Behavioural economist Yi-Tsung (Hugo) Hsieh investigates how people deal with risk. His study, Integrating Risk in the Timing of Outcomes examines how people make decisions when they don’t know exactly when or what the outcomes will be. To investigate, the PhD candidate at the Department of Economics conducted a novel experiment through the Toronto Experimental Economics Laboratory (TEEL). Referring to existing literature, Hsieh tested four different models of decision making that have been put forward to explain the how people integrate or evaluate elements of risk in their decisions.
“The whole study is motivated primarily by two other studies,” Hsieh explained. “One found that people are risk-averse to time uncertainty. In economics, we already know people are risk-averse to outcome uncertainty. For example, if we toss a fair coin and I give you $100 for heads and $0 for tails, or you can get $50 for sure, most people prefer the certain outcome. There’s another study showing that people are also risk-averse to timing uncertainty. For example, if I need to borrow $100 from you, and I tell you I can pay it back any time between one month and one year from now, or I can pay it back in six months exactly, most people prefer the precise timing even if they forego the chance of getting paid back earlier. A more precise example would be, if I purchase some products from Amazon, and I can choose to pay extra fee to ensure the fixed delivery date, or simply have it delivered at some random timing, some people would be willing to pay extra even if the product may still come earlier without paying extra fee. Current economic models cannot explain this phenomenon.”
To test these results, Hsieh designed his experiment using a lottery system with two stages: the first stage decides when the outcome will happen, and the second stage decides what the outcome will be. He then tracked which of the established models of decision making. One of those models, recursive rank-dependent utility theory or RRDU, holds that people evaluate outcome risk first and then timing risk before deciding. According to Hsieh’s results, almost two-thirds of people make decisions according to this model. Just under the remaining third of participants made their decisions according to the rank-dependent utility on discounted payment or RDUDP model in which decision-makers integrate risk in outcome and timing to evaluate the discounted values of outcomes. There are other models of decision-making Hsieh tested too, with fewer than 7% of participants drawing upon those methods.
The results of Hsieh’s study are especially helpful in understanding how people make long term decisions where the outcome won’t be known for some time.
“When we think about retirement, for example, most of us do not know exactly when we will retire or how much money we will have at that time,” Hsieh explained. “We face both time uncertainty (when we will retire) and outcome uncertainty (how much money we will have). Also, if you make investment, or buy bonds or stocks, you don’t know when you will realize a gain or loss (time uncertainty) and whether it will be a gain or loss (outcome uncertainty). Another application is job seeking. For example, when I look for a job, I don’t know how much pay I will get or when I will get the offer. These are the kinds of scenarios my research addresses.”
The experiment was the first Hsieh conducted himself and he was especially grateful for the support of the team and his colleagues at TEEL.
“The work we conduct at TEEL is deeply collaborative and cooperative,” he said. “I had classmates who came to help supervise my experiments to make sure nothing went wrong and would help them run their experiments.”
That collegiality extended to presentations of results when the colleagues were ready to share. While critical thinking and analysis are always Hsieh’s priority in scholarly discussion, he never loses track of that spirit of collaboration and cooperation in the field.
“All our opinions, whether against or for supporting each other, are totally about the research itself. Right is right, and wrong is wrong because this is research. There’s no room for confusion or lack of clarity. When I feel something is wrong in your research, or when I disagree with something you say, if I have evidence to support my idea, I must speak out,” Hsieh said. “It’s impersonal evaluation, but we all feel good when we are being treated in a friendly way. The whole culture and atmosphere of the Department of Economics at U of T made me feel that people are nice to each other. That taught me that I also need to make others feel good when they deal with me.”
As a result, Hsieh finds the opportunity to present as enjoyable as it is informative.
“I really enjoy discussions with people and cherish every opportunity to present in our seminars,” he said. “It’s very difficult to gather a group of intelligent individuals to spend their time, an hour, or 90 minutes, to listen to you carefully and understand your research. Every time I present in a seminar, I recognize that I am using others’ time, and everyone’s time is valuable.”
The experience of presenting and discussing results shares several features with how Hsieh teaches when he leads tutorials in behavioural economics and game theory.
“In both peer-led presentations and in tutorials, we try to communicate with people in a really precise and clear way,” Hsieh said. “When we teach, we try to explain complicated concepts, whether mathematical or economic intuition, in a general, plain language that is easy to understand. This is the same in research. When we present at seminars, we present something new to our audience, making things clear and organized, similar to designing course materials in a rational and organized way.”
That rational and organized way has been a feature of Hsieh’s time at the University of Toronto.
“Hugo wrote two solo-authored papers on scenarios where people face uncertainties both over the outcomes and the timing in which those risky outcomes are received, which he calls Compound Time Lotteries,” said Professor Yoram Halevy, Hsieh’s supervisor. “This adds a novel and important dimension to the environments in which choice under risk takes place. He studies these scenarios theoretically by proposing three alternative models to evaluate these objects and conducts a sophisticated experiment through which he can measure the empirical relevance of each model. In his work Hugo demonstrates creativity, independence, and perseverance.”
Return to the Department of Economics website.
Scroll more news.