What's your research about?
'The short version: I measure democracy. The long version: I am Director of the V-Dem Institute, which is the largest ever social science endeavour to measure democracy. V-Dem was developed because existing resources were comparatively small, used very narrow conceptions of democracy, had various methodological flaws, and sometimes questionable face validity.'
What makes the V-Dem dataset so special?
'The V-Dem dataset is the world’s largest on democracy and unprecedented with more than 19 million observations on human rights, media, civil liberties, and political institutions. Behind the dataset is a complex research infrastructure and many years of effort by a large organization comprising six Principal Investigators, 17 Project Managers, and over 3,200 Country Experts.'
'V-Dem radically alters the way scholars can address questions of the causes and consequences of democracy. V-Dem is also distinctive because it provides five indices of democracy, and dozens of indices of separate components of these, that are specifically designed to map onto the most prevalent theories of democracy: electoral democracy as "polyarchy", liberal, deliberative, participatory, and egalitarian democracy. V-Dem is also one of the very few democracy measurement programmes that takes measurement error seriously. Using Bayesian methods, we estimate how certain we can be about each data point and we make this information freely available to the public. In sum, V-Dem radically alters the way scholars can address questions of the nature, causes, and consequences of democracy and autocracy.'
How did you become interested in this?
'I believe that a lot of good science originates from a personal engagement that can be either simply a fascination with a natural or social phenomenon, or coming from some personal experiences and convictions. I come from a family where engagement in the cause of human freedom and autonomy at both the individual and societal level has always been very strong. It eventually led me to focus my research on democracy and the causes of democratization.'
'Our work provides results that are used by pro-democratic forces.' 'While democracy in practice is always flawed, the alternatives often offer so much more oppression and denial of human dignity. Our work provides results that can be (and are being) used by pro-democratic forces such as state departments, development organizations, international and national NGOs, to further democracy and human rights in the world.'
Have you made any interesting discoveries/results?
'Yes, actually quite a few, and a couple that I find particularly important. First, the relationship between democracy and development has been analysed and conflicting results debated over the past 60 years. With the reach and depth of V-Dem data we have been able to provide a resolution to this debate, showing that democracy is good for growth and for human development measured in terms of infant mortality rates, for example. And we have shown that it is in particular the electoral aspect—that elections are clean, free and fair—that is doing “the job” in terms of the positive effects. Previous research, based on the less sophisticated data available, was inconclusive.'
'We have shown that it's free and fair elections that's doing “the job” in terms of the positive effects. Secondly, we have also discovered something about the importance of order in the development of various aspects of democracy. The question is: when a country is in an early phase of opening up (liberalization), should they invest support primarily in strengthening civil society, or the media, or the judicial system, or something else? We have now fresh results from the FASDEM project which demonstrates that successful episodes of democratization are distinguished from failed ones primarily by the early establishment of a high degree of autonomy for election management bodies.'
'We have also developed a new set of advanced methods for causal identification using observational data.'
'We have also developed a new set of advanced methods for causal identification using observational data, which shows that there is a true causal effect of such autonomy on the success of a democratization process. This is a very important finding. This is highly significant for organizations and actors that want to support democratization.'
Have these results surprised you?
'Yes and no. Both findings have strong theoretical intuitions. But at the same time, we are surprised by the size and robustness of the results, since that is quite exceptional in the social sciences.'
What are the future challenges in your research field?
'We really have to learn from other scientific areas how to make better use of observational data for making both descriptive and causal inferences to recognize the multi-causal and interactive, and sometimes recursive, effects that drives the true world of social phenomena. We also have to be much better at making our findings available and inform decision makers and media, while at the same time fiercely protect our academic freedom.'
