Working against helicopter science, one step at a time

The phrase ‘helicopter science’ is used to describe the practice in which a researcher with superior resources, typically from a developed country, goes into a developing country to collect data, only to travel out and use the data elsewhere for whatever they want.

This practice is problematic for several reasons, not least because it leaves almost nothing of persistent value behind for local researchers and residents, other than, say, remuneration for time and logistics. It also often limits the contextual relevance of research while making ‘heroes’ out of alien researchers who know little about their study context.

Helicopter science is a problem in the social sciences too

The existing discussion around helicopter research has been concentrated on the natural and biological sciences. Action has been limited to these disciplines, such as when a group of African scientists developed a guide for the use of genomics data from the continent.

Nonetheless, it is a real problem in the broad social sciences, and specifically in development research. Common sense tells us that local researchers will bring a deeper appreciation of the social, political and economic contexts to policy discussions on economic development. Yet, they are only marginally involved even by the so-called development partners.

A recent study reports that only 0.2% of the authors of the World Bank Development Report between 1978 and 2020 reside in a developing country. Of the authors who wrote or advised on the UNDP Human Development Report from 1990 to 2020, only 12.3% lived in a developing country.

Data availability helps to overcome helicopter science

Overcoming the challenge of helicopter science requires much more than mere talk. One of the reasons often advanced for the marginalisation, apart from a pathological paucity of research funds in developing countries, is lack of good data. My colleagues and I recently completed a study of the social science research system in Nigeria and we found poor data access to be a major problem even for in-country researchers. Out of the approximately 500 researchers that we sampled, about 50% were dissatisfied with their access to primary data for research. This is likely worse in disciplines that involve expensive laboratory work.

One take-away from this finding is that we can speed up the decline of helicopter science by collecting, curating and disseminating high-quality data on topical issues from developing countries.

In the last two years, my colleagues and I have made some contributions in this regard.

1. With funding from the Global Development Network (GDN), in the framework of the Doing Research programme, we created the dataset on the production, dissemination and uptake of social science research in Nigeria. The data was used in this report and this article which we summarise here. We used the data in this article too. While we await some more output, I believe the data has potentials for much more.

2. With funding from the Private Enterprise Development in Low-income Countries (PEDL) programme of the Centre for Economic Policy Research (CEPR), we created a pooled dataset on entrepreneurial characteristics of undergraduates in selected Nigerian universities. The dataset was used in this article which I summarise here, but has potentials for much more.

3. With funding from PEDL, I merged, cleaned and standardised data from two rounds of Nigeria’s official national innovation surveys to produce a pooled cross-sectional micro-level innovation dataset in Nigeria. This dataset has become popular, having been used multiple times by several authors not affiliated with my home institute: here, here, here, and here.

Building the capacity of local researchers for globally relevant research can also help

Some argue that helicopter science exists in the first place due to weak capacity for research in developing countries. Researchers from developed countries are therefore left with no choice but to implement their own studies but recruit local researchers only for low-value activities such as transcription, interpretation or tedious fieldwork.

Recently, I authored a chapter for a collection on African entrepreneurship research. The chapter draws upon two decades of research experience in Nigeria and other parts of Africa. I argue that researchers need high-quality national survey data to answer a wide array of policy-relevant questions, but such survey data are sparse in developing countries. While small surveys help to fill the gap, they are generally difficult to implement and often of poor quality.

I then proceed to highlight four major challenges that researchers face in implementing surveys in Africa:

1. weak demand for research evidence among policymakers;

2. poor infrastructure;

3. weak domestic capacity for well-designed and well-implemented surveys; and

4. survey overload that seems to have precipitated some form of societal apathy.

I offer seven tips to overcome these challenges:

1. use careful and rigorous sampling methods;

2. resolve population frame problems thoroughly;

3. ensure survey instrument are well designed;

4. adopt solid survey administration techniques;

5. engage with stakeholders;

6. offer some incentives, preferably non-pecuniary; and

7. adopt a sequential procedure.

For the interested reader, a preprint of that chapter is available upon request via my ResearchGate profile. For those who wish to take a step further, I have a 2017 piece that provides a primer on how to choose the right statistical tests for empirical research. It is freely available.