To better understand an issue, researchers look to figure out cause and effect. Drawing that causal connection in studies of real-life situations, however, isn’t always clear-cut. Variables are also at play.
And while conducting randomized controlled trials are the gold standard for research in many fields, that’s often not possible for economists probing central questions facing society, such as how minimum wages or immigration affect the labor market or how a new policy impacts economic activity.
But Stanford economist Guido Imbens — together with newly minted fellow Nobel laureates David Card of the University of California, Berkeley, and Joshua Angrist of MIT — have revolutionized empirical research in the economic sciences. Imbens, a pioneering econometrician, helped develop innovative mathematical and statistical methods that have vastly improved the ability of researchers to glean causal insights from both field and experimental data.
Their complementary contributions to the profession, dating back to the early 1990s, have given economists foundational tools to conduct “natural experiments” by using observational data to estimate the causal effect of a program or intervention. In fact, much of the rise in applied economics can be traced to methodological approaches that Imbens and Monday’s two other Nobel winners developed.
To this day, Imbens, a senior fellow at the Stanford Institute for Economic Policy Research (SIEPR), continues to expand on mathematical and statistical methods to help researchers interpret data.
“The work that Guido and his colleagues have done has helped to ignite nothing short of an empirical revolution in the social sciences,” said Jonathan Levin, dean of the Stanford Graduate School of Business. “It’s fueled by the methods they’ve developed and the ever-increasing amounts of data that we have access to. In particular, the methods Guido developed to infer cause and effect for natural experiments have been adopted by thousands of researchers across the social sciences.”
In an innovative study from 1994, Angrist and Imbens created a framework to show what conclusions about causation can be drawn from “natural experiments” — settings in real life in which people are randomly participating or not, or where there is a delineation, such as from program or policy changes.
Under this approach, researchers could then analyze data to tease apart the causal inferences when looking at the two groups, similar to comparisons drawn between randomized and controlled groups in a randomized trial.
Imbens and Angrist applied a two-step process to estimate causal effects. First they used “instrumental variables” — a source of variation that economists can use to mimic the threshold difference between two separate groups for comparison.
Second, when they evaluated the effects, Imbens and Angrist elaborately clarified the assumptions needed — developing what is known as the local average treatment effect (LATE). This model of establishing causal effect has helped boost the transparency and credibility of empirical research, and their Nobel Prize recognizes this foundational contribution.
"They’ve helped economists like me tease out causal relationships," said Mark Duggan, the Trione Director of SIEPR and an expert on the economics of health care and government entitlement programs. "If you don’t have good methods to do that, the estimates will be unreliable."