We found that in the EU, poverty and poor mental health, among the unemployed and even the employed, are positively impacted by the generosity of Unemployment Insurance (UI) in terms of its criteria for eligibility, duration and wage replacement levels in the EU. Though UI is not intended to compensate fully for a loss of earnings, generous UI programs can moderate harmful social and mental health consequences of unemployment.

2) How did you find the experience of applying a realist lens to the study of the topics covered in SOPHIE?

Very rewarding scientifically since it forced us to identify the mechanisms linking social protection to health. Most studies just report associations without much thinking behind them: they do not care about how things work and thus provide weak support for policies. Whether this attitude obeys to a pragmatic view of social epidemiology to focus on maximizing papers, or to avoid policy positions or both is unclear to me. Yet we found a sufficient number of studies that contain enough information where we could glean mechanisms and allowed us to understand why some policies are helpful and others aren't. The methods we applied in SOPHIE have been a great opportunity to advance our knowledge on how some of the most important social protection policies work. In my view realism adds depth to empiricism, while pragmatism, at the end of the day, is not fully scientific (although it can get you more papers).

3) You have worked in Europe, United States and Canada. In your opinion, what are the main differences between the three in terms of political strategies to tackle health inequalities?

I know better the US, where I worked close to 20 years, then Catalonia and Canada of late. I would say that, at least before the current recession where (mostly) German banks have forced a dismantling of Spanish social protections, the US welfare regime would have the most tolerance for inaction around health inequalities. I lived in and around Baltimore, and the recent events involving police repression do not surprise me at all. On the other hand, one can see that those affected most by these inequalities in Baltimore and similar US cities are not as they are depicted in the racist show 'The Wire'. The criminal element is a minority, and people are organized around social determinants of health. Unfortunately, they have no means to influence politics. Canada has less tolerance for health inequalities due to a single payer health care system and a public educational system. Given the current political climate, whether these institutions will be there in ten years remains to be seen.

4) You have been working in this field for 30 years. How would you rate the 'health' of research on the determinants of health inequalities today?

I started as a behavioural pharmacologist and social psychologist, so it's more like 25 years still quite a stretch. It's been rewarding to see how the field of SDOH and Social Epidemiology has grown. In the US the 90s were very exciting in that respect. Later on, and even more important, the WHO Commission on Social Determinants of Health among other efforts brought global acceptance to research on social inequalities in health. Although politics can be responsible for the ups and downs of this field, the body of knowledge is too large and the social impact too critical to vanish easily. There is a cumulative effect like in any other science. Unfortunately many leading researchers focus more on the complexity of methods than on the complexity of how society shapes health and disease. You can improve endlessly the way in which you show an association between 'SES', 'race' and mortality, yet it is still a 'black box', pre-scientific finding. Yet I am optimistic, as far as there are new questions and new findings, SDOH and social epidemiology will continue to grow.

5) Finally, you are a professor at the University of Toronto. What do you say to your students to encourage them to think about how structural policies influence health and health inequalities?

I feel lucky to be at UoT where I get students from all over the world, literally. Half of the residents of Toronto are born outside Canada. I have students from Latin America, Asia, Europe, Africa and, of course, North America. Thus, many students are aware of global inequalities in health. Yet I have often to provide evidence and debate 'small intervention' approaches, modern philanthropy, 'descriptive epidemiology' mindsets that skew policy, and postcolonial worldviews. To encourage my students to think about how social policies influence health inequalities I remind them that this is part of their scientific duty. The ethics of public health as a social technology and of epidemiology as an applied science require action on the world. Social epidemiology is not a basic science like astronomy or quantum physics. You also need to work with stakeholders since the people who are affected by your research should have a say. Democratic values are part of public health ethics. So you need to study the best politics and policies that can provide greater health equity. In order to do that you need to understand social causation of health inequalities (what is it in a policy that produces the desired change in population health), which brings us back to realism.