ID Conference: Michael Dorsch on Evidence-Based Policymaking

Dr. Michael Dorsch, the associate professor of Economics and the director of the Master of Public Administration program at the School of Public Policy in Central European University, will deliver a keynote presentation on evidence-based policymaking. We have talked with Dr. Dorsch about his research project, as well as his views on policymaking during the COVID-19 pandemic and shift of academic activities to online.

Photo courtesy of Michael Dorsch / Central European University

Photo courtesy of Michael Dorsch / Central European University

– What topic are you going to present at the Conference?

 – I am presenting a paper that is coauthored with one of my CEU colleagues – a political scientist and economist from the University of Paris. This is the second one in the series of papers that we have on the broader topic of political economy in evidence-based policymaking. In this paper, we are looking at the role of polarization within this context, and specifically at randomized control trial experiments as outcome variables. The broad idea that runs throughout the entire project, not just in this paper, is that evidence-based policymaking and its tools have political implications for the implementers, and therefore are subjected to be used in political ways. And if you want to think statistically about the implications of that, this is actually quite serious. The universe of randomized control trial experiments that we have is not a random sample, it is biased due to political selection. We think that it is actually a quite substantial thing to say. And this project is really trying to establish the broad methodological point and to work through a mechanism that is driving this political selection effect. The paper I am presenting is a step in that direction.

 – Can you briefly describe the role of evidence-based policymaking in contemporary governance?

 – Since our focus is on randomized control trials, and most of these experiments are done by development agencies, our theoretical and empirical focus, at least in the paper, is more on developing economies. When you think about the politics of randomized control trial experiments, you have a supply-side effect and a demand-side effect. On the supply-side we have academics, scientists, researchers, NGOs, development agencies, charitable foundations, who are financing these types of experiments oftentimes to promote their own preferred policies.

On the demand side we have governments who are also commissioning randomized control trial experiments to figure out what works and to use experiments as advocacy. Governments may have policies that they prefer, but they may have opponents who are unconvinced that this is a good policy. Randomized control trial experiments are a seemingly non-partisan method of advocating a policy.

If you stick with the demand side, even when a policy experiment is initiated by the supply side, a local government has to sign off on it. They are always going to have a veto over a proposed experiment in their jurisdiction. Even though researchers may think that they are the ones driving the way of randomized control trial experiments, at the end of the day the government is able to put a stop to any policy. The government has to be willing to host the policy. In Hungary, for example, some of my colleagues were proposing to do experiments on social inclusion in primary school education. They wanted to see if Roma kids were affected differently by a public-school reform, but the government would not let them do it.

This paper that I will present is drilling into this demand side. In this paper, we do an in-depth case study of Liberia. I will elaborate on that during the presentation itself, but the outcomes of the experiment were taken by the government into consideration, and politicians even campaigned with them for the elections. Also, the paper has a formal model that sets up the politicians’ decisions, and then we have a cross-country panel analysis.

 –  Are there any signs of evidence-based policymaking in the current COVID-19 policies? Are there any prospects for its application in the future?

–   In terms of policy responses, I am not aware of any explicit experiments that have been done. I think that would be very difficult. Think of the ethics of that. You would want to know how effective mask-wearing is, for example, and you would literally have to tell some citizens: «you are a treatment group and you are going to wear masks», and nothing to another group, and then compare the outcomes in terms of contagion rates.

More broadly though, it seems that we are starting to gather evidence about effective and ineffective policies. The politics of this particular experience seems pretty messy. I am originally from the US, so I follow politics there pretty closely. And there, the politics do not seem to be in a state where the two sides can be convinced of what is a good policy based on scientific research and evidence. The division between the two sides has broadened and deepened to such a degree that it is not clear, what kind of evidence about good COVID policy would be salient to both sides. In Europe, there seems to be more respect for the role of data and experts in this debate. And the segment of the population that may not be convincible by data seems to be a more fringe kind of segment of the population. In a lot of ways, the extent to which evidence-based policy responses to COVID will be implemented through a democratic political process depends on the extent to which the political debates still have a respect for empirical facts.

– What are your expectations of the Conference, and what do you think about the online conference format?

– I was supposed to come to the Conference last year, but I was not able to come. At that time, I thought that it was not a big deal – I will get to go to St. Petersburg next year. Unfortunately, it did not pan out that way and I will be joining virtually. Hopefully, I will have a chance to make it to St. Petersburg some other time. By now I have quite a diverse set of academic interactions over Zoom and online. From finishing the spring semester on the fly (nobody even heard of Zoom back then) to a series of seminars, a short course, and conferences’ cancellations that got moved on Zoom. And this academic year in the CEU we have half classes live and half on Zoom. My prior was that online education and academic interaction can never be as good, but I am starting to shift. We can still have meaningful and intense academic exchanges. In a lot of ways, it democratizes the whole enterprise. As an example, we have a lot of students from the Global South. Some of them, even with nice scholarships, simply could not afford to move to Vienna. But now they are following the classes online and this is satisfying, I think.

Interview by Angelina Silaeva