Irrational Economist_ Making Decisions in a Dangerous World - Erwann Michel-Kerjan [57]
How does time affect ambiguity preferences, especially with respect to the two sources of ambiguity—probabilities and outcomes? One framework that may be useful in answering this question is construal level theory, developed by two Israeli psychologists, Nira Liberman and Yaacov Trope. They propose that the mental representations of events in the near future are fundamentally different from those of events in the far future and that this difference can affect our choices. Specifically, we tend to focus on the desirability of an event when it is in the far future and to switch to feasibility concerns when it is imminent. For instance, when planning our summer vacation, we focus on the desirable aspects such as going to a nice location, seeing our friends, and so on. But when the date approaches, we start focusing on the feasibility dimension: reserving the hotel, packing our bags, and so on. In the context of risky choice, then, probabilities are subordinate to outcomes: The possible outcomes (i.e., desirability) are more salient in the long run, but as time passes the probability of an event happening (i.e., feasibility) becomes more important. This would imply the following.
1. People prefer to seek information about the desirability of future prospects (i.e., outcomes) before seeking information about the feasibility of attaining them (i.e., probabilities): If there is a possibility that I will receive some money in the future, I would initially like to know how much I may get, rather than what my chances of receiving it might be.
2. The probability dimension becomes more salient as time passes: As the date approaches, I would focus more on my chances of receiving the money.
Following these observations, we can conjecture that time has an impact not only on risk preferences but also on preferences for ambiguous events. Regarding events that are currently happening we are averse to uncertain probabilities and somewhat more tolerant of uncertain outcomes. If probabilities become less salient over time and outcomes become more salient, the negative influence of uncertain probabilities would decrease; that is, individuals would become less ambiguity-averse for future events. In short, in cases where both probabilities and outcomes are highly unknown—as with global warming—individuals would be more tolerant of missing information when the event takes place in the future. Postponing a vague event causes less anxiety than an immediate one.
My colleagues Dolchai La-ornual and Selcuk Onay and I recently conducted an experiment in a European business school. Preliminary results from our experiment supported our prediction. In some of the scenarios, we gave the participants full information about the risk involved, such as receiving $10,000 with a probability of 50 percent. In others, there was missing information about outcomes (receiving an amount between $5,000 and $15,000) and/or probabilities (with a probability of 25 percent to 75 percent). The results consistently showed that ambiguity attitudes change over time. Both for positive outcomes (e.g., winning the lottery, getting a tax rebate) and negative ones (e.g., paying a traffic ticket, losing in the stock market), ambiguity aversion decreases with time. More people choose the scenario “Winning between $5,000 and $15,000 with a chance of 25-75 percent” than the scenario “Winning $10,000 with a chance of 50 percent” when both options occur within three months from the day they are asked to make this choice. In short, we are more likely to take on unknown risks or costs when these pertain to the future than