Psychological connectedness and intertemporal choice

Daniel M. Bartels, Lance J. Rips
2010 Journal of experimental psychology. General  
People tend to attach less value to a good if they know a delay will occur before they obtain it. For example, people value receiving $100 dollars tomorrow more than receiving $100 in 10 years. We explore one reason for this tendency (due to Derek Parfit, 1984): In terms of psychological properties, such as beliefs, values, and goals, the decision maker is more closely linked to the person (his or her future self) receiving $100 tomorrow than to the person receiving $100 in 10 years. For this
more » ... 0 years. For this reason, she prefers her nearer self to have the $100 than her more remote self. Studies 1 and 2 show that the greater the rated psychological connection between two parts of a participant's life, the less she discounts future monetary and nonmonetary benefits (good days at work) over that interval. In Studies 3-5, participants read about characters who undergo large lifechanging (and connectedness-weakening) events at different points in their lives, and then make decisions about the timing of benefits on behalf of these characters. All five studies reveal a relation between perceived psychological connectedness and intertemporal choice: Participants prefer benefits to occur before large changes in connectedness, but prefer costs to occur after these changes. Psychological Connectedness and Intertemporal Choice 3 Psychological Connectedness and Intertemporal Choice People often choose to consume a smaller amount of some good now, rather than a larger amount later. You might, for example, prefer to receive $150 dollars now over $500 in 25 years, or to pay more in shipping to have a DVD delivered sooner rather than later. Normative accounts of this tendency-called "temporal discounting"-provide the logic by which a rational actor should arrive at such a preference, whereas descriptive accounts try to explain how people actually choose. The current studies examine how well the prescriptions of a particularly innovative normative account describe how people choose when faced with intertemporal decisions. In brief, this account predicts (and we find) that when people anticipate psychological discontinuities between their current and future selves, they feel less concern for those future selves. They therefore prefer benefits to occur before the discontinuities and burdens to occur after them. Temporal Discounting You might think that a person should always choose the larger benefit of two or more options, regardless of timing, as long as the outcome occurs within the person's lifetime. All else equal, choosing the larger reward confers greater utility, serving to maximize your total lifetime utility. It seems reasonable for you to want your life, as a whole, to go as well as possible. But even if rationality demands impartiality towards all parts of your life, economists argue that temporal discounting is consistent with rationality. For example, if an option's value constantly increases over time (as does, e.g., money in a savings account), then at no point within your lifetime is consumption more profitable than postponement. Normative theories of discounting explain why (and at what points in time) the non-deferral of consumption is rationally Psychological Connectedness and Intertemporal Choice 4 justified, typically by showing how the time of consumption can affect total utility over a person's life. For example, you might rationally prefer to consume sooner, rather than later, because delays entail opportunity costs. Moreover, on average, people grow wealthier over time, so an increase in your standard of living will mean that the utility of $150 will decrease. Also, unpredictable events-changes in inflation rates, in tastes and preferences, uncertainty about realization of the benefit-all affect total utility. Some normative accounts specify the degree of discounting that one or more of these influences rationally require. For example, one theory (Fisher, 1930) argues that people should discount investable goods at a rate equal to the market interest rate. People's tendency to choose early options over later, larger options often exceeds what this family of models is able to justify (for a review, see Frederick, Loewenstein, & O'Donoghue, 2002) . In this light, people's choices are impatient. For example, market evidence suggests that people discount long-term costs (e.g., the higher energy costs associated with less expensive, lower-efficiency home air conditioners) by more than what the market interest rate predicts (Thaler & Shefrin, 1981) . Even psychological models, which typically describe rather than prescribe behavior, have adopted implicit normative standards. For example, situations in which people maximize short-term goals at the expense of long-term goals are often branded "failures" of selfcontrol (e.g., Ainslie, 1986; Baumeister, Vohs, & Tice, 2007; Metcalfe & Mischel, 1999; Strotz, 1956 ). Personal Identity and Intertemporal Choice Most accounts of rational action and rational choice argue that rationality demands acting in a manner consistent with self-interest: acting to achieve my goals, which include not only those realizable in the present, but also those my future self will obtain. Benefits to your future self can even compensate for burdens imposed on your present self; this is, of course, the basis of such Psychological Connectedness and Intertemporal Choice 5 practices as saving money, dieting, and studying. Reconceptualizing what constitutes a person or a lifetime, therefore, could motivate very different principles for behavior and choice, perhaps justifying as rational tendencies that seem impatient and irrational by the standards of most normative accounts. What are the future selves to which you should direct your interest? The philosophy of personal identity attempts to answer a set of interrelated questions about which properties and relations determine identity of selves over time. One traditional view of identity comports with the traditional view of temporal discounting. Consider a person P at a particular time point t a . If she survives to a later time t b , then there is some person P′ at t b who is numerically identical to P. (Numerical identity is the relation that each individual entity bears to itself, the relation symbolized by "="; in this case, P = P′. Numerical identity contrasts with "qualitative identity," the relation between things that have the same qualitative properties, like color or shape. Thus, two different 2009 Nissan Sentras can be qualitatively but not numerically identical.) Which properties of P and P′ determine this identity is a controversial matter: In order for P to be identical to P′, they may have to be related in certain psychological ways (e.g. Newman, 2006, for empirical studies of how people judge identity over time.) In either case, the relation of P to her future self P′ is one of identity. If P is trying to decide whether to consume some good now (at t a ) or to postpone consuming until t b , the issue is whether P's current utility for herself consuming now, u(P, t a ) is greater than her current utility for P′ consuming at t b , u(P′, t b ). However, if P = P′, this choice reduces to deciding whether her current utility for her own consuming at t a is greater than that for her own consuming at t b -whether u(P, t a ) is greater or less than u (P, t b ). This in turn depends only on how utility changes as a function of time. Psychological Connectedness and Intertemporal Choice 6 However, some contemporary views of personal identity call into question these assumptions about present and future selves and, thus, have direct implications for rational choice. One account that differs radically from standard economic views is offered by Derek Parfit (1984). He describes persons as a sequence of partially-overlapping selves and argues that the number, strength, and quality of psychological connections, which constitute the overlap between the present self and future selves, tend to decrease over time. What matters for choice is not the identity relation between persons (e.g., P = P′), but the strength of these psychological connections. Although Fred today is the same person as Fred 1 year from now and Fred 20 years from now (Fred 0 = Fred 1 = Fred 20 ), Fred today is likely to be more closely connected psychologically to his self in 1 year than to his self in 20 years. If so, Fred 0 should care more about Fred 1 than Fred 20 , and he should desire more positive and fewer negative outcomes for the former than the latter. As mentioned earlier, identity may itself depend on psychological connections; nonetheless, degree of connectedness can vary even within the identical person in ways that are important for choice. The practical import of Parfit's theory can be stated as an analogy: You are not rationally required to care as much about most others' welfare as your own. So, too, if your future self is sufficiently different in terms of personality and values from your current self, you are not rationally required to care as much about your future self's welfare. Thus, impatience can be justified insofar as you anticipate changes in your psychological connectedness over time. You ought to be more interested in having a closer future self consume a good than in having a more distant self consume the same good. Some evidence supporting this analogy between decisions for future selves and decisions for other people comes from a number of recent experiments (e.g.,
doi:10.1037/a0018062 pmid:20121312 fatcat:gtga7z6tejbsvelumkeyc2znr4