In this video, Geert Mesters (UPF and BSE) talks about his ERC Starting Grant project that proposes a new framework for evaluating macroeconomic policy decisions.

Video transcript

Geert Mesters:

My name is Geert Mesters, and I work at UPF in the Barcelona School of Economics and my ERC project develops new econometrics methods for evaluating macroeconomic policy decisions. I focus on answering two key questions. First: is the proposed macroeconomic policy choice optimal? And second: if I find that it’s not optimal, why not? Which variables, shocks, dynamics were overlooked when setting policy?

Now, as an example, you can think about a central bank that proposes an interest rate choice, or a government who sets a carbon tax. In each case, you can ask: is that proposed policy choice optimal? And if not, why is it not optimal? These are pretty simple questions to state, but it turns out that they are quite difficult to answer using the existing methods. The reason is that in practice policymakers tend to combine many inputs when reaching policy decisions. They combine multiple models, instincts, judgments, and they put all that together to reach their final policy decision.

Somewhat in contrast, the traditional approach in economics when doing policy evaluation is to analyze the specific model for the economy. We can learn a lot from this approach, but a fundamental concern is that such a model can be mis specified and therefore you cannot use it to evaluate those practical policy decisions.

Now, in my project I aim to address these concerns and the key new idea is that I will approach policy evaluation as a hypothesis testing problem. The key benefit of this approach is that it allows me to fix proposed policy choices under the null and therefore I can avoid the evaluation of counterfactual policy choices and I do not need to specify a specific model for the economy.

Now, concretely, I will use this hypothesis testing approach to handle many practical aspects of policymaking. In practice, these methods can be used by at least two different parties. First, policy makers themselves can use them to avoid non-optimal policy decisions. Second, researchers more broadly can use these methods to retrospectively better understand policy decisions and possibly make model improvements.

I will use these methods myself, as well, to evaluate modern monetary fiscal and climate policy decisions and using this approach then we can construct a unique non-model-based overview of modern macro- policy decisions.