Three years ago I was asked as a non-economist to give a talk about econ teaching in public policy. I gave my talk the title Making Econ a Softer Science. My econ colleagues were surprised because they thought econ needs to be more empirical and even harder, i.e. more scientific. Now, as a political economist, I like the new turn to empirical and behavioural economics. And I like (big) data, but as a public policy person I don’t think that the image of hard sciences is particularly helpful.
As a matter of fact throughout the years I have encountered two extreme and radically opposed ideologies in my field(s): Ideology A states that social reality cannot be measured and that statistics in social sciences is useless; ideology B states that social sciences will only be real sciences if they emulate X with X = econ < physics < maths. In other words: the harder the better. Needless to say that stated in such absolute terms I find both equally rubbish.
Ideology A is often found in the context of international relations and some spins of political theory. I don’t want to engage in an epistemological debate about how to approach social reality. I just think that, first, there is a lot of stats in social sciences and that, second, not all of it is bad (although I do admit that a surprisingly large share is really bad).
Ideology B is often found among my fellow regression-junkies in political science and economics. I think they suffer from the famous condition of physics envy. In a strong dose this condition is very dangerous. There is a lot of reason to believe that the global financial crisis was in part the result of famous macroeconomists too readily believing that they know the iron laws of monetary policy. In a field that is dear to me, microfinance, there are also examples of scientific studies (in top US journals) that have been used to legitimize the boom in the industry although these studies later on turned out to have severe methodological issues. The politics of evidence based policy making is all too clear in these examples.
I think, given the ‘nature’ of social reality, social sciences have to be soft sciences. I also think that there is nothing wrong about this, as long as you admit it. I hasten to add that there is nothing wrong with scientific rigor. Trying to be precise, transparent and to talk in a language anyone could understand is good practice in all fields of social sciences. However, the only perfect case of rigor in the social world is rigor mortis. You kill the subject if you scrutinize it too hard.
For me social reality in social sciences is (ontologically) soft or, more precisely a mixed bag: some bits are tougher and lend themselves more towards counting and estimation, others not so much. The best simile for me is to compare social reality to the human body. Yes, there are many parts that are hard and easily measurable, such as the spine or the rib cage. These parts are hard enough to make them quantifiable. They also give some structure and stability to social reality. But that doesn’t mean that there are universal: bones can rot, rips can break. Similarly, there are relatively few hard laws in empirical science (can you name one? I sometimes believe there is one between violence and inequality link), and even these don’t exist in all temporal and spatial contexts.
Looking from the outside, what defines the human body is its organic shape, its fleshy bits, which vary from person to person, and even from day to day. These bits are harder to measure precisely, but they are still an important part of social reality. Nobody would be able to identify a person by looking at her bone structure. Similarly nobody would be able to identify a country or a political party without the fleshy bits, there idiosyncratic leaders its contorted history.
And in similar ways to good (holistic) human medicine, social reality is best understood using the diversity of methodological tools at your disposal, including quantitative and qualitative methods. This also implies that instead of squeezing social reality into a methodology that was developed for the physical world, methods in social science have to adapt to its own reality. This might also imply that we have to say farewell to the idea of cumulative knowledge, since social reality is reflexive and mobile and includes sciences themselves. But it is better to embrace this softness and deal with it, than conflating a skeleton for a human being.