Causation in demography: procedures, problems and alternatives

Ignacio Pardo

Formal demography used to limit its scope to the measurement of different population processes and dynamics, describing and predicting trends. But within the field of social demography broad and accurate explanations are needed, because our discipline is becoming much more integrated and theory – driven. Then, causal attribution is a key component of many studies and should be based on explicit procedures. In this paper, I will reflect on the different notions of causation. Besides, in terms of research design, the new framework of Mixed Methods research (developed in the last decade upon an old debate between quantitative and qualitative traditions) will be introduced as an extremely useful approach. In medical trials and in social sciences in general, contrafactual causality is the “traditional” way of achieving a valid link between cause and effect, embedded in experimental and quasi-experimental strategies. But these strategies have certain shortcomings (narrow account of phenomena, context blindness) and in many cases they are just impossible to develop, for ethical or practical reasons. Then, alternative and complementing notions of causation, such as generative causality, are being taken into account. In terms of research designs for demographic issues, Mixed Methods strategies should be designed when needed (this is usually the case in fertility and migration studies), embedding a broader account of causation. Considering that in methodological and mixed methods reflection, increasing attention is being paid to notions and techniques such as “modus operandi”, “modus narrandi” or “reasons as causes”, demographic studies can benefit from this innovations. With Mixed Methods designs it should be easier to achieve broader and better explanations, regarding migration, family change or fertility. For example, we can complement findings on fertility determinants with those on the “culture of reproduction” (Laura Bernardi) and see the whole picture for causal attribution: structural determinants and subjective reasons.