Carnegie Mellon University

Understanding Decisions

April 18, 2017

Understanding Decisions: The Power of Combining Psychology and Economics

By Shilo Rea

Adolescents face many challenging decisions. So, do consumers.

A new paper published in the Proceedings of the National Academy of Sciences shows how collaborations between psychologists and economists lead to better understanding of such decisions than either discipline can on its own.

"Psychology and economics are both interested in how people make decisions, but have different theories and methods. In our work with economists at Northwestern, Michigan, the Federal Reserve and elsewhere, we have found ways to complement each other’s expertise," said Wändi Bruine de Bruin, professor of behavioral decision making at Leeds’ University Business School, who received her Ph.D. from Carnegie Mellon University, where she is collaborating professor in the Department of Engineering and Public Policy.

In two series of studies that focused on individuals’ expectations for major life events, Bruine de Bruin and CMU’s Baruch Fischhoff worked with economists to design survey questions that were simple enough for laypeople to answer but precise enough to inform economic models.

The first project examined adolescents’ expectations for life events that would affect their psychological and economic development, such as finding work, being arrested and having children. Their colleagues in economics, led by Charles Manski, a former CMU faculty member, wanted to ask precise questions on a major national survey but were meeting resistance from survey researchers, who claimed that they were too hard for teens to answer.

Bruine de Bruin and Fischhoff supported the economists’ concerns with studies showing that questions using seemingly simpler language were actually more difficult for respondents and less useful for researchers, because the simpler wording was more ambiguous. The team then developed questions that teens could understand and provide answers that economists could use. The process included asking for numerical probabilities (e.g., 70 percent), rather than verbal quantifiers, such as "very likely."

"We found that kids were better at judging their futures than people may have thought, that they could estimate with numerical probabilities just fine and their answers were generally sensible," said Fischhoff, the Howard Heinz University Professor in the Institute for Politics and Strategy and Department of Engineering and Public Policy at CMU.

The second project involved consumers’ expectations of inflation, which play a central role in predicting financial decisions for the overall economy. Economists at the U.S. Federal Reserve worried that the questions that they had used for decades did not mean the same thing to consumers as they did for economists.

Bringing psychological methods to bear on economics problems, the team found that here, too, it was better to ask more precise questions.

"When you ask people directly about 'inflation,' it led to less confusion, and more accurate expectations, than when you use vague terms, like 'prices in general,'" said Bruine de Bruin.

Moreover, asking about 'prices in general' led people to think of prices for specific goods, bringing more extreme prices to mind. As a result, expectations for ‘prices in general’ were higher than expectations for 'inflation.'

Bruine de Bruin and Fischhoff point to four conditions that made such transdiciplinary research possible: having a shared research goal, which neither discipline could achieve on its own; finding common ground in shared methodology; sharing effort throughout, with common language and sense of ownership; and gaining mutual benefit from both the research process and its products.

"Successful collaborations across fields can be done. You need to find willing partners, and create trusted partnerships," Fischhoff said. "For people interested in decision research, you should look for work that combines both psychology and economics because neither can provide the complete picture."

The research was supported by grants from the U.S. National Science Foundation, Swedish Foundation for the Humanities and Social Sciences and the European Union Seventh Framework Program.