" at all, so went to the main program to check. Apart from the keynote, where it is parenthetical, just one paper title out of 80 mentions "poverty". Probably it would be the same at other development conferences.

Perhaps the World Bank should come up with a new slogan, since "Working for a World Free of Poverty" seems passe going on the evidence from what the profession currently finds interesting.

Submitted by Dave Evans on

Nice point. I think that you're right that other conferences would be similar (e.g., NEUDC 2013 had just 2 papers with poverty in the title). But I wouldn't trust title analysis as a sufficient indicator of interest within the profession. Many of the papers are focused on increasing incomes and reducing poverty, whether through education, health, fertilizer use, cash transfers, or other methods. None of the PacDev papers have development in the title, but just about all of them are about development on some level.

That said, this may suggest that relatively few papers are focused on measuring directly the impact of interventions on reducing poverty.

Submitted by Berk on

Hey Dave,

Modrek et al. links to wrong paper...

For some reason the Berkeley website puts all the papers from one session into a single file - so scroll through the pdf that the link goes to and you should find the right paper.

Submitted by Chris Nelson on

Thanks Dave, always good to get an overview like this. One quick observation from me; the prevalance of papers on research in Africa and South Asia. It seems like the reast of the World is being left out and the Pacific entirely ignored in a Pacific conference (which I gather is more a reference to the Pacific coast of the USA). Only working from this list though and haven't explored the conference website.

Submitted by Dave Evans on

Thanks, Chris. You're right that Pacific means the West Coast (the analog to NEUDC's "Northeastern"). Your comment reminds me of Das et al's 2013 JDE paper "U.S. and Them: The Geography of Academic Research", which looks at how much economic research is published on different countries. They find, both globally and within regions, a stark positive correlation between GDP and research carried out (controlling for population). On average, there seems to be slightly less research on Africa than on East Asia & the Pacific together, but the East Asia average is pulled up by China, Indonesia, etc., so the Pacific may be not very well represented.

(Here's the ungated working paper of Das et al: http://www-wds.worldbank.org/servlet/WDSContentServer/WDSP/IB/2009/12/10/000158349_20091210112013/Rendered/PDF/WPS5152.pdf.)

Add new comment