Some of the more impressive results I have seen have come out of papers which take a long term view. Three examples of this are Esther Duflo’s work on school construction in Indonesia (ungated version here), Hoddinott, et. al’s paper on an early childhood nutrition intervention in Guatemala, and Baird, et. al’s paper on the long term effects of deworming in Kenya.
Later in the post, he also lays out the obstacles to putting together and funding such studies, especially the challenge of tracking respondents over the life of such a project. Here at IPA Ghana, one of our studies is looking to evaluate the returns to secondary schooling in Ghana.
…2,068 students, 682 have been selected through random assignment to each receive a 4-year scholarship to attend the senior secondary school they were admitted to. Between 2009 and 2018, we will keep track of all 2,068 students and conduct follow-up surveys with them, and the households that they reside in, every three years. The follow-up surveys will include questions on health, labor market outcomes, and fertility and marriage. This will enable us to measure the medium and long-term effects of acquiring secondary education on multiple aspects of life and well-being.
Currently in the midst of this project, its really interesting for us to plan different strategies to keep in touch with respondents in this study. Tracking youth is particularly tricky – especially those respondents in the sample who are out of school – since they often migrate far out in search of work. So far, the team has managed to keep in touch with most of these students – however, not without any challenges. Our project team has shared with us that some of these respondents might have even migrated outside the country looking for work. It appears that the recent disturbances in neighbouring Ivory Coast, Nigeria and Libya might have brought some of our respondents back to Ghana. I feel bad for those who might have lost their livelihoods in the process and wish them the best back home in Ghana – but the point is about the unpredictable nature of tracking respondents in the long-term.
Another effort at long-term data collection I am familiar with in Ghana is the Ghana Socio-economic Panel Survey. The first wave of this survey was completed last year and there is some really interesting data being put together that we should all be watching out for. This 15-year survey covering households spread all over the country will surely see huge tracking challenges as we go along. In this survey, we put in place many of the possible long-term tracking measures – GPS readings, phone numbers of two non-family contacts etc. Tracking households is supposed to be easier than tracking individuals since they are likely to be less mobile than individuals.
Also, this survey is a collaboration between Yale and ISSER, a research institute in University of Ghana – a great way to invest in building local research capacities in the long-term. Not only the ISSER research fellows, but a whole generation of present and upcoming grad students will have the opportunity to work on this survey and on analysing the data it will generate. This also highlights the importance of having a strong field team and transferring knowledge over the years within the survey teams in long-term tracking studies.
Another question these long-term tracking studies bring up is that of the relationship between researchers and respondents. There are usually boundaries to how much information researchers want/ought to reveal to their study respondents regarding the objectives of their study. But in long-term studies with repeated interactions, the nature of engagement between the researchers and their study respondents probably need to be different from the usual – an important issue to keep an eye on. Do respondent incentives become more important? Would sharing results periodically help? I would be interested in learning of any experiences on this subject.
Don’t forget to keep track of the respective study links for updates in the coming years.