HARVARD GAZETTE ARCHIVES

The Eric M. Mindich Conference on Experimental Social Science demonstrated how methods from psychology and economics can combine to achieve common research goals. Princeton University's Jeffrey Kling spoke at an afternoon session. (Staff photo Stephanie Mitchell/Harvard News Office)

Tools that travel

Previously known as the Center for Basic Research in the Social Sciences (CBRSS, pronounced seabreeze), the Institute for Quantitative Social Science was launched on March 1 to create and circulate statistical and analytical tools for the social and health sciences. Social scientists from across Harvard are often at work on projects that can't be readily accomplished through any single department, explains institute director Gary King, the David Florence Professor of Government. "You solve a problem in one field and the solution also works for unrelated problems in different fields," he explains. "Our goal is to support this type of research and provide infrastructure."

King, together with fellow government Professors Jim Alt and Ken Shepsle, co-founded CBRSS in 1998.

The center, housed in a comfortable, 18th century Victorian on Kirkland Street, went on to become an essential resource for quantitative social scientists. The upgrade to institute status reaffirms Harvard's commitment to their efforts. Along with the name change has come a shift in structure. The institute's organization is threefold: Scientific programs that unite investigators working in related areas; technology platforms that provide infrastracture to support research; and educational opportunities for Harvard students and postdocs.

Harvard economist Sendhil Mullainathan (Staff photo Rose Lincoln/Harvard News Office)

Choosing the cuter bank loan

At last week's conference, researchers in social psychology and economics were paired in joint presentations designed to show how research methods from the different fields can complement each other.

The first presentation, by Marianne Bertrand, professor of economics at the University of Chicago, and Eldar Shafir, professor of psychology and public affairs at Princeton University, was a dramatic demonstration of the intrusion of the emotions in supposedly rational decision making. Bertrand and Shafir are collaborating with Mullainathan on a study of how psychological factors influence an individual's decision to apply for a bank loan. The study population includes 60,000 residents of South Africa, who were each sent a loan solicitation through the mail. Loan offers varied by interest rate - the economic component - as well as by certain "psychological manipulations," such as a photo of an attractive man or woman on the offer letter.

Shafir said economic theorists assume decision makers are well informed and driven to maximize their self-interests. But behaviorists see things differently, he said. "Their view is that decision makers are driven by a number of factors of which they are not fully aware. The implications are that they may not [act in their self-interest]." Study results show that psychological factors do, in fact, play an important role in economic decision making. Bertrand said, "When you make the offer more psychologically attractive people seem less sensitive to interest rates."

Even in matters of health, rationality doesn't necessarily rule, as Rebecca Thornton, a Ph.D. candidate in economics at Harvard, demonstrated in her presentation. Thornton studies the demand for information about personal HIV status among rural villagers in Malawi, an African country where the infection rate nears 14 percent. Although 90 percent of survey respondents in Malawi say they want to know their status, only 9 percent actually have this information, Thornton said. This discrepancy, she explained, is the result of a web of psychological influences.

Forecasting feelings

Conference-developer Wilson joined colleague Dan Gilbert, Harvard professor of psychology, in a discussion of "affective forecasting," through which individuals predict their emotional response to a hypothetical event, such as a diagnosis of disease. Affective forecasting is important for health economics because it forms the basis for treatment decisions, Wilson said.

Usually, individuals overestimate the magnitude of their response, in a phenomenon called "impact bias." Actual responses are often milder than anticipated ones, he said, in part because individuals adapt to new conditions over time. He then described a model created with Gilbert that seeks to explain this process. Known as AREA, the model dissects the process into four phases: In the first, individuals "attend" to the event; in the second, they have an intense, emotional reaction to it; in the third, they seek to explain it; and in the fourth, they finally adapt. "Our goal is to speed up the process of adaptation to negative events," he explained. "This is a fascinating question: Can people be taught about adaptation in advance? This is important because critical decisions are based on affective forecasts."

As the conference illustrated, economic policies are ultimately about people - their fears, hopes, and motivations, that is, their psychology. By incorporating these human elements into economic research, scholars can achieve more comprehensive results and policy-makers can make sounder decisions. Through its many programs - including the separately funded Harvard-M.I.T. Data Center, the endowed Henry A. Murray Research Archive (formerly of Radcliffe), and the new University Initiative in Geospatial Analysis - the new institute will continue to foster the kind of collaboration that achieves these goals.