People are often characterized as poor savers; an attentional asymmetry away from money-saved relative to money-earned, potentially contributes to decreased everyday salience and future wealth

From 2018: Differential temporal salience of earning and saving. Kesong Hu, Eve De Rosa & Adam K. Anderson. Nature Communications, volume 9, Article number: 2843 (2018). https://www.nature.com/articles/s41467-018-05201-9/

Abstract: People are often characterized as poor savers. Here we examined whether cues associated with earning and saving have differential salience for attention and action. We first modeled earning and saving after positive and negative variants of monetary reinforcement, i.e., gains versus avoiding loss. Despite their equivalent absolute magnitude in a monetary incentive task, colors predicting saving were judged to appear after those that predicted earning in a temporal-order judgment task. This saving posteriority effect also occurred when savings were framed as earnings that come slightly later. Colors predicting savings, whether they acquired either negative or positive value, persisted in their posteriority. An attentional asymmetry away from money-saved relative to money-earned, potentially contributes to decreased everyday salience and future wealth.

---
Introduction

In the parable of the ant and the grasshopper, the ant’s assiduous collection of food, saving for the winter, contrasts with the grasshopper’s pursuits of immediate gratification. Cast more as grasshopper than ant, humans are often characterized as poor savers. This reputation may be well earned, in particular in America. According to a 2016 analysis of the Federal Reserve’s 2013 survey of consumer finances, the median American working-age couple has saved only $5000 for retirement with 43% of working-age families estimated to have no retirement savings at all1. On a long downward trend, the personal savings rate (expressed as a percentage of disposable personal income, DPI), dropped to <3% of DPI at the close of 20172. Contrasting with a near 96% employment rate, we have ant-like work ethic, yet earnings are rarely converted into savings.

Here we assess, in the context of monetary reinforcement, whether earning and saving reflect an asymmetry in value-derived attention3,4,5, asking whether the attentional scales are tipped in one’s favor. The way we attend has important interactions with value, perception, decision making, and ultimately behavior3. To assess their comparative behavioral and attentional salience, we considered two potential conceptual models for the distinction between earning and saving. First, we modeled earning and saving after positive and negative reinforcement, i.e., gains versus avoiding losses. Second, we considered earning and saving as variants of positive reinforcement in which gains accumulate at the same rate, but differ according to a conceptual framing of manifesting immediately or a short time later.

Inspired by language, one meaning of “to save” is to avoid loss. Saving may represent an aversion to losing one’s earnings. The assessment of gains and losses is central to our most basic physiological needs and drives6. There are evident asymmetries in the weight we place on gains and losses, with potential losses having an incommensurate influence when people evaluate identical outcomes7. In field experiments, monetary incentives framed as losses (“avoid losing A by doing B”) increases factory workers’ productivity relative to those as gains (“gain A by doing B”)8. Such loss aversion has received empirical support from a variety of studies9,10,11,12, and when directly experienced, losses outweigh gains13. Loss aversion is related to biases such as the endowment effect14 and the status quo bias15, suggesting that individuals should place greater value on savings they have already earned. But quite to the contrary, poor saving behavior16 suggests loss aversion is likely not at work in limited savings. One must be motivated to accrue savings before being concerned about losing them.

Loss aversion and related biases are thought to reflect the asymmetric weighting of punishment and reward17. Losses are punishing, resulting in an exaggerated avoidance response, biasing both decisions and the amount of attention devoted to them17,18. While the act of avoiding loss is the removal of punishment, and thus is reinforcing19. Motivation to earn versus save may, more directly, be a comparison of positive and negative variants of reinforcement20, comparing earnings with the avoidance of losing one’s earnings. Positive and negative reinforcement refer to increasing the likelihood of a behavior with the addition or subtraction of an outcome, not their positive or negative utility for an individual. Positive and negative reinforcement have been shown to similarly recruit the reward system, suggesting that both have positive utility21. Nevertheless, they may have asymmetric motivational power12. Psychologically, and in our daily experience, individuals believe they are paid for their performance rather than arranging conditions to avoid moneyless periods of time22. Savings, in this context, should motivate individuals to avoid being without money. Earning and saving should then align with different concerns. Moreover, individuals can differently experience pleasure or utility according to their promotion versus prevention orientation23, through either promoting desired versus preventing undesired outcomes.

On the other hand, the meaning of “to save” could be understood in terms of expected utility in the future, hence currently inaccessible. In line with this, efforts to avoid moneyless periods of time highlight the importance of temporal perspectives on one’s earnings. Maintaining an orientation toward saving may result in temporal discounting of today’s earnings in the future24. Discounting of future value is captured by individuals who prefer $5 now compared to $10 three months from now25,26. Participants often make choices of smaller but immediate rewards relative to rewards that are larger but delayed. Such temporal or delay discounting is also considered a marker of impulsive behavior, assessing the degree to which the subjective value of an offering decreases as a function of delay in its delivery27. While the rate of discounting depends on the individual, it is a fundamental to the representation of value, observed in human and nonhuman animals28. Temporal distance of saving for the future may also modulate value representations such that they are more abstract29. This may cognitively distance individuals from the reality of the undesirable outcomes of not saving, i.e., extended moneyless periods of time. Saving in these contexts reflects an orientation toward the future, as well as the limits of imagination on behavior30. Accordingly, while earning may reflect the here and now, savings may reflect earnings as a discounted and abstract future.

Whether earning and saving reflect varieties of reinforcement or differentially reflect temporal discounting, they involve making a choice between options31. Our nervous system is confronted, at each moment, by choices in terms of where to invest or allocate its resources in the currency of attention3,32. By paying attention, an individual is able to impact the salience33 and value34 of sensory events. While unpleasant events typically evoke relatively stronger changes in affect and attention in both perceptual and decision studies9,12,17,18,35,36, gains also play a similar role3,4,37,38,39. Importantly, value not only alters attention, but attention is also central to value, with attention-boosting34 and inattention-reducing value40. Attention can both follow and influence preference41, predicting consumer choice42. Thus, the choice to what we attend is central to value and behavior17,18,43. While multiple studies have characterized value-derived attention3,4,5, much less is known about how different variants of reinforcement and temporal framing regulate attention. Here we examined how earning and saving, according to different models, regulate the paying of attention. If earning and savings represent differential concerns to the individual, then this should be reflected in attentional choice, having an asymmetric regulatory influence on salience and awareness.

Mirroring how value is scaled relative to time44, time is also scaled relative to attention. Attention shapes not only what is perceived but also when45. Attention can warp the judgment of temporal order, with attended events appearing to occur before non-attended events, called “prior entry”46,47. Similarly, individuals attend to more immediate events and outcomes than those in the more distant future29, and this asymmetry in attention may modulate temporal discounting48. We took advantage of how attention can influence judgments of temporal order to examine how individuals perceive events predicting earning and saving. Just as how individuals may put off saving due to decreased salience and relative inattention, monetarily reinforced colors associated with savings may be less attentionally salient and appear to come later. As a model for earnings and savings, we first examined the power of positive (gains) and negative (avoiding losses) monetary reinforcement of color patches and the relationship between action and attentional salience (experiments 1a–c). In a further study (experiment 2), through distinct temporal framings of positive reinforcement, we modeled earning and saving after gains that come immediately versus gains to come later (i.e., saving for future).

Figure 1 illustrates the core tasks and the example colored circle used as stimuli. Participants started with value reinforcement trials, where equiluminant colors (red, blue, or yellow) were 100% reinforced, or received no reinforcement, for fast and accurate color discriminations. One color was associated with “earning,” gaining 30 cents, and another associated with “saving,” avoiding loss of 30 cents. The task was sufficiently easy to enable reinforcement on the majority trials, whereby earning would increase one’s balance and saving would preserve those earnings. Participants received their performance-based earnings at the end of the experiment. Color-reinforcement associations were counterbalanced across participants. The temporal-order judgment (TOJ) task required participants to judge which of the side-by-side colored stimuli appeared first, when presented in varying temporal proximity (8–98 ms). TOJ trials were pseudo-randomly intermixed with value reinforcement trials to ensure that any acquired salience for colors was maintained throughout. Similar to indifference points in temporal discounting to establish value25,26, we estimated the participant’s point of subjective simultaneity (PSS), which indicates the estimated time interval to perceive the two stimuli as arriving simultaneously, i.e., 50%47,49,50,51.
Fig. 1
Fig. 1

Illustration of the display sequences and target stimuli examples. a Monetary reinforcement task. Exp. 1a involved a color discrimination (red, blue, and yellow), while Exp. 1b and 1c involved a gap side (left and right) discrimination. After response, participants were informed about gain or loss, together with the total cash bonus accrued (in white). b Temporal-order judgment task. Following fixation, colored circles were presented either on the left or on the right side of the fixation followed by a second different color circle, which appeared on the opposite side after a variable SOA (8, 18, 38, 68, and 98 ms). Participants were required to indicate which color (Exp 1a) or which side (Exp 1b and 1c) appeared first
Full size image

Despite their equivalent absolute magnitude in a monetary incentive task, we find that saving results in less behavioral salience and decreased payout. In a temporal-order task, saving-associated color cues are also judged to appear after those that predict earning, consistent with the decreased attentional salience of saving. This saving temporal posteriority effect generalizes to when saving is framed as earnings that come slightly later. Across studies, saving-associated cues persisted in their relative inattention whether the cues acquire negative or positive valence. Thus, saving posteriority is not simply explained by acquired affective value. We conclude that decreased attentional salience related to money-saved relative to money-earned is a fundamental information-processing bias. That saving has less moment-to-moment attention attracting potential may contribute to reduced saving behavior. Attentional interventions to enhance the everyday salience of saving may be gainfully employed to improve saving behavior.