Scholars should not fall for the safety of numbers

PEGNet was started in 2005. The idea was to link academic development research to policy makers and the staff of development agencies. This is a worthy cause. German scholars tend to be proud of focusing on nothing but their research discipline and many feel that, if lay persons understand their academic work, it cannot be very good. Too few professors disagree with this notion, and Germany lacks a tradition of public intellectuals. In this setting, PEGNet provides open-minded scholars with a forum and sets a good example for young researchers.

I have personally attended three PEGNet conferences, and all of them were impressive. This time, however, it seemed to me that too many contributions were a bit misguided. I'll give an example. One young researcher presented an econometric model she had designed with a colleague to find out why exactly women with higher education in a particular African country have fewer babies than those who are forced to marry in their teenage years.

This phenomenon can be observed all over the world. Reasons include that women who go to university are better informed about the use of contraceptives, and that they are also better placed to convince their partners of using them. Women with university degrees, moreover, have opportunities to earn high incomes, which is an incentive to postpone becoming a mother. Obviously, women who complete secondary school and university are busy and cannot afford to take care of children at the same time. In many developing countries, female students live in dormitories that are strictly regulated, so they hardly have opportunities to have a sex life.

I don't think we really need to know which of these reasons is the most important one in country X or country Y. It is certainly fascinating to develop fancy mathematical models and find statistical data to test hypotheses. Getting robust results this way is certainly an intellectual achievement. But is that knowledge helpful? Social norms that are valid today may be outdated tomorrow. Gender roles are changing fast. Data collected in an urban environment is not representative of predominantly rural societies.

My point is not to single out a particular scholar, but her example was telling. There were many other presentations of a similar nature, and PEGNet participants seemed happy to discuss the intricacies of econometric modelling in great detail, rather than cast doubt on the relevance of that exercise. My impression is that many young researchers are eager to prove causality beyond doubt and accept the assumption that numbers crunching is the way to do that. The problem is that no econometric work can deliver results that are more interesting than the hypotheses it is designed to test. If the hypotheses do not have policy relevance, the econometric calculations won't either.

We are living in uncertain times. Advanced economies have not returned to normal since the collapse of Lehman Brothers. In view of the euro crisis, the EU looks weaker than ever before. The economies of most emerging markets and developing countries are witnessing a downturn. Fragile statehood and civil strife are affecting many world regions. Climate change is a challenge the international community has not managed to get a grip on, even though permanent negotiations on the matter have been held since the Earth Summit in Rio in 1992.

Yes, population growth is a huge problem too, but we know that educating girls and women is the answer. We need to know how to educate many more girls and women, and do so fast. Why exactly education has this impact doesn’t matter that much. I’d happily leave it to the pope to worry about whether educated women have fewer babies because they use contraceptives, make their partners use them or abstain from sex altogether, but I’m not even sure Pope Francis is very interested in the matter.

Humankind needs research that delivers answers to pressing problems. Young scholars who focus on these questions, however, risk to not come up with mathematically clean results. For career purposes, inside or outside academia, many seem to prefer the safety of numbers that prove some kind of causal relation. That is absurd in a multi-causal world, but the incentives young scholars have make them play it safe in uncertain times.

There is an even more depressing angle to this matter: I'm not sure many political leaders are interested in sophisticated, science-based policies. Too many seem happy to resort to simplistic populism. Just consider the Republican Party in the USA today – many of its leaders are science-deniers, especially, but not only, in regard to climate change.