Edição Especial da Revista. Psiquiatria, Psicologia & Justiça

Edição Especial da Revista

Psiquiatria, Psicologia & Justiça

ii

SUMÁRIO

Intertemporal Choices and Deviant Behavior: Brief Note on Four Decades of

Studies

Escolhas Intertemporais e Comportamento Desviante: Breve Nota sobre Quatro

Décadas de Estudos

Diana Moreira/Fernando Almeida/Marta Pinto/Susana Barros/Fernando Barbosa ... 1

Gender-Based Violence: Relationships of power and domination reproduced in

the statements of women victims of violence

Violência de Género: relações de poder e dominação reproduzidas no discurso

da mulher vítima de violência

Priscylla Cassol/Jaci Augusta Neves de Souza/Gustavo Pachoal Teixeira de Castro ... 17

Involuntary commitment – 6-month follow up

Internamentos compulsivos – Follow up a 6 meses

Maria do Céu Ferreira/Catarina da Costa Campos/Teresa Sousa-Ferreira/Joana

Mesquita Machado/Beatriz Santos

... 35

Conjugal visitation in the context of female incarceration: a comparison of the

perceptions and meanings of inmates participating and not participating in the

program

A visita íntima no contexto de reclusão feminina: uma comparação das perceções

Rita Pinto/Alexandra Oliveira ... 43

Confinement without risk – HIV Prevention and AIDS

Reclusão Sem Risco – Prevenção do VIH e SIDA

Filomena Frazão de Aguiar/Alexandra Duque/Carla Oliveira/Duarte Barros/Joana

Martins/Helena Teixeira/Helena Vilaça/Laura Aguiar

... 59

Violence in the Context of Prostitution: Psychological Impact

Violência no Contexto da Prostituição: Impacto Psicológico

iv

Corpo Editorial

Editor-Chefe

Prof. Doutor Fernando Almeida

Comissão Científica

Prof. Doutor Adriano Vaz Serra

Prof. Doutor Agostinho Santos

Prof. Doutora Ana Sani

Prof. Doutora Ana Sofia Neves

Prof. Doutora Anita Santos

Prof. Doutor António Pacheco Palha

Prof. Doutora Carla Antunes

Prof. Doutor Carlos Mota Cardoso

Prof. Doutora Célia Ferreira

Mestre Diana Moreira (Doutoranda)

Prof. Doutor Duarte Nuno Vieira

Prof. Doutor Eurico Figueiredo

Prof. Doutor Fernando Almeida

Prof. Doutor Fernando Barbosa

Prof. Doutor Francisco Machado

Prof. Doutor Gualberto Buela-Casal

Prof. Doutora Helena Grangeia

Dr. Hernâni Vieira

Prof. Doutor João Marques-Teixeira

Prof. Doutor João Salgado

Prof. Doutor Jorge Costa Santos

Prof. Doutor Jorge Trindade

Prof. Doutor José Pinto da Costa

Prof. Doutora Laura Nunes

Prof. Doutor Luís Fernandes

Prof. Doutor Luís Gamito

Juiz Conselheiro Manuel Simas Santos

Prof. Doutora Maria José Carneiro de Sousa

Prof. Doutora Maria Luísa Figueira

Prof. Doutor Mário Simões

Prof. Doutora Marisalva Fávero

Prof. Doutora Marta Pinto

Prof. Doutora Olga Cruz

Prof. Doutora Paula Gomide

Prof. Doutor Pio Abreu

Prof. Doutora Rita Conde Dias

Prof. Doutora Sónia Caridade

Prof. Doutora Teresa Magalhães

Intertemporal Choices and Deviant Behavior: Brief Note on Four Decades

of Studies

Diana Moreira1,2, Fernando Almeida2, Marta Pinto2, Susana Barros1, & Fernando Barbosa1 1. University of Porto

2. ISMAI – Maia University Institute

Abstract Intertemporal choices (ICs) are choices that involve trade-off between costs and benefits that take place at different moments in time. The first aim of the current article is to present a conceptual framework on IC, namely the theoretical basis of discounting and discounting functions (discounted utility; quasi-hyperbolic, and hyperbolic discounting), concluding with a brief description of patterns of preference in ICs (delay effect; magnitude effect; sign effect; and the delay-speedup asymmetry). Secondly, we intend to present a summary on methodological options regarding IC research, namely differences concerning modalities of reward, delay intervals, and attempts to manipulate delay discounting. Finally, we emphasize the interest of IC research to the study of antisocial and deviant behavior.

Keywords: Intertemporal choices; Discounting; Reward; Delay intervals.

INTRODUCTION

Intertemporal choices (ICs) are choices that involve exchanges between costs and benefits that occur at different points in time (Johnson, 2012; Pimentel, Gonçalves, Scholten, Carvalho, & Correia, 2012). The ICs have been studied by experimental and behavioral economy for several decades, using delay-discounting paradigms. Experimental economy is intimately associated with behavioral economy and, despite having slightly different goals, they share many common points. These fields analyze the fundamental human choice of distribution of scarce resources in an ascending perspective, taking into account individual economic agents and trying to understand how important economic decisions are made. There is practical importance pertaining to ICs, and researchers have been applying delay-discounting protocols to the study of diverse deviant behaviors, namely when these behaviors are strongly associated with impulsivity disorders.

CONCEPTUAL FRAMEWORK Theoretical Basis of Discounting

A pioneer paper on the concept of temporal preference was presented by Koopmans (1960), despite research on the subject having already been performed in 1912 (Fishburn & Rubenstein, 1982). It was followed by many contributions that expanded the knowledge on this subject. The concept of discounting is substantiated on the presupposition of the behavior of economic agents and is represented by a number of standard economic axioms. These axioms frame individuals’ intertemporal preferences by introducing various formalized assumptions, which are at the center of any discounting model. As the behavioral axioms alter themselves, the discounting model does as well.

One of the crucial behavioral axioms defines impatience and procrastination. If a result is positive, then a smaller time interval is preferred over a larger one. If the result equals zero, the person is indifferent to the time period in which the null result occurs, while if the result is negative and, thus, something that causes displeasure, the longer period is preferred, therefore, implying procrastination (Farrugia, 2010).

Using the standard presuppositions on people’s behavior we can express their preferences in terms of utility functions. The utility functions reflect preferences instead of material goods, time, and uncertainty (Andersen, Harrison, Lau, & Rutstrom, 2008). Discounting rates are estimated with the equation of the utilities of two different payments received in different time periods. To be able to isolate the effect of time on preferences, the preferences instead of goods and uncertainty must be known, or assumed. In the majority of discounting experiments, monetary rewards are the only good in question. This simplifies the analysis by maintaining the constant “good” instead of the choices. However, there is a decision that still needs to be made regarding the uncertainty associated with the payments before the analyst is capable of isolating the discounting rate. The uncertainty in an experiment arises from the presence of delay in the reception of the payment. In most cases, the researchers assume risk neutrality and, thus, a linear relationship between the result and its utility (Farrugia, 2010, p.52).

When two results are presented, an earlier payment and a delayed payment, the discounting rate may be identified as the point where the utility of the present payment equals the utility of the second discounted payment. Assuming the neutrality of the risk and the exponential discounting, the individual discounting rate may be obtained by working with the present value of the monetary results instead of the utility (Andersen et al., 2008).

The effects of temporal discounting in subjective value are generally considered predictable and intuitive: as the delay until a reward increases, its subjective value decreases. The generality of this phenomenon – which operates on an ample variety of rewards and time scales – allowed the researchers to use a vast variety of methods, reward modalities, choice options, and analysis parameters in their experiments.

Discounting Functions. While the effects of delay on value may be normatively modeled using a single

exponential decay function, more complex functions are necessary to describe actual choice behavior (Fudenberg & Levine, 2006; Kalenscher & Pennartz, 2008; Loewenstein, O’Donoghue, & Rabin, 2003; O’Donoghue & Rabin, 2000; Ok & Masatlioglu, 2007). Next, we will present three proposed models for ICs - discounted utility, quasi-hyperbolic, and hyperbolic discounting - resorting to studies by Samuelson (1937), McClure et al. (2004), and Kable and Glimcher (2007). These studies differ in many aspects of their methods; for example, they differ in the amount of training the participants received. Even so, various methods can be applied to obtain individuals’ intertemporal preferences.

Model of Discounted Utility. Samuelson (1937) proposed the model of discounted utility. According to this

model, individuals assign a utility to a future consequence, integrating it in the basal level of consumption and they discount the utility of the consequence by a constant rate per unit of time (Strotz, 1956). Samuelson‘s model (1937) describes delay discounting (DD) using an exponential decay function, a simple solution that used common mathematics in calculating the interests at the time. However, many violations of this model are evident (Prelec & Loewenstein, 1991; Thaler, 1981), some of them noticed by Samuelson (1937) around the time of his original work. For example, individuals with deviant behaviors such as substance abuse or pathological gambling tend to exhibit more pronounced discounting, thus contrasting to the Samuelson model. Collectively, these violations imply a function with a much more accentuated discount throughout short time scales.

Quasi-hyperbolic Discounting Model. In a study on ICs by McClure, Laibson, Loewenstein, and Cohen (2004),

the participants chose between pairs of real monetary rewards that differed in their value (between a few tens of dollars) and delay before the delivery (in the period of a few weeks). To guide their analyses, the authors adopted a two-process framework that models discount using a quasi-hyperbolic function of subjective value (U), for a reward of amount A, a specific delay D, such as U = A βδD (Laibson, 1997; Loewenstein, 1996; Shefrin & Thaler, 1988). It is worth noting that this perspective corresponds to the widespread dual systems account of behavior control (Bechara, 2005; Ernst & Paulus, 2005; Kahneman, 2003; Loewenstein, 1996). Their analyses sought brain regions whose activation was consistent with each of the two model parameters. The first parameter, β, is also referred to as the “hot” or visceral system. It reflects a strong preference for immediate rewards when compared with any delay. Therefore, the β parameter, or “hot” system, seems to be prominent in deviant behaviors, as there is a strong preference for immediate rewards (Moreira, Pinto, Almeida, Barros, & Barbosa, 2016). The second parameter, δ, also referred to as the “cold” or rational system, produced a flat discount rate (and, as such, reduced the discriminability between reward values) in longer delays; in their analyses, the delta-related regions were defined based on the similar activation levels regardless of the intertemporal delay.

The results by McClure et al. (2004) have been highly influential, in the sense that they were interpreted to provide evidence that the impulsive choices reflect an over-activation of the reward system (compared to mechanisms of control). This increased activation of the reward system is also observed in substance abuse, pathological gambling, among other addictive behaviors, thus explaining the pattern of impulsive choices in these individuals (Grant & Chamberlain, 2014).The same group replicated the results in a 2007 study (McClure, Ericson, Laibson, Loewenstein, & Cohen, 2007) with liquid rewards (juice and water). In this more recent study, the participants were given the choice between a small amount of a liquid reward now, and a reward up to three times as large after a maximum of a 20-minute time delay. Using a beta/delta model, they observed the activation of different brain systems for immediate or delay rewards. These results were interpreted as strong neural evidence in favor of a two-component model of ICs.

Hyperbolic Discounting Model. Originally proposed by Loewestein and Prelec (1992), this model suggests that

people are, tendentiously, extremely impatient for consequences occurring in a near future, but are much more patient for consequences occurring in a more distant future (Benzion, Rapoport, & Yagil, 1989; Thaler, 1981). Therefore, in a generic manner, individuals tend to discount the future hyperbolically. The models that supported the idea of the existence of exponential discounting (e.g., Model of Discounted Utility) were gradually abandoned.

Moreover, different functions do not lead to drastically different predictions for the majority of the situations of ICs, reflecting the general similarity between quasi-hyperbolic and hyperbolic functions. Due to this imprecision, the results of various studies, would be consistent with any of the models. A potential strategy would be to focus the adjustment tests of the model to only specific ranges of values, where strong discrepancies are predicted. Secondly, processes such as an impulsive choice system, may have upstream or downstream influences on a subjective value representation. As summarized by Tesch and Sanfey (2008), a variety of behavioral effects might modulate the subjective value in the IC: the effect of reward magnitude - greater discounting for smaller rewards (Raineri & Rachlin, 1993; Thaler, 1981); an effect of number of choices - more choices increase the

discount (Read, 2001); an effect of order - increasing sequences are more valued than decreasing sequences (Loewenstein & Sicherman, 1991); a preference for accelerating instead of delaying the payment dates (Loewenstein, 1988); a preference for distributed rewards (Loewenstein & Prelec, 1993); an effect of type of reward - food is more intensely discounted than money (Odum & Rainaud, 2003); an effect of reward valence - gains are discounted in a more accentuated manner than losses (Thaler, 1981). Additionally, the preference for discount may be modified by distraction (Mischel, 1972). There are also cases in which the mental state of the participant preceding the choice models the discount (Berns, Laibson, & Loewenstein, 2007; Montague & King-Casas, 2007). Discounting rates increase with the deprivation of self-control (Baumeister & Heatherton, 1996), as is observed, for example, when smokers are deprived of nicotine (Field, Santarcangelo, Sumnall, Goudie, & Cole, 2006; Mitchell, 2004) and in addiction in general. Schweighofer et al. (2006) propose a more general conclusion where participants’ choices reflect the optimum discounting function for the task. Very recent data also presents difficulty for the simple conclusion that hyperbolic discounting behavior reflects the hyperbolic activation of the reward system.

Patterns of Preference in Intertemporal Choices

There are various patterns of preference in ICs. Taking the literature into consideration, it is pertinent to focus attention on four of these patterns, which were reported, for the first time, by Thaler (1981): (a) delay effect; (b) magnitude effect; (c) sign effect; and the (d) delay-speedup asymmetry. The aforementioned patterns patent observed variations in discounting rates, wherein a larger discounting rate refers to a greater impatience for gains and a greater procrastination for losses (Pimentel et al., 2012).

The delay effect refers that discounting rates tend to be larger for shorter waiting periods rather than longer waiting periods. To this respect, Thaler (1981) concludes that an individual can both prefer one apple today and two tomorrow, or, simultaneously, prefer two apples in 51 days’ time instead of one apple in 50 days’ time. This effect has been pointed out as a trigger of dynamically inconsistent, or impulsive behavior given that it is related to the individual’s decreasing impatience (Pimentel et al., 2012). In this sense, individuals with impulsive behavior or psychiatric conditions characterized by high impulsivity display greater aversion to delay, thus preferring choices that involve immediate gratification.

The magnitude effect refers that discounting rates tend to be larger for minor consequences than for major consequences (Chapman & Elstein, 1995; Holcomb & Nelson, 1992). People are sensitive to relative differences (contemplated by discounting rates), but also to absolute differences (Loewenstein & Prelec, 1992). To receive $100 now or $150 in a year involves a greater difference than receiving $10 now or $15 in a year, although the discounting rate is equal in both choices. For the larger difference between the consequences, people tend to be more patient, e.g., they tend to prefer the larger consequence later on (Pimentel et al., 2012). However, this pattern of choice may not be observed in highly impulsive individuals, since they tend to prefer more immediate consequences, even if it entails a smaller gain.

The sign effect tells us that people’s behavior is qualitatively different for frameworks of gains and losses. Using the example mentioned in the previous paragraph, for example, a person that prefers to receive $100 now instead of receiving $150 in a year (revealing himself impatient) may prefer to pay $100 now instead of paying $150 in a year (not revealing himself as a procrastinator). The phenomenon of aversion to losses, which is known to be altered in psychopathy (De Martino, Camerer, & Adolphs, 2010; Shiv, Loewenstein, Bechara, Damasio, & Damasio, 2005), is characterized by an individual’s behavior directed to avoid reductions of current well-being, for the displeasure associated with a loss weighs more than the pleasure associated to the gain (Thaler & Benartzi, 1995). Therefore, it is possible to conclude that individuals will be more patient in a framework of losses and more impatient in a framework of situations of gain (Pimentel et al., 2012).

The sign effect interacts frequently with the magnitude effect, in the sense that the sign effect is larger for smaller consequences than for larger consequences (Loewenstein & Prelec, 1992).

The delay-speedup asymmetry (Loewenstein, 1988) consists in the asymmetric preference between advancing and delaying consumption. Loewenstein (1988) observed that the necessary amount for people to delay a reception in a certain time interval was considerably larger than the amount that people were willing to sacrifice in order to advance consumption in that interval. Therefore, when an individual postpones a reception of $100 per year, he may also demand to receive at least $150, but when that same individual postpones the reception of $150 he may demand to receive more than $100 today. Both pairs of choices are different representations of the same pair of underlying options, thus, the results constitute a classic framework effect, which then is inconsistent with any normative theory, like the discounted utility model (Samuelson, 1937).

METHODOLOGICAL FRAMEWORK

There are several methodological differences in studies of DD, but such differences refer mainly to reward parameters, delay intervals, and manipulations of DD.

Modalities of Reward. There has been a notable variety in the reward parameters used in studies on ICs. Some

of this diversity is guided by differences in parameters of delay. In order to obtain variability in ICs, it is necessary to structure the magnitudes of delays and rewards so that participants may sometimes choose the delayed option and other times the immediate option. Therefore, longer delays are often correlated with greater

rewards (Pimentel et al., 2012). However, the choice of reward is also driven by payment method. For reference, almost all studies in behavioral economics analyze choices for monetary rewards, most of which involve hypothetical rewards (Frederick, Loewenstein, & O’Donoghue, 2002). In the study by Carter, Meyer, and Huettel (2010), the rewards offered included direct monetary payments (cumulative or random), gift certificates or debit cards, hypothetical monetary payments, or juice/water. The range of symbolic reward magnitude across studies is very ample, from a minimum of $.01 (Ersner-Hershfield, Wimmer, & Knutson, 2008) to a maximum of $40000 (Shamosh et al., 2008).

There are examples of previous behavioral research that provide evidence that hypothetical rewards may be treated differently from real rewards, especially with short delays (Camerer & Hogarth, 1999; Smith & Walker, 1993), although there are examples where no difference was found (Baker, Johnson, & Bickel, 2003; Johnson & Bickel, 2002; Lagorio & Madden, 2005; Madden, Begotka, Raiff, & Kastern, 2003; Madden et al., 2004). Studies also refer that random single-trial payments evoke different discount rates than those with cumulative methods of payment (Kagel & Roth, 1997). Most of the neuroeconomic studies on risky choices, for comparison, use a standard system of payment: the choices involve real monetary rewards that range up to a few tens of dollars (or the equivalent in another currency) and pay the participants for a subset of their choices, selected randomly (Clithero, Carter, & Huettel, 2009; Hsu, Bhatt, Adolphs, Tranel, & Camerer, 2005; Huettel, Stowe, Gordon, Warner, & Platt, 2006).

Delay Intervals. There has been little consistency in temporal intervals throughout studies. Most studies use

delay intervals ranging from weeks to years, and the remaining use much smaller intervals ranging from seconds to minutes, for rewards obtained during the experiment and for symbolic rewards that are accumulated. The former intervals are generally compatible with those used in research on behavioral economics (Hayden & Platt, 2007; Kalenscher et al., 2005; Kalenscher et al., 2004; Rosati, Stevens, & Hauser, 2006; Stevens, Hallinan, & Hauser, 2005; Stevens, Rosati, Ross, & Hauser, 2005). A study by Wittmann et al. (2007) compared delays (for hypothetical rewards) of less than one year with delays of more than one year, and found that areas of the brain involved in the evaluation of rewards differentiated between these two categories. In addition, the use of real but not immediate rewards (e.g., those that are delivered after participants leave the laboratory session) risks incurring transaction costs that may influence the subjective value, which are not present in research with animals. It should be noted that McClure et al. (2007) used both monetary rewards with a delay of weeks (McClure et al., 2004), and juice rewards with delay of minutes (McClure et al., 2007) and presented relatively similar action patterns across both studies.

Attempts to Manipulate Delay Discounting Although there have been many attempts to manipulate DD using

pharmacological interventions, virtually no study succeeded in such manipulation. An exception to this is the study by de Wit, Enggasser, and Richards (2002), in which aversion to delay was reduced after the administration of amphetamines. Similarly, few studies tried to manipulate cognitive functioning to influence DD. In a study that used a “within subjects” design, Hinson, Jameson, and Whitney (2003) observed that increased working memory load was associated with more accentuated discounting functions. Moreover, task context and presentation per se may affect the DD rates. For example, Loewenstein and Prelec (1993) found that when decisions were presented as a series of results instead of individual results, people preferred results that bettered with time. This corresponds to a real inversion of conventional discounting of delayed results. Studies also indicated that people presented lower discounting rates for decisions that involved large and important results (Chapman, 2002) or long delays (Roelofsma & van der Pligt, 2001). Based on this research, Ortendahl and Fries (2005) recommend framing health messages in terms of sequences of results, with large and important results or long delays, which should induce lower implicit discounting rates.

APPLICATIONS TO ANTISOCIAL BEHAVIOR

Every day we must decide between options that have immediate or delayed consequences. The process of deciding whether to opt for an immediate reward earlier or for a delayed but higher reward is strongly related to academic, professional or social success in life. Adaptive behavior implies, voluntarily, postpone impulsive urges of immediate gratification and instead persist in goal-oriented actions to achieve positive results in the future.

Studies on ICs suggest that people often avoid risk when they have to choose between options associated with probable versus guaranteed results (Grace, 1999), and the time interval between choice and realization of the gain/loss is a paramount factor in the decision, wherein people prefer to receive gains earlier and losses later on (Pimentel et al., 2012). This tendency seems to be more marked in people showing certain antisocial personalities or deviant behaviors (De Martino et al., 2010). Literature shows that antisocial behavior is characterized by patterns of disinhibition and impulsivity, and these characteristics may affect the way these individuals make decisions and organize them temporally.

The patterns of preference in ICs observed variations in discounting rates, wherein a larger discounting rate refers to a greater impatience for gains and a greater procrastination for losses (Pimentel et al., 2012). This tendency is observed, for example, in psychopaths (De Martino et al., 2010). These individuals have difficulties in behavior inhibition, even in situations where the probability of punishment following their actions is high,

especially if there is a competing reward (Carver & White, 1994). It can be assumed that they wait less time for delayed rewards, even if the value of these rewards increase through this delay, or they prefer to delay the losses, even if they worsen in the future. They evidenced a pattern of preference more pronounced than the normative pattern for a smaller immediate reward or larger future loss than normative pattern (Patrick, 2007).

Individuals differ in the strength with which they retain the subjective value as a function of time. In addition, the extent to which people are able to delay gratification is generally associated with the notion of willpower, whereas those who choose immediate gratification rather than important long-term interests are judged as being impulsive. However, impulsivity is determined by a complex set of processes and comprises multiple components and this observation is consistent with the conclusion that the correlations between different impulsivity assessment methods are modest. Nevertheless, impulsivity can be conceptualized as a pattern of behavior for which the potential negative consequences have limited influence in action planning. The external validity for using DD paradigms to quantify impulse control disorders is based on the conclusions that children with ADHD, smokers, alcoholics, and individuals addicted to substances exhibit an increasing discount of delayed rewards.

Despite the well-established conclusion that impulsive individuals discount delayed rewards more strongly than individuals with more self-control, the underlying cognitive and affective processes that explain this phenomenon are still poorly understood.

CONCLUSIONS

Studies on DD have been widely used in experimental economics, but have also interesting applications in psychology, for example in order to analyze how individuals showing deviant behaviors organize their intertemporal choices. In fact, several studies have been conducted to their purpose, in some cases with conflicting results. The variability in models, payment methods, and delay intervals used in studies of ICs may contribute to the differences in their results. However, it is important to mention that the use of a standardized set of conditions also poses significant limitations. For example, when studying the impulsivity of the choice, the promise of an abstract reward later on may not elicit the desired behavior. The use of substantive rewards large enough to elicit compensation for longer delays may also raise ethical concerns. Nonetheless, the consistency in methods may often improve links to other research without compromising the purpose of the studies (Carter et al., 2010).

The anomalies in human behavior of ICs suggest potential constraints for different models. We described three models: discounted utility, quasi-hyperbolic, and hyperbolic discounting.

The current focus of attention lies on the descriptive models of hyperbolic discounting that incorporate the decrease of discounting rates with the progression of time (Ainslie, 1975; Loewenstein & Prelec, 1992). In this sense, the notion of decreasing impatience has become more common, because it constitutes an explanatory alternative for the individual’s behavior in temporal decision-making (Thaler & Benartzi, 2004).

Therefore, temporal preferences are driven by the discounted values of the options. These discounted values are defined by assigning values to changes in face of the basal consumption level and by the hyperbolic discounting of these values according to the delay of the consequences, the hyperbolic discounting being the decrease in value rates according to the delay of the consequences (Pimentel et al., 2012).

In the determination of discounting functions, the majority of studies vary the magnitude of both the temporal distance and the immediate reward, while the remaining vary the magnitude of only one or neither of the rewards. The practice of varying only the delay reward (fixed immediate reward) elicits a larger discounting by the participants than only varying the immediate reward (fixed delay reward; Tesch & Sanfey, 2008).

Studies that use hypothetical payment methods have a median maximum reward of $312 while studies with direct payment (in cash or gift certificate, normally in some randomly chosen trials) have a median maximum reward of $32.50. These studies that resort to a direct cumulative method of payment have the lowest median maximum reward ($.95).

Despite several attempts involving the use of pharmacological agents, there were essentially no studies that successfully displayed direct manipulations of DD.

ACKNOWLEDGEMENTS

This work was supported by the Fundação para a Ciência e Tecnologia (reference grant SFRH/BD/108216/2015). The authors declare that no competing interests exist.

CORRESPONDING AUTHOR

Diana Moreira, Laboratory of Neuropsychophysiology, Faculty of Psychology and Educational Sciences, University of Porto, Rua Alfredo Allen, 4200-135 Porto, dianapmoreira@gmail.com

REFERENCES

Ainslie, G. (1975). Specious reward – behavioral theory of impulsiveness and impulse control. Psychology Bulletin, 82, 463-496.

Andersen, S., Harrison, G., Lau, M., & Rutstrom, E. (2008). Eliciting Risk and Time Preferences. Econometrica, 76, 583-618.

Baker, F., Johnson, M., & Bickel, W. (2003). Delay discounting in current and never-before cigarette smokers: Similarities and differences across commodity, sign, and magnitude. Journal of Abnormal Psychology, 112, 382-392. Ballard, K., & Knutson, B. (2009). Dissociable neural representations of future reward magnitude and delay during temporal

discounting. NeuroImage, 45, 143-150.

Baumeister, R., & Heatherton, T. (1996). Self- regulation failure: An overview. Psychological Inquiry, 7, 1-15.

Bechara, A. (2005). Decision making, impulse control and loss of willpower to resist drugs: A neurocognitive perspective.

Nature Neuroscience, 8, 1458-1463.

Benzion, U., Rapoport, A., & Yagil, J. (1989). Dis- count rates inferred from decisions: An experimental study. Management

Science, 35, 270-284.

Berns, G. S., Laibson, D., & Loewenstein, G. (2007). Intertemporal choice: Toward an integrative framework. Trends in

Cognitive Sciences, 11, 482-488.

Camerer, C., & Hogarth, R. (1999). The effects of financial incentives in experiments: A review and capital-labor-production framework. Journal of Risk and Uncertainty, 19, 7-42.

Carter, R., Meyer, J., & Huettel, S. (2010). Functional Neuroimaging of Intertemporal Choice Models: A Review. Journal of

Neuroscience Psychology, and Economics, 3(1), 27-45.

Carver, C. & White, T. (1994). Behavioral inhibition activation, and affective responses to impending reward and punishment: The BIS/BAS Scales. Journal of Personality and Social Psychology, 67, 319-333.

Chapman, G. (2002). Your money or your health: Time preferences and trading money for health. Medical Decision Making,

22, 410-416.

Chapman, G., & Elstein, A. (1995). Valuing the future: Discounting health and money. Medical Decision Making, 15, 373-386.

Clithero, J., Carter, R., & Huettel, S. (2009). Local pattern classification differentiates processes of economic valuation.

NeuroImage, 45, 1329-1338.

De Martino, B., Camerer, C., & Adolphs, R. (2010). Amygdala damage eliminates monetary loss aversion. Proceedings of

the National Academy of Sciences, 107(8), 3788-3792. doi:10.1073/pnas.0910230107

de Wit, H., Enggasser, J., & Richards, J. (2002). Acute administration of d-amphetamine decreases impulsivity in healthy volunteers. Neuropsychopharmacology, 27, 813-825.

Ernst, M., & Paulus, M. (2005). Neurobiology of decision-making: A selective review from a neurocognitive and clinical perspective. Biological Psychiatry, 58, 597-604.

Ersner-Hershfield, H., Wimmer, G., & Knutson, B. (2008). Saving for the future self: Neural measures of future self-continuity predict temporal dis- counting. Social Cognitive and Affective Neuroscience, 4, 85-92.

Farrugia, J. (2010). Heterogeneities In Time Discounting Preferences Of Environmental Projects: Some Experimental Results. Bank of Valletta Review, 41, 47-68.

Field, M., Santarcangelo, M., Sumnall, H., Goudie, A., & Cole, J. (2006). Delay discounting and the behavioural economics of cigarette purchases in smokers: The effects of nicotine deprivation. Psychopharmacology, 186, 255-263. Fishburn, P., & Rubenstein, A. (1982). Time Preference. International Economic Review, 23, 677-694.

Frederick, S., Loewenstein, G., & O’Donoghue, T. (2002). Time discounting and time preference: A critical review. Journal

of Economic Literature, 40, 351-401.

Fudenberg, D., & Levine, D. (2006). A dual-self model of impulse control. American Economic Review, 96, 1449-1476. Grace, R. (1999). The matching law and amount-dependent functions as account for choice in human and pigeon. Journal of

the Experimental Analysis of Behavior, 70, 1-19.

Grant, J., & Chamberlain, S. (2014). Impulsive action and impulsive choice across substance and behavioral addictions: cause or consequence? Addictive Behaviors, 39(11), 1632-1639. doi:10.1016/j.addbeh.2014.04.022

Green, L., & Myerson, J. (2004). A discounting framework for choice with delayed and probabilistic rewards. Psychological

Bulletin, 130, 769-792.

Hayden, B., & Platt, M. (2007). Temporal discounting predicts risk sensitivity in rhesus macaques. Current Biology, 17, 49-53.

Hinson, J., Jameson, T., & Whitney, P. (2003). Impulsive decision-making and working memory. Journal of Experimental

Psychology: Learning, Memory, & Cognition, 29, 298-306.

Holcomb, J., & Nelson, P. (1992). Another experimental look at individual time preference. Rationality and Society, 4(2), 199-220.

Hsu, M., Bhatt, M., Adolphs, R., Tranel, D., & Camerer, C. (2005). Neural systems responding to degrees of uncertainty in human decision-making. Science, 310, 1680-1683.

Huettel, S., Stowe, C., Gordon, E., Warner, B., & Platt, M. (2006). Neural signatures of economic preferences for risk and ambiguity. Neuron, 49, 765-775.

Johnson, M. (2012). An efficient operant choice procedure for assessing delay discounting in humans – initial validation in cocaine-dependent and control individuals. Experimental and Clinical Psychopharmacology, 20(3), 191-204. doi:10.1037/a0027088

Johnson, M., & Bickel, W. (2002). Within- subject comparison of real and hypothetical money rewards in delay discounting.

Kable, J., & Glimcher, P. (2007). The neural correlates of subjective value during intertemporal choice. Nature Neuroscience,

10, 1625-1633.

Kagel, J., & Roth, A. (1997). The Handbook of Experimental Economics. Princeton: Princeton University Press.

Kahneman, D. (2003). Maps of bounded rationality: Psychology for behavioral economics. American Economic Review, 93, 1449-1475.

Kalenscher, T., & Pennartz, C. (2008). Is a bird in the hand worth two in the future? The neuro-economics of intertemporal decision-making. Progress in Neurobiology, 84, 284-315.

Kalenscher, T., Windmann, S., Diekamp, B., Rose, J., Gunturkun, O., & Colombo, M. (2005). Single units in the pigeon brain integrate reward amount and time-to-reward in an impulsive choice task. Current Biology, 15, 594-602. Kalenscher, T., Windmann, S., Gunturkun, O., Rose, J., Diekamp, B., & Colombo, M. (2004). Single neurons code delay-

and amount-dependent subjective reward value and relative reward preference during impulsive decision-making.

Journal of Psychophysiology, 18, 202-202.

Koopmans, T. (1960). Stationary Ordinal Utility and Impatience. Econometrica, 28, 287-309.

Lagorio, C., & Madden, G. (2005). Delay dis- counting of real and hypothetical rewards III: Steady-state assessments, forced-choice trials, and all real rewards. Behavioural Processes, 69, 173-187.

Laibson, D. (1997). Golden eggs and hyperbolic discounting. Quarterly Journal of Economics, 112, 443-477. Loewenstein, G. (1988). Frames of mind in intertemporal choice. Management Science, 34, 200-214.

Loewenstein, G. (1996). Out of control: Visceral influences on behavior. Organizational Behavior and Human Decision

Processes, 65, 272-292.

Loewenstein, G., & Prelec, D. (1992). Anomalies in intertemporal choice: Evidence and an interpretation. Quarterly Journal

of Economics, 107, 573-597.

Loewenstein, G., & Prelec, D. (1993). Preferences for sequences of outcomes. Psychological Review, 100, 91-108.

Loewenstein, G., & Sicherman, N. (1991). Do workers prefer increasing wage profiles. Journal of Labor Economics, 9, 67-84.

Loewenstein, G., O’Donoghue, T., & Rabin, M. (2003). Projection bias in predicting future utility. Quarterly Journal of

Economics, 118, 1209-1248.

Madden, G., Begotka, A., Raiff, B., & Kastern, L. (2003). Delay discounting of real and hypothetical rewards. Experimental

and Clinical Psychopharmacology, 11, 139-145.

Madden, G., Raiff, B., Lagorio, C., Begotka, A., Mueller, A., Hehli, D., … Begotka, A. (2004). Delay discounting of potentially real and hypothetical rewards: II. Between- and within- subject comparisons. Experimental and

Clinical Psychopharmacology, 12, 251-261.

McClure, S., Ericson, K., Laibson, D., Loewenstein, G., & Cohen, J. (2007). Time discounting for primary rewards. Journal

of Neuroscience, 27, 5796-5804.

McClure, S., Laibson, D., Loewenstein, G., & Cohen, J. (2004). Separate neural systems value immediate and delayed monetary rewards. Science, 306, 503-507.

Mischel, T. (1972). Explanation in Behavioural Sciences – Borger, R. and Cioffi, F. Contemporary Psychology, 17, 51-52. Mitchell, S. (2004). Effects of short-term nicotine deprivation on decision-making: Delay, uncertainty and effort discounting.

Nicotine & Tobacco Research, 6, 819-828.

Montague, P., & King-Casas, B. (2007). Efficient statistics, common currencies and the problem of reward harvesting.

Trends in Cognitive Sciences, 11, 514-519.

Moreira, D., Pinto, M., Almeida, F., Barros, S., & Barbosa, F. (2016). Neurobiological bases of intertemporal choices: A

comprehensive review. Aggression and Violent Behavior, 26, 1-8.

doi:http://dx.doi.org/10.1016/j.avb.2015.11.003O’Donoghue, T., & Rabin, M. (2000). The economics of immediate gratification. Journal of Behavioral Decision Making, 13, 233-250.

Odum, A., & Rainaud, C. (2003). Discounting of delayed hypothetical money, alcohol, and food. Behavioural Processes, 64, 305-313.

Ok, E., & Masatlioglu, Y. (2007). A theory of (relative) discounting. Journal of Economic Theory, 137, 214-245.

Ortendahl, M., & Fries, J. (2005). Framing health messages based on anomalies in time preference. Medical Science Monitor:

International Medical Journal of Experimental and Clinical Research, 11, 253-256.

Patrick, C. (2007). Antisocial personality disorder and psychopathy. In W. O’Donohue, K. A. Fowler, & S. O. Lilienfeld (Eds.), Handbook of personality disorders (pp. 109-166). New York: Sage.

Pimentel, D., Gonçalves, G., Scholten, M., Carvalho, P., & Correia, M. (2012). Efeitos contextuais na escolha intertemporal: evidência contra modelos de desconto. Análise Psicológica, xxx(3), 269-284.

Prelec, D., & Loewenstein, G. (1991). Decision- making over time and under uncertainty: A common approach. Management

Science, 37, 770-786.

Raineri, A., & Rachlin, H. (1993). The effect of temporal constraints on the value of money and other commodities. Journal

of Behavioral Decision Making, 6, 77-94.

Read, D. (2001). Is time-discounting hyperbolic or subadditive? Journal of Risk and Uncertainty, 23, 5-32.

Roelofsma, P., & van der Pligt, J. (2001). On the psychology of time preference and human decisions: Introduction to the special issue. Acta Psychologica, 108, 91-93.

Rosati, A., Stevens, J., & Hauser, M. (2006). The effect of handling time on temporal discounting in two New World primates. Animal Behaviour, 71, 1379-1387.

Samuelson, P. (1937). A note on measurement of utility. Review of Economic Studies, 4, 155-161.

Schweighofer, N., Shishida, K., Han, C. E., Okamoto, Y., Tanaka, S. C., Yamawaki, S., … Doya, K. (2006). Humans can adopt optimal discounting strategy under real-time constraints. Plos Computational Biology, 2, 1349-1356.

Shamosh, N., DeYoung, C., Green, A., Reis, D., Johnson, M., Conway, A., … Gray, J. (2008). Individual differences in delay dis- counting: Relation to intelligence, working memory, and anterior prefrontal cortex. Psychological Science,

19, 904-911.

Shefrin, H., & Thaler, R. (1988). The behavioral life-cycle hypothesis. Economic Inquiry, 26, 609-643.

Shiv, B., Loewenstein, G., Bechara, A., Damasio, H., & Damasio, A. (2005). Investment behavior and the negative side of emotion. Psychological Science, 16(6), 435-439.

Smith, V., & Walker, J. (1993). Monetary rewards and decision cost in experimental economics. Economic Inquiry, 31, 245-261.

Stevens, J., Hallinan, E., & Hauser, M. (2005). The ecology and evolution of patience in two New World monkeys. Biology

Letters, 1, 223-226.

Stevens, J., Rosati, A., Ross, K., & Hauser, M. (2005). Will travel for food: Spatial dis- counting in two new world monkeys.

Current Biology, 15, 1855-1860.

Strotz, R. (1956). Myopia and inconsistency in dynamic utility maximization. Review of Economic Studies, 23, 165-180. doi:http://dx.doi.org/10.2307/2295722

Tesch, A., & Sanfey, A. (2008). Models and methods in delay discounting. Annals of the New York Academy of Sciences,

1128, 90-94.

Thaler, R. (1981). Some empirical evidence on dynamic inconsistency. Economics Letters, 8, 201-207.

Thaler, R., & Benartzi S. (2004). Save more tomorrow: Using behavioral economics to increase employee saving. Journal of

Political Economy, 112, 164-187.

Thaler, R., & Benartzi, S. (1995). Myopic loss aversion and the equity premium puzzle. The Quarterly Journal of Economics,

11, 73-92.

Wittmann, M., Leland, D., & Paulus, M. (2007). Time and decision-making: Differential contribution of the posterior insular cortex and the striatum during a delay-discounting task. Experimentelle Hirnforschung Expérimentation

Escolhas Intertemporais e Comportamento Desviante: Breve Nota sobre

Quatro Décadas de Estudos

Diana Moreira1,2, Fernando Almeida2, Marta Pinto2, Susana Barros1, & Fernando Barbosa1 1. Universidade do Porto

2. ISMAI – Instituto Universitário da Maia

Resumo Escolhas intertemporais (EIs) são escolhas que envolvem trocas entre custos e benefícios que ocorrem em diferentes momentos no tempo. O primeiro objetivo deste artigo é apresentar um enquadramento conceptual das EIs, nomeadamente a base teórica do desconto e as funções de desconto (utilidade descontada, desconto quase-hiperbólico e desconto hiperbólico), concluindo com uma breve descrição dos padrões de preferência nas EIs (o efeito de diferimento, o efeito de magnitude, o efeito de sinal, e a assimetria atraso-adiantamento). Em segundo lugar, pretendemos apresentar um resumo das opções metodológicas na investigação sobre EIs, em particular, as diferenças relativamente às modalidades de recompensa, intervalos de atraso, e tentativas de manipular o desconto diferido. Finalmente, salientamos a relevância da investigação sobre EIs para o estudo do comportamento antissocial e desviante.

Palavras-chave: Escolhas intertemporais; Desconto; Recompensa; Intervalos de atraso.

INTRODUÇÃO

Escolhas intertemporais (EIs) são escolhas que envolvem trocas entre custos e benefícios que ocorrem em diferentes momentos no tempo (Johnson, 2012; Pimentel, Gonçalves, Scholten, Carvalho, & Correia, 2012). As EIs têm sido estudadas pela economia experimental e comportamental durante décadas, usando paradigmas de desconto diferido (DD). A economia experimental está intimamente associada à economia comportamental e, apesar de terem objetivos ligeiramente diferentes, partilham vários pontos comuns. Estas áreas analisam a escolha humana fundamental de distribuição de recursos escassos numa perspetiva ascendente, tendo em conta os agentes económicos individuais e tentando perceber como são tomadas as decisões económicas importantes. Existe uma importância prática relativamente às escolhas intertemporais, e investigadores têm vindo a aplicar protocolos de DD ao estudo de diversos comportamentos desviantes, em particular quando estes comportamentos estão fortemente associados com distúrbios de impulsividade.

ENQUADRAMENTO CONCEPTUAL Base Teórica do Desconto

Um trabalho pioneiro sobre o conceito de preferência temporal foi apresentado por Koopmans (1960), apesar de a investigação sobre o assunto ter sido realizada já em 1912 (Fishburn & Rubenstein, 1982). Seguiram-se várias contribuições que expandiram o conhecimento sobre este tema. O conceito de desconto está fundamentado no pressuposto do comportamento de agentes económicos e é representado por um número de axiomas económicos-padrão. Estes axiomas enquadram as preferências intertemporais dos indivíduos ao introduzir vários pressupostos formalizados, que estão no centro de qualquer modelo de desconto. À medida que os pressupostos comportamentais se alteram o modelo de desconto também se altera.

Um dos axiomas comportamentais cruciais define a impaciência e a procrastinação. Se um resultado é positivo, então um intervalo de tempo mais curto é preferido a um mais longo. Se o resultado é igual a zero, a pessoa é indiferente ao período de tempo no qual o resultado zero ocorre, enquanto se o resultado for negativo e, portanto, algo que confere desprazer, é preferido o período mais longo implicando, assim, procrastinação (Farrugia, 2010). Utilizando os pressupostos padrão no comportamento dos indivíduos, podemos exprimir as suas preferências em termos de funções de utilidade. As funções de utilidade refletem as preferências em vez dos bens, do tempo, e da incerteza (Andersen, Harrison, Lau, & Rutstrom, 2008 As taxas de desconto são estimadas com a equação das utilidades de dois pagamentos diferentes recebidos em diferentes períodos de tempo. Para ser capaz de isolar o efeito do tempo nas preferências, as preferências em vez dos bens e da incerteza têm de ser conhecidas, ou assumidas. Na maior parte das experiências de desconto, as recompensas monetárias são o único bem em questão. Isto simplifica a análise ao manter a constante “bem” em vez das escolhas. No entanto, há uma decisão que ainda precisa de ser feita sobre a incerteza associada com os pagamentos antes de o analisador ser capaz de isolar a taxa de desconto. A incerteza numa experiência surge da presença do atraso na receção do pagamento

.

Quando são apresentados dois resultados, um pagamento mais próximo e um pagamento atrasado, a taxa de desconto pode ser identificada como o ponto onde a utilidade do pagamento presente é igual à utilidade do segundo pagamento descontado.Assumindo a neutralidade do risco e o desconto exponencial, a taxa de desconto

individual pode ser obtida ao trabalhar com o valor presente dos resultados monetários em vez da utilidade (Andersen et al., 2008).

Os efeitos de descontos temporais em um valor subjetivo são, geralmente, considerados previsíveis e intuitivos: à medida que o atraso até à recompensa aumenta, o seu valor subjetivo diminui. A generalidade deste fenómeno – que funciona através de uma ampla variedade de recompensas e de escalas de tempo – permitiu aos investigadores utilizar uma ampla variedade de métodos, modalidades de recompensa, opções de escolha, e parâmetros de análise nas suas experiências.

Funções de Desconto. Enquanto os efeitos de atraso sobre o valor podem ser modelados normativamente

utilizando uma única função de decaimento exponencial, são necessárias funções mais complexas para descrever o comportamento de escolha real (Fudenberg & Levine, 2006; Kalenscher & Pennartz, 2008; Loewenstein, O’Donoghue, & Rabin, 2003; O’Donoghue & Rabin, 2000; Ok & Masatlioglu, 2007). De seguida, apresentaremos três modelos propostos para as EIs – utilidade descontada, desconto quase-hiperbólico, e desconto hiperbólico – recorrendo aos estudos de Samuelson (1937), McClure e colaboradores (2004), e Kable e Glimcher (2007). Estes estudos diferem em vários aspetos metodológicos; por exemplo, diferem na quantidade de treino que os participantes receberam. Ainda assim, vários métodos podem ser aplicados para obter as preferências intertemporais dos indivíduos.

Modelo de Utilidade Descontada. Samuelson (1937) propôs o modelo de utilidade descontada. Segundo este

modelo, os indivíduos atribuem uma utilidade a uma consequência futura, integrando-a no nível de consumo base, e descontam a utilidade da consequência por uma taxa constante por unidade de tempo (Strotz, 1956). O modelo de Samuelson (1937) descreve o DD usando uma função de decaimento exponencial, uma solução simples que utilizou a matemática comum no cálculo de interesses da altura. No entanto, são evidentes várias violações deste modelo (Prelec & Loewenstein, 1991; Thaler, 1981), algumas delas notadas por Samuelson (1937) na altura do seu trabalho original. Por exemplo, indivíduos com comportamentos desviantes, como o abuso de substâncias ou o jogo patológico, tendem a apresentar um desconto mais acentuado, contrastando, assim, com o modelo de Samuelson. Coletivamente, estas violações implicam uma função com um desconto muito mais acentuado ao longo de escalas de curto tempo.

Modelo de Desconto Quase Hiperbólico. Em um estudo sobre EIs elaborado por McClure, Laibson,

Loewenstein, e Cohen (2004), os participantes escolheram entre pares de recompensas monetárias reais que diferiam no seu valor (entre algumas dezenas de milhar de dólares) e atraso antes da entrega (no espaço de algumas semanas). Para guiar as suas análises, os autores adotaram um enquadramento de dois processos que modela o desconto usando uma função quase-hiperbólica para valor subjetivo (U), para uma recompensa de quantidade A, a um determinado atraso D, como U = A βδD _{(Laibson, 1997; Loewenstein, 1996; Shefrin &} Thaler, 1988). De notar que esta perspetiva corresponde à descrição difundida de sistemas duplos de controlo comportamental (Bechara, 2005; Ernst & Paulus, 2005; Kahneman, 2003; Loewenstein, 1996). As suas análises procuraram regiões cerebrais cuja ativação foi consistente com cada um dos dois parâmetros modelo. O primeiro parâmetro, β, é também referido como o sistema “quente” ou visceral. Reflete uma forte preferência por recompensas imediatas quando comparadas com qualquer atraso.Assim, o parâmetro β, ou sistema “quente” , parece ser proeminente em comportamentos desviantes, uma vez que existe uma forte preferência por recompensas imediatas (Moreira, Pinto, Almeida, Barros, & Barbosa, 2016). O segundo parâmetro, δ, também referido como o sistema “frio” ou racional, produz uma taxa de desconto plana (e, como tal, reduziu a discriminabilidade entre valores de recompensa) em atrasos mais longos; nas suas análises, as regiões delta-relacionadas foram definidas com base nos níveis de ativação semelhantes, independentemente do atraso intertemporal.

Os resultados de McClure e colaboradores têm sido altamente influentes, na medida em que foram interpretados para fornecer provas para duas conclusões: em geral, que o valor de uma opção de atraso depende da interação de dois conjuntos distintos de sistemas neuronais e especificamente, que as escolhas impulsivas refletem uma sobre-ativação do sistema de recompensa (comparada com mecanismos de controlo). Este aumento de atividade do sistema de recompensa é também observado no abuso de substâncias, jogo patológico, entre outros comportamentos aditivos, explicando, assim, o padrão de escolhas impulsivas nestes indivíduos (Grant & Chamberlain, 2014).Este mesmo grupo replicou os resultados num estudo de 2007 (McClure, Ericson, Laibson, Loewenstein, & Cohen, 2007), usando recompensas líquidas (sumo e água). Neste estudo mais recente, foi dado a escolher aos participantes entre uma pequena quantidade de uma recompensa líquida agora, e uma recompensa até três vezes maior após um tempo máximo de atraso de 20 minutos. Utilizando um modelo beta/delta, verificaram a ativação de diferentes sistemas cerebrais para recompensas imediatas ou atrasadas. Estes resultados foram interpretados como uma forte evidência neural a favor de um modelo de dois componentes das EIs.

Modelo de Desconto Hiperbólico. Originalmente proposto por Loewestein e Prelec (1992), este modelo sugere

que as pessoas são, tendencialmente, extremamente impacientes para consequências num futuro próximo, mas são muito mais pacientes para consequências que ocorrem num futuro mais longínquo (Benzion, Rapoport, & Yagil, 1989; Thaler, 1981). Assim, os indivíduos tendem, de forma genérica, a descontar o futuro

hiperbolicamente. Os modelos que apoiavam a ideia da existência de desconto exponencial (e.g., Modelo da utilidade descontada) foram, de uma forma gradual, abandonados.

Além disso, diferentes funções não levam a previsões drasticamente diferentes para a maioria das situações de EIs, refletindo a semelhança geral entre funções quase hiperbólicas e hiperbólicas. Por causa desta imprecisão, os resultados de vários estudos seriam consistentes com qualquer um dos modelos. Uma estratégia potencial seria focar testes de ajuste do modelo apenas a gamas de valores específicos, onde são previsíveis fortes discrepâncias. Em segundo lugar, processos como um sistema de escolha impulsiva podiam ter influências a montante ou a jusante de uma representação de valor subjetivo. Como resumido por Tesch e Sanfey (2008), uma variedade de efeitos comportamentais pode modular o valor subjetivo na escolha intertemporal: o efeito de magnitude de recompensa - maior desconto para recompensas menores (Raineri & Rachlin, 1993; Thaler, 1981); um efeito do número de escolhas - mais escolhas aumentam o desconto (Read, 2001); um efeito de ordem - sequências crescentes são mais valorizadas do que sequências decrescentes (Loewenstein & Sicherman, 1991); uma preferência por acelerar em vez de atrasar as datas de pagamento (Loewenstein, 1988); uma preferência por recompensas distribuídas (Loewenstein & Prelec, 1993); um efeito de tipo recompensa - a comida é descontada mais intensamente que o dinheiro (Odum & Rainaud, 2003); e um efeito de valência da recompensa - ganhos são descontados de forma mais acentuada que as perdas (Thaler, 1981). Adicionalmente, as preferências de desconto podem ser modificadas por distração (Mischel, 1972). Existem também casos em que o estado mental do participante precedente à escolha modela o desconto (Berns, Laibson, & Loewenstein, 2007; Montague & King-Casas, 2007). As taxas de desconto aumentam com a privação do autocontrolo (Baumeister & Heatherton, 1996), como é observado, por exemplo, quando os fumadores são privados de nicotina (Field, Santarcangelo, Sumnall, Goudie, & Cole, 2006; Mitchell, 2004) e nas dependências em geral. Schweighofer e colaboradores (2006) propõem uma conclusão mais geral onde as escolhas dos participantes refletem a função de desconto ótima para a tarefa. Dados muito recentes também apresentam dificuldades para a simples conclusão de que o comportamento de desconto hiperbólico reflete ativação hiperbólica do sistema de recompensa.

Padrões de Preferência nas Escolhas Intertemporais

Existem diversos padrões de preferência nas EIs. Tendo em conta o que a literatura tem referido a este propósito, considerou-se pertinente focar a atenção em quatro destes padrões, que foram reportados, pela primeira vez, por Thaler (1981): (a) o efeito de diferimento; (b) efeito de magnitude; (c) efeito de sinal; e a (d) assimetria adiamento-adiantamento. Os padrões supramencionados patenteiam variações nas taxas de desconto observadas, onde uma maior taxa de desconto se refere a uma maior impaciência em ganhos e a uma maior procrastinação em perdas (Pimentel et al., 2012).

O efeito de diferimento refere que as taxas de desconto tendem a ser maiores para períodos de espera mais curtos do que para períodos de espera mais longos. Thaler (1981) conclui, a este propósito, que um indivíduo tanto pode preferir uma maçã hoje a duas amanhã como, em simultâneo, preferir duas maçãs daqui a 51 dias a uma maçã daqui a 50 dias. Este efeito tem sido apontado como um propulsionador de comportamentos dinamicamente inconsistentes, ou impulsivos, pois está relacionado com uma impaciência decrescente por parte dos indivíduos (Pimentel et al., 2012). Neste sentido, indivíduos com comportamentos impulsivos ou condições psiquiátricas caracterizadas por elevada impulsividade apresentam uma maior aversão ao atraso, preferindo, assim, escolhas que envolvam gratificação imediata.

O efeito de magnitude refere que as taxas de desconto tendem a ser maiores para consequências menores do que para consequências maiores (Chapman & Elstein, 1995; Holcomb & Nelson, 1992). As pessoas são sensíveis às diferenças relativas (contempladas pela taxa de desconto) mas também às diferenças absolutas (Loewenstein & Prelec, 1992). Receber $100 agora ou $150 daqui a um ano envolve uma diferença maior do que receber $10 agora ou $15 daqui a um ano, embora a taxa de desconto seja igual em ambas as escolhas. Pela maior diferença entre as consequências, as pessoas tendem a ser mais pacientes, e.g., tendem a preferir a consequência maior mais tarde (Pimentel et al., 2012). Contudo, este padrão de escolha pode não ser observado em indivíduos altamente impulsivos, visto que tendem a preferir consequências mais imediatas, mesmo que impliquem um ganho menor.

O efeito de sinal indica-nos que o comportamento dos indivíduos é qualitativamente diferente para enquadramentos de ganhos e de perdas. Utilizando o exemplo abordado no parágrafo anterior, por exemplo, uma pessoa que prefere receber $100 agora a receber $150 daqui a um ano (revelando-se impaciente) pode preferir pagar $100 agora a pagar $150 daqui a um ano (não se revelando procrastinador). O fenómeno de aversão às perdas, que se sabe estar alterado na psicopatia (De Martino, Camerer, & Adolphs, 2010; Shiv, Loewenstein, Bechara, Damasio, & Damasio, 2005), caracteriza-se por um comportamento do indivíduo direcionado para evitar reduções do bem-estar atuais, pois o desprazer associado a uma perda pesa mais do que o prazer associado a um ganho (Thaler & Benartzi, 1995). Desta forma, é possível concluir que os indivíduos serão mais pacientes num enquadramento de perdas e mais impacientes num enquadramento de situações de ganhos (Pimentel et al., 2012).

O efeito de sinal interage de uma forma frequente com o efeito de magnitude, no sentido do efeito de sinal ser maior para consequências pequenas do que para consequências grandes (Loewenstein & Prelec, 1992).

A assimetria adiamento-adiantamento (Loewenstein, 1988) consiste na preferência assimétrica entre adiantar e adiar o consumo. Loewenstein (1988) verificou que a quantia necessária para as pessoas adiarem um recebimento num intervalo de tempo era consideravelmente maior do que a quantia que as pessoas estavam dispostas a sacrificar para adiantar o consumo nesse intervalo. Assim, um indivíduo quando adiar um recebimento de $100 por ano pode exigir receber pelo menos $150, mas esse mesmo indivíduo quando adiantar o recebimento de $150 pode exigir receber mais do que $100 hoje. Os dois pares de escolhas são diferentes representações do mesmo par de opções subjacente, portanto, os resultados constituem um efeito de enquadramento clássico que assim é inconsistente com qualquer teoria normativa, como o modelo de utilidade descontada (Samuelson, 1937).

ENQUADRAMENTO METODOLÓGICO

Existem várias diferenças metodológicas nos estudos de DD, mas estas diferenças reportam principalmente aos parâmetros de recompensa, aos intervalos de atraso, e às manipulações de DD.

Modalidades de Recompensa. Tem havido uma notável variabilidade nos parâmetros de recompensa utilizados

nos estudos sobre EIs. Alguma desta diversidade é acionada por diferenças nos parâmetros de atraso. Para obter a variabilidade na escolha intertemporal, é necessário estruturar as magnitudes de tempo e de recompensa para que os participantes possam por vezes escolher a opção de atraso e outras vezes a opção imediata. Assim, atrasos mais longos estão muitas vezes correlacionados com recompensas maiores (Pimentel et al., 2012). Contudo, a escolha de recompensa é também acionada pelo método de pagamento. Para referência, quase todos os estudos dentro da economia comportamental analisaram escolhas por recompensas monetárias, a maioria das quais envolviam recompensas hipotéticas (Frederick, Loewenstein, & O’Donoghue, 2002). No estudo de Carter, Meyer, e Huettel (2010), as recompensas oferecidas incluíam pagamentos monetários diretos (cumulativos ou aleatórios), vales-presente ou cartões de débito, pagamentos monetários hipotéticos, ou sumo/água. A variedade de magnitude de recompensa simbólica, para uma única experiência, através dos estudos é muito ampla, desde o mínimo de $.01 (Ersner-Hershfield, Wimmer, & Knutson, 2008) até ao máximo de $40000 (Shamosh et al., 2008).

Existem exemplos de investigação comportamental prévia que fornecem indícios de que as recompensas hipotéticas podem ser tratadas de forma diferente das recompensas reais, especialmente em atrasos curtos (Camerer & Hogarth, 1999; Smith & Walker, 1993), apesar de existirem também exemplos em que não foi verificada nenhuma diferença (Baker, Johnson, & Bickel, 2003; Johnson & Bickel, 2002; Lagorio & Madden, 2005; Madden, Begotka, Raiff, & Kastern, 2003; Madden et al., 2004). Os estudos também referem pagamentos aleatórios de experiências únicas para evocar diferentes taxas de desconto das dos métodos de pagamento cumulativo (Kagel & Roth, 1997). A maioria dos estudos neuroeconómicos de escolha de risco, para comparação, utiliza um sistema de pagamento padrão: as escolhas envolvem recompensas monetárias reais que variam até algumas dezenas de milhar de dólares (ou o equivalente noutra moeda) e pagam os participantes por um subconjunto das suas escolhas, escolhidas aleatoriamente (Clithero, Carter, & Huettel, 2009; Hsu, Bhatt, Adolphs, Tranel, & Camerer, 2005; Huettel, Stowe, Gordon, Warner, & Platt, 2006).

Intervalos de Atraso. Tem havido pouca consistência no intervalo temporal através dos estudos. A maioria dos

estudos utiliza atrasos na ordem de semanas a anos, com os restantes a utilizarem intervalos de atraso muito menores que variam entre segundos a minutos. Para recompensas obtidas durante a própria experiência e para recompensas simbólicas que são acumuladas. Os primeiros intervalos são geralmente compatíveis com os utilizados na investigação da economia comportamental (Hayden & Platt, 2007; Kalenscher et al., 2005; Kalenscher et al., 2004; Rosati, Stevens, & Hauser, 2006; Stevens, Hallinan, & Hauser, 2005; Stevens, Rosati, Ross, & Hauser, 2005). Um estudo de Wittmann e colaboradores (2007) comparou atrasos (para recompensas hipotéticas) de menos de um ano com atrasos de mais de um ano, observando que regiões envolvidas na avaliação de recompensas, entre outras, diferenciavam estas duas categorias. Além disso, a utilização de recompensas reais mas não imediatas (e.g., as que são apresentadas após os participantes deixarem a sessão de laboratório) arrisca incorrer em custos de transação que podem influenciar o valor subjetivo e que não estão presentes em investigações em animais. Note-se que, por outro lado, McClure e colaboradores (2007) utilizaram tanto recompensas monetárias com atrasos de semanas (McClure et al., 2004), como recompensas de sumo com atrasos de minutos (McClure et al., 2007) e apresentaram padrões de ação relativamente semelhantes entre os dois estudos.

Tentativas de Manipulação do Desconto Diferido Apesar de ter havido várias tentativas para manipular o DD

utilizando agentes farmacológicos, virtualmente nenhum estudo apresentou com sucesso uma manipulação direta. Uma exceção é um estudo de De Wit, Enggasser, e Richards (2002), no qual a aversão ao atraso foi reduzida após a administração de anfetaminas. De modo semelhante, poucos estudos tentaram manipular o funcionamento cognitivo para influenciar o DD. Em um estudo que utilizou um desenho “entre sujeitos”, Hinson, Jameson, e Whitney (2003) observaram que uma carga de memória de trabalho aumentada estava associada a funções de desconto mais acentuadas. Além disso, o contexto e a apresentação das tarefas por si podem afetar as taxas de DD. Por exemplo, Loewenstein e Prelec (1993) verificaram que quando as decisões eram apresentadas como uma série de resultados em vez de resultados individuais, as pessoas preferiam