IOP Institute of Physics

For physics • For physicists • For all


Members can try our new IOP Career Development Hub

visit the new Hub

Once a physicist: Dave Donaldson

Dave Donaldson is a professor of economics at the Massachusetts Institute of Technology, US, who conducts research on topics related to international and inter-regional trade in low-income countries. He grew up in Toronto, Canada and obtained an MPhys degree at the University of Oxford, UK, in 2001

What sparked your interest in physics? 
I was always curious about some of the basic physics of everyday life, and my father (who did an undergraduate degree in physics, and was himself raised by a PhD physicist father) did an inspirational job at stoking that basic interest over the years. Later on in school, I saw the beautiful role of mathematical thinking in physics and fell in love with the way these simple equations could formalize the intuition my father had taught me (and yet also be used to extend, bravely, into less intuitive and everyday domains such as quantum physics), and could do so with such remarkable empirical success.

Did you ever consider an academic career in physics? 
Yes, absolutely. And my physics tutors were very supportive in that. But, looking back, I think I just didn’t have the maturity to see what I now recognize as the huge social value of fundamental research. I was attracted to other, more applied problems – like the ones I saw in economics or business – and it took a while before I realized that research was what I really wanted to do. By then I was hooked on economics, but I still try to keep up (obviously in an amateurish and simplistic way) with modern physics research.

How did you get interested in economics? 
I was captivated by just how different economic living standards can be across countries – how could it be that a barber delivers exactly the same haircut in London for ten pounds and in Lima for one? Books like William Easterly’s The Elusive Quest for Growth showcase the daunting intellectual challenge posed by the puzzle of global inequality and the trail of failed policy thinking on how to alleviate it. At the same time (this was the early 2000s) there were all these prominent “anti-globalization” protest movements – and unmissable events like currency and exchange-rate crises in emerging markets – and I wanted to learn the scientific understanding that economists had of the issues at stake.

What are some of the challenges of working with big data when it comes to economics? 
We face two big challenges. One is that we are trying to understand a relatively complicated system, and the other is that (for the most part) we don’t have direct experimental control. So we try to find and focus attention on what we think of as “natural experiments” – settings where it is plausible that one element of the system was (in a sense) randomly allocated, so that we can study its effects on other outcomes. It isn’t easy to find cases to satisfy that criterion, and there are many phenomena we’d love to study this way but can’t, but I don’t think there is really any other way to figure out how a complex system works.

I came across a simple case of this at work in a recent paper that tries to estimate the effect on a US firm (for example, on its profits and behaviour) of being granted a patent. Of course, no social scientist has direct experimental control over the allocation of patents – for good reason! But the authors of the paper spotted that there is a part of the US patent application process that is effectively random – namely, the process that assigns a patent examiner to individual patent-application files. Since some examiners are, on average, stricter than others, being randomly assigned to a lenient examiner effectively gives a random boost to the chances of your patent being granted. Isolating just that variation allowed the authors to mimic a randomized experiment in patent allocation, and to thereby start to figure out how the patent system could be better designed.

You recently won the American Economic Association’s John Bates Clark Medal – what was that like? 
It was a huge honour and an even bigger surprise. These kinds of awards are arbitrary at the best of times, but in this case, it feels especially so because of the fact that I have done virtually all of my work with collaborators who deserve just as much (if not far more) of the credit. But it is certainly a big boost of confidence for the style of work that my collaborators and I, and the many others who work on related themes, have been doing.

What are you working on now? 
Our access to data has recently exploded thanks to the fact that so many organizations are collecting data for their own purposes, and are sometimes willing to share (anonymized, confidential) versions of it. That development has had a big impact on the ways I work. For example, I have projects that seek to model and measure how road improvements affect traffic congestion (using cell-phone data), and whether China’s incredible high-speed rail network was worth its equally incredible construction cost (using geocoded data on all credit card transactions in China for the past five years).

At a more methodological level, I am troubled by the daunting task that economists face in using data from the past to answer important policy questions about the future – for example, what are the economic consequences of a change in trade policy under Brexit – and have been trying to find ways to improve our ability to do that. Given a policy question and given some data on the past, what is the minimal set of theoretical assumptions about the way the economy works that are needed to answer the question? If we understood that frontier then we could more accurately and transparently convey to the public the limits of our knowledge.

How has your physics background been helpful in your work, if at all? 
It has been a massive help. Fundamentally, it taught me the power of putting trust in logical, mathematical thinking that takes you all the way from first principles to the final conclusions they imply. My physics tutors at Oxford taught us that above all else, and it stuck.

Any advice for today’s students? 
Work hard! There is so much to learn in a physics degree and I don’t think anyone could really predict how any particular piece of it will help you with whatever you end up working on in later life. In my case I can definitely say that, looking back, anything that was hard was worth fighting through – since that was where I learned the most – and anything that seemed irrelevant was pretty likely to end up useful sooner or later.