Special News

A Joyful Revolution in Econometrics: How Josh Angrist Transformed a Field, While Making It More Fun

I was a junior in college when the textbook Mostly Harmless Econometrics by Josh Angrist and Steve Pischke first hit bookshelves, and its timing could not have been better. I had just finished my first course in econometrics—the statistical toolkit economists use to bring data to theory—and I was excited to start applying these tools to my undergraduate thesis. But putting this coursework into practice turned out to be challenging.

Despite attending hours of lectures and completing pages of problem sets, on rather dry mathematical foundations, I had little clue how econometrics was actually used to answer complex real-world questions. Mysteriously named methods like “instrumental variable regression” were taught as powerful levers for disentangling cause and effect from the various correlations we observe in data. Yet harnessing this power to separate correlation and causation, using only the arcane mathematical formulas from class, seemed like a daunting task.

In many ways, Mostly Harmless represented a sea change in econometrics, a movement towards practical and credible tools for causal inference, which Josh Angrist and many others have helped usher in over the past three decades. Today, his service in this methodological revolution has earned Josh, David Card, and Guido Imbens the 2021 Sveriges Riksbank Prize in Economic Sciences in Memory of Alfred Nobel.

Much of the official background for the prize focuses on the three laureates’ many methodological contributions, which have given economists new tools to measure the causal effects of real-world polices. There is no doubt these advancements have shaped modern empirical practice and improved the credibility of inquiries into several important topics, from the value of college to the effects of minimum wage laws on employment and more.

But to me, and I suspect to many other recent students of economics, Josh’s impact on the field goes well beyond these methodological insights. Through his teaching and mentoring, Josh has been equally triumphant in making the recent econometric revolution accessible, useful, and downright fun. This democratization of the field is typified by Mostly Harmless, as a clear and approachable handbook to how modern causal inference methods can actually be used and not just studied (one that sits on the bookshelf of virtually every applied economist I know).

Alongside its clarity and practicality, Josh’s teaching also has a unique humor which seems central to its effectiveness. For such a serious scholar, Josh clearly has fun doing econometrics. I suspect this infectious joy for rigorous empiricism has led many students, researchers, and policymakers to the revolution celebrated by this year’s Nobel.

Examples of Josh’s fun with econometrics are everywhere in Mostly Harmless, perhaps the only textbook ever to riff on Douglas Adams’ comedic sci-fi novel The Hitchhiker’s Guide to the Galaxy. Like the novel’s eponymous encyclopedia, the textbook markets itself as a practical and only “infrequently inaccurate” guide (with the latter compensated by it being “quite a bit cheaper” than other econometrics textbooks).

Josh and Steve begin their book by discussing what economic questions can be learned from empirical investigation by referencing Deep Thought, the supercomputer from Hitchhiker which gave a famously funny answer to “life, the universe, and everything” (“42”).  The deep thinkers behind Mostly Harmless give their own comedic conclusion about the limits to empiricism. They draw a distinction between the Frequently Asked Questions (FAQs) of good research—including what ideal “experiment” defines the causal relationship of interest—and what the authors term Fundamentally Unidentified Questions (FUQs). A research question which cannot be answered by any hypothetical experiment is, they dryly conclude, FUQ’d. The puns only get worse from there.

Josh’s humor is memorable, despite—or perhaps because of—its occasional corniness. I often hear quotes from Mostly Harmless in academic seminars, from students and colleagues who remember the puns surrounding its profound methodological insights. Sometimes one also hears references to “Master Joshway” – Josh’s whimsical Kung Fu inspired persona introduced in his follow-up book, Mastering ‘Metrics.

The occasional silliness of Josh’s econometrics courses is also the stuff of graduate school legend. Several years after getting my first copy of Mostly Harmless in the mail, I was lucky to start an economics PhD at MIT and learn from the master himself. Among the decades of wisdom packed into each of his lectures, I was never sure when an elaborate analogy to a Battlestar Galactica plotline might slip in. In lectures on recent econometric analyses of gig worker labor supply, students might instead hear funny mishaps from Josh’s (quite poorly reviewed) stint as an Uber driver.

All this silliness has serious pedagogical power, in both cementing difficult econometric concepts and humanizing the often-intimidating mathematical field. Recently, the power has itself been causally established in an experiment. Exposure to Josh’s style of econometric teaching, it turns out, can increase policymakers’ responsiveness to (and demand for) compelling causal evidence.

I was lucky to have been exposed to Josh’s joyful econometric revolution in college and graduate school, but I am now even more fortunate to help continue it. At Blueprint Labs, the research group Josh co-founded and directs, we put his many methodological discoveries to work daily–finding answers to challenging questions in education, healthcare, and labor markets.

On especially exciting days, we learn something new about how econometrics can be used to untangle complex causal relationships. A recent Blueprint Labs paper, for example, shows how instrumental variables regression (the same method I struggled to understand, before Mostly Harmless) can be used to improve high-stakes measures of public school quality.

But beyond these new discoveries, Josh continues to work tirelessly every day to make the magic of econometrics accessible and fun to millions around the globe. It is an honor to help continue this mission in my own research and teaching, occasional corny puns and all.