As Adam McCloskey delved deeper into his undergraduate honors thesis, he found himself becoming increasingly frustrated. He was focusing on big data sets, trying to tease out relationships between different variables (volatility in exchange rates and how volatility causes certain types of investments to flow into and out of countries, for example). As he applied various mathematical models to the data, he realized that each model was giving him a fairly plausible answer, and that there wasn’t one that stood out as better than the another. Looking at other papers on applied economics, he noticed that other researchers were having some of the same issues.
“How did they know this was the right model and how did they know these results are accurate?” McCloskey wondered. “Every model we write down is an approximation of the true data-generating process.”
It’s that frustration that led him to the research he’s currently working on, finding ways to account for the fact that applied economists use data to select models and then work under the assumption that the model is correct, a process called inference after model selection. “I wanted to contribute to being robust about the fact that we don’t really know the true data-generating mechanism. Our models are probably all incorrect and even if they are correct, there’s uncertainty involved in deciding what the true model is,” McCloskey said. Eventually he’d like to take his research even further, establishing a process that can determine the best model selection procedures and inference procedures after model selection that are the most informative to the user.
“I’m an econometrician. I lean more toward being on the theoretical side, but everything I do I want to be useful, I want applied economists to be able to use it. A lot of what econometricians seem to do these days is maybe not so useful for an applied economist, so this is something that I have in mind when I do research: I want it to be useful.”
As an undergraduate at the University of Colorado–Boulder, McCloskey said that mathematics was a natural fit. His particular interest was in time series analysis, looking at data recorded over time to forecast future trends. McCloskey found that he took a particular liking to analyses of economic and financial data and decided that econometrics was the field for him. He earned his master’s and Ph.D. in economics from Boston University. He’s spent the last year teaching at Brown as a visiting assistant professor in the economics department.
He said he’s excited to be making the position a more permanent one and to be coming into the economics department at a time when it’s growing its econometrics group.
“It’s a great economics department, one of the best in the country. Usually econometrics makes up a small portion of an economics department, but there are five of us, so it’s comparatively big,” McCloskey said, who adds that much of his research overlaps with others in econometrics, which will give him an opportunity to learn from his colleagues and maybe collaborate on future projects.