Kantian Autonomy and the Role of the Will
Abstract and Keywords
Economists, psychologists, and philosophers have written extensively on the causes of procrastination. While their work helps us understand why people procrastinate, it does not explain how they can resist the urge to procrastinate. This chapter argues that this is a result of a refusal to acknowledge a faculty of choice, or a will, separate from preferences or desires, which can overwhelm their pull. Based on the moral philosophy of Immanuel Kant, as well as contemporary work in volitionism, this chapter applies to procrastination a Kantian-economic model of decision making that emphasizes strength of character and willpower, which can serve as alternatives to the coping strategies offered elsewhere in the literature.
Procrastination may be the most common and widespread instance of weakness of will.1 Many of us, if not most of us, procrastinate with respect to particular tasks some of the time, and some of us persistently procrastinate at certain tasks most of the time. Common cases of weakness of will can also be understood to incorporate features of procrastination; one component of the failure to control one’s eating, for instance, can be considered procrastination with regard to starting a diet or exercise program.
Economists, psychologists, and philosophers—including many of the contributors to this book—have written extensively and brilliantly on the causes of procrastination. But I would argue that they have dealt with only one side of the problem. They have described, often in excruciating detail (especially the economists), the reasons procrastination is attractive to us at the time we choose it: salience, hyperbolic discounting, and so on. But how do they recommend that people combat the urge to procrastinate? It is the same remedy they offer for any other case of weakness of will: manipulate the relevant costs and benefits, often through changes made in the choice environment, so that the akratic action is no longer the action picked by (myopic) rational deliberation. But this does not solve weakness of will as much as it avoids or sidesteps it altogether. Someone who locks her refrigerator to block late-night binges has not exhibited a strong will; she has merely accounted for the weak will she will have later by changing the payoffs now, before temptation strikes (and when her will seems stronger).
(p.217) While such scholars can explain why agents do or do not have incentives or preferences to procrastinate, they cannot explain why agents sometimes resist the urge to procrastinate—because they do not recognize the existence of the will as a faculty of choice independent of preferences or desires. While most philosophers, in the spirit of Gilbert Ryle’s “ghost in the machine” remark, do not see any reason to posit a will, economists have almost completely ignored the concept.2 The resulting deterministic model of choice denies the “agent” any true choice or agency at all. If the agent’s payoffs are such that procrastination maximizes their utility, then that agent must procrastinate, almost by definition. In this model, there can be no choice contrary to the result of one’s calculation of utility-maximizing action; the only thing the agent can do is to take another action preemptively to manipulate the relevant payoffs, because there is no way she can resist them once they are “active.”
In contrast to traditional economic and philosophical theories of decision making, I posit an active faculty of choice, a will, through which the agent makes a final choice in any decision-making situation. By incorporating the will into the economic model of choice, I refute the psychological determinism inherent in mainstream models, and allow for choices that contradict the agent’s best considered reasons for action. Taken in a Kantian context, I consider these true choices to be a function of strength of will or character, emphasizing the ethical aspect of choice itself (regardless of the ethical nature of the options among which the agent is choosing).
Relying on this conception of choice, I hope to provide an alternative to the various coping strategies suggested in other chapters in this book (and elsewhere).3 Rather than circumvent or account for her weakness of will, the agent can exercise her strength of will; simply put, she can try harder. In the same way that modern labor-saving devices have made us physically weaker (and heavier) compared with previous generations, I argue that the proliferation of coping mechanisms has made our wills weaker. This is not entirely a negative thing, of course; technological and institutional developments that economize on effort can be very beneficial, but only if they allow effort to be redirected to a more productive use. Most of us have little need for significant physical strength in our everyday lives; those of us who exercise do so primarily to improve our health or appearance. But we have no gyms or health clubs for our will, and I would argue that in the modern world, we have occasion to need strength of will more often than muscular strength. If we ever lose access (p.218) to our coping mechanisms, our willpower is all we have to fall back on, and we will be sorely disappointed if we find it missing as a result of sustained neglect.
In this chapter, I will reinterpret the phenomenon of procrastination using a Kantian-economic model of choice (incorporating a concept of the will) that I developed elsewhere.4 This model is broadly consistent with much of the previous work on procrastination by economists and philosophers, in that it includes this work as a special case and expands on it. (My contention is not that this other work is incorrect, but rather incomplete.) I begin by surveying some of the work on procrastination by economists, pointing out its strength and weaknesses. Next, I summarize the work of volitionist philosophers who maintain the existence of an independent will or faculty of choice, before explaining my Kantian-economic model of choice based on volitionism. I then explain how procrastination is understood in the context of the Kantian-economic model, compare it with similar approaches, and conclude with limitations of the exercise of willpower.
Economists’ Perspectives on Procrastination
Previous work on procrastination by economists has focused on the structure of preferences that leads to such behavior, and they have analyzed a fair range of possible circumstances in which procrastination will arise.5 The first prominent economist to address procrastination directly was George Akerlof, who based his understanding of procrastination on the salience, or vividness, of the costs of the arduous present task: “Procrastination occurs when present costs are unduly salient in comparison with future costs, leading individuals to postpone tasks until tomorrow without foreseeing that when tomorrow comes, the required action will be delayed yet again.”6 Akerlof emphasizes the costs of delaying the task, resulting from the exaggerated weight given the salient costs.
In a series of papers since 1999, Ted O’Donoghue and Matthew Rabin have offered the most extensive analysis of procrastination on the part of economists. To explain procrastination, they focus on present-biased preferences, which they consider “a more descriptive term for the underlying human characteristic that hyperbolic discounting represents”7 and functionally equivalent to Akerlof’s salience (with less implication that they (p.219) are not “real” preferences). Consistent with hyperbolic discounting, “when considering trade-offs between two future moments, present-biased preferences give stronger relative weight to the earlier moment as it gets closer.”8 In another 1999 paper, they analyze principals’ options for countering their agents’ tendencies to procrastinate, basically arguing that penalties for delay must be increased for such agents to counteract their present-bias effects.9 In a later paper, they elaborate on their basic model by introducing multiple tasks and tasks of varying importance.10 They argue that introducing a new option (a new task) with higher long-run benefit may prompt the agent to switch to that one, but the new option may also have high present costs, which will lead one to procrastinate. This is based on the observation that task choice and timing choice are made differently and can counteract each other. Also, more important projects, which may require higher up-front costs, may incite more procrastination. Most recently, they discuss procrastination in multistage projects, claiming, for instance, that projects with high start-up costs but lower finishing costs never get started (because of procrastination), while projects with low start-up costs but high finishing costs get started but not finished (again, because of procrastination).11
In “Read This Paper Later,” Carolyn Fischer eschews time inconsistency, developing a model of procrastination with time-consistent preferences, using simple marginal analysis of the work/leisure trade-off.12 Given a fixed amount of time to perform a task (such as writing a paper) and a (time-consistent) preference for present leisure, the agent will postpone the task until there is just enough time to finish it. Furthermore, this is utility-maximizing. The problem, however, lies in interpreting this behavior as procrastination, rather than rational time-allocation (based on time preference), as a later paper of Fischer’s acknowledges.13 In that paper, she utilizes time-inconsistent preferences instead, focusing on hyperbolic discounting and “differential discounting,” by which the utility from leisure is discounted at a higher rate than returns from work. (She links this to Akerlof’s salience, as well as arguing that preferences based on differential discounting can appear hyperbolic.)
This is just a sampling of economists’ work on procrastination, but I think it represents the dominant approach to studying the phenomenon. The problem with all of these explanations is that they focus on preferences or utility; in these models, it is the conflict among different sets or types of preferences that leads to the self-control problem. These models provide truly fascinating insights into the motivations behind procrastination, (p.220) but they cannot escape the tyranny of preferences and therefore cannot explain how the agent may resist the pull of his preferences and choose not to procrastinate. For that, we need a model that acknowledges that agents can somehow override their preferences—for instance, by exerting willpower.
Kantian-Economic Model of Will
Volitionism and Economics
The standard economic model of choice can best be described as constrained preference satisfaction (given information that may have been acquired rationally or provided exogenously). The economic agent chooses the option (among those available) that satisfies the highest-ranked preference within her constraints and given her information. It is in this sense that the economic agent has no true choice or agency; her decision is determined wholly by her preferences, constraints, and information. This resembles the basic Humean desire-belief model in philosophy, in which desires and beliefs determine choice (and action). J. David Velleman describes it as follows: “There is something that the agent wants, and there is an action that he believes conducive to its attainment. His desire for the end, and his belief in the action as a means, justify taking the action, and they jointly cause an intention to take it, which in turn causes the corresponding movements of the agent’s body. Provided that these causal processes take their normal course, the agent’s movements consummate an action, and his motivating desire and belief constitute his reasons for acting.”14
While fairly standard among philosophers (albeit in much more elaborate versions), nonetheless some prominent philosophers reject this model. R. Jay Wallace calls it the “hydraulic interpretation of human motivational psychology,” which is ultimately grounded in psychological determinism, leaving “no room for genuine deliberative agency. Action is traced to the operation of forces within us, with respect to which we as agents are ultimately passive, and in a picture of this kind real agency seems to drop out of view.”15 Velleman also criticizes the standard conception of choice for the passivity it implies on the part of agents; he argues that it “fails to include an agent—or, more precisely, fails to cast the agent in his proper role. In this story, reasons cause an intention, and an intention causes bodily movements, but nobody—that is, no person—does anything. Psychological and physiological events take place inside a person, but the person serves merely as the arena for these events: he takes no (p.221) active part.”16 On a more positive note, as a supplement to the standard model, Richard Holton posits a “distinct faculty of will-power,” emphasizing that it is a faculty “that the agent actively employs,” and further explores the concept of strength of will (see below).17
John Searle, in Rationality in Action, refers to the Humean picture as “the Classical Model,” which he claims “represents human rationality as a more complex version of ape rationality” (referring to experiments that showed apes to be rational decision makers).18 Searle argues that rationality requires a true act of choice or agency, and he locates this agency in “gaps” in the decision-making process, one of which exists “between the reasons for making up your mind, and the actual decision that you make.”19 In other words, the gap “occurs when the beliefs, desires, and other reasons are not experienced as causally sufficient conditions for a decision,”20 a description reminiscent of Wallace and Velleman’s criticisms of psychological determinism.21 But obviously, the activity in the gap is not itself reducible to desires, beliefs, or other reasons, so Searle asks, “What fills the gap? Nothing. Nothing fills the gap: you make up your mind to do something, or you just haul off and do what you are doing to do.”22 In Searle’s view, an “irreducible notion of the self” is necessary for understanding “our operation in the gap,” “a self that combines the capacities of rational and agency,” where agency implies “consciously try[ing] to do something.”23
In the spirit of the scholars cited above, I incorporate true agency or will into the economic model of choice. This involves modeling the activity of the will as choosing between two (or more) alternatives, which may be defined as simple options, lotteries, or preference rankings; in the interest of generality, let us refer to the options as “paths” among which the agent chooses. In interesting choice situations, these paths will be in some sort of conflict, with the two paths having qualitatively different allures. For instance, one path may be to pursue a giving life characterized by a nonegoistic moral code, while the alternative path may be one of pursuing one’s narrowly defined self-interest. One path may represent living a (p.222) healthy lifestyle of proper diet and exercise, and the other may be an easier life of sloth and gluttony. Or one path may involve long-range planning and commitment to one’s chosen goals, while the other may represent myopia and impulsive action.
We assume that the agent has a metapreference or second-order preference over the two paths, but the specific nature of that metapreference is not of concern here. However, we can imagine that in the pairings above, the agent would normally prefer to follow the first path, but the second path nonetheless has its own appeal; otherwise, there would be no conflict of interests. Most people would rather be altruistic, healthy, and resolute than the alternatives, but behaving so is a matter of willpower; one has to be able to avoid the more immediate, salient pull of the “lesser” path.
We can represent an agent’s strength of will with a simple probability distribution, in which pH is the likelihood that the agent will choose the “higher” or preferred path, while pL ≡ 1 − pH is the likelihood that the agent will choose the “lower” path. (Again, the designation of paths as higher or lower is a personal matter, not imposed from the outside, and generally speaking not necessarily a moral distinction; it is more an issue of the agent setting goals and aspirations for herself.) It is very important to note that the use of a probability distribution should not be taken to imply that the agent’s choice is random, though it may appear that way to the outside observer (or behaviorist). Rather, this technique is necessary because there is no way to model the operation of the will in a deterministic fashion—it is a matter of truly free choice.24 As Searle writes, “What makes the action a psychologically free action is precisely that the antecedent psychological causes were not sufficient to cause it.”25 According to this framework, we would expect a person with pH = 80 percent to choose the better path four times out of five and the lesser path one time out of five. Once again, this does not describe a stochastic process; rather, this person’s will is such that she simply fails to choose to do the “better” thing 20 percent of the time.
Kant and Heteronomy
What does it mean when someone fails to follow her (chosen) higher path? In precisely what way has she failed? If we take the higher path to be one consistent with the moral law, then, in Kantian terms, we would say the person has been heteronomous or has allowed inclination to affect her moral decisions. This stands in opposition to autonomy, the ideal of (p.223) making moral choices according to (and out of respect for) duty, with no undue influence from inclinations or desires (much less external authority). Every person, because of her rationality, has the capacity for autonomy (or inner freedom), but her strength or virtue in this regard can vary: “For while the capacity to overcome all opposing sensible impulses can and must be simply presupposed in man on account of his freedom, yet this capacity as strength is something he must acquire.”26 In the Kantian-economic model described in the last section, pH represents this strength of character, or virtue, which Kant defines as “the strength of a human being’s maxims in fulfilling his duty.”27
There are several ways in which a person can be heteronomous. One is “the general weakness of the human heart in complying with the adopted maxims, or the frailty of human nature.”28 I will henceforth call this simple weakness, which describes the person who is generally of strong will but occasionally lapses in her moral duty. Consider the person who cheats on his diet once in a while (assuming dieting accords with duty) but for the most part sticks to it.29 As the term implies, this is merely weakness, resulting from the imperfect rationality and morality of human beings and the constant, relentless pull of inclination, which no person can resist all of the time (although some are more successful—stronger—than others). As Kant writes, “I incorporate the good (the law) into the maxim of my power of choice; but this good, which is an irresistible incentive … is subjectively the weaker (in comparison with inclination) whenever the maxim is to be followed.”30 More precisely, this weakness does not imply any viciousness on the part of the agent but “only a lack of virtue … which indeed can coexist with the best will.”31 In such cases, inclination has not influenced the determination of the person’s maxim but has only interfered with executing the maxim itself.
The other version of heteronomy is more troubling; this is the case of the impure will, which describes the person who allows consideration of inclination to influence her deliberations over maxims.32 Impurity of the (p.224) will involves a conscious choice to be heteronomous, a surrender in the endless fight against inclination, as opposed to simple weakness, which represents merely a temporary loss of control. Since an impure will allows inclination to participate in the determination of maxims (and, through them, actions), such a person is much less virtuous than the merely weak one, sliding dangerously close to vice and letting inclination take over completely (threatening to drive pH to zero).33
Kant emphasizes the distinction between simple weakness and impurity of the will in his discussions of affect and passion. To Kant, affect (sometimes translated as “emotion”) is more akin to a momentary impulse, “feelings” that, “preceding reflection,” are temporarily powerful enough to interfere with the execution of our maxims but then pass quickly (“this tempest quickly subsides”).34 Passion, on the other hand, is a “sensible desire that has become a lasting inclination” and therefore more stable and long-lasting.35 Simple weakness can be understood as an agent succumbing to affect (such as a sudden, irresistible craving for chocolate or a burst of anger), which overwhelms her action, while the impure will is based on a strong, persistent taste or passion (such a lifelong love of chocolate or perpetual anger), which she admits—even welcomes—into the determination of her maxims.36 In other words, while affect overwhelms our rational faculties, passions become an intrinsic part of them, corrupting the very process rather than just the result. The impure will involves a deliberate submission, a choice to admit the influence of inclination—as Kant writes, “the calm with which one gives oneself up to it permits reflection and allows the mind to form principles upon it”—and is therefore more blameworthy than simple weakness (which nonetheless must be fought).37
In the context of the Kantian-economic model, both of these types of heteronomy contribute to a lower pH, but I argue they do so in different ways, which will prove important for the upcoming discussion of procrastination. The degree to which an agent is “simply weak,” insofar as this is a determinant of pH, would seem to be more stable, since the impulses (affects) that trigger it are transitory and do not necessarily signal a significant change in the agent’s strength of character or her process of moral (p.225) deliberation. Nevertheless, maintenance of a certain pH involves effort even in the face of mere weakness, and any relaxation of moral resolve can result in more frequent lapses and a lower pH. To use the dieting example again, the strong-willed person will not give in to every craving he experiences. With a pH of 80 percent (based only on simple weakness), he will only give in to temptation one time out of five. This does take an exertion of will—without it, he would give in every time—but it is his “baseline” exertion of will, or what he has become used to. As Kant says, strength of will develops through practice (as well as contemplation), so if one can avoid temptations nine times out of 10, we would say his will has strengthened, and his pH will have risen to 90 percent. Character is only as strong as the obstacles it overcomes, which we can safely interpret as frequency, as well as degree, of temptation.38
Compared with simple weakness, impurity of will poses a more serious threat to strength of character, based on its essential corruptive nature. To some extent, impurity (based on the influence of passions) overrules—and eventually lowers—an agent’s strength of will or pH. (Not for nothing does Kant refer to passions as “cancerous stores for pure practical reason.”)39 The deliberate consideration of inclination in decision making weakens the separation of the higher and lower paths. In a sense, rather than represent a probability, pH may become more of a linear combination term, such as in computing expected utility; the agent may construct his maxim on a linear combination of the two paths (rankings, “utility” functions, etc.), and in that way incorporates inclination from the lesser path into his deliberation. To the extent that the impure will allows inclination to matter, it is as if the agent is choosing the lesser path more often, which implies a lower pH.
While affect “produces a momentary loss of freedom and self-control,” passion “surrenders both, and finds pleasure and satisfaction in a servile disposition.”40 So, a simpler way of thinking of the workings of the impure will is that it simply gives up or submits to inclination, not even exerting willpower, and clearly diminishing pH. As strength “can be recognized only by the obstacles it can overcome, and in the case of virtue these obstacles are natural inclinations, which can come into conflict with the human being’s moral resolutions,” an act of an impure will would be one that lacks strength, because it does not even try to restrict the obstacle of temptation by passion.41
(p.226) Finally, regarding virtue or strength, Kant writes “if it is not rising, [it] is unavoidably sinking.”42 As soon as the agent stops trying to resist inclination, it expands its grasp on him, and the pH thereby falls as he is more likely to succumb to temptation in the future. But by the same token, his will can also be made stronger, and Kant says that “the way to acquire [strength] is to enhance the moral incentive (the thought of the law), both by contemplating the dignity of the pure rational law in us and by practicing virtue.”43 Thereby, the effect of exertions of willpower are cumulative, which will help explain the persistence of procrastination (while also previewing the willpower-as-muscle theory discussed below).
For the purposes of this chapter, I regard procrastination as a temporally oriented variation of weakness of will or akrasia, in which an agent is likely to put off performing (relatively) disagreeable tasks, against her better judgment, rather than enduring the displeasure now.44 As summarized above, economists also see procrastination as a variation of weakness of will, representing it as a self-control problem and explaining it by detailing the nature of the preferences that make procrastination attractive to the agent. Either their choice situation leads them to procrastinate or it does not—if it does, then they can manipulate the choice situation ahead of time such that they are no longer led to procrastinate when the time comes. But none of these scholars (with one qualified exception noted below) incorporates the role of the will in their models of procrastination, and therefore none of them can explain how people, while facing these strong incentives to procrastinate, nonetheless sometimes resist the urge to do so.
Application of the Kantian-Economic Model
In the context of the Kantian-economic model of choice, procrastination (with respect to a generic task or goal) takes the form of the lesser path, and timely action is the higher path. If this distinction between higher and lesser paths were purely prudential, this would not be controversial, but I am claiming that procrastination is an issue of character and therefore moral in nature. For this purpose, I follow Thomas E. Hill in maintaining (p.227) that procrastination—as with weakness of will in general—represents a failure of self-respect, a violation of duty to oneself, insofar as one fails to follow through on one’s goals and plans (at the appropriate time) as previously deliberated upon.45
Understood this way, the model implies that an agent has a certain likelihood of resisting procrastination based on her strength of will, virtue, or character. Regardless of the incentives she faces, procrastination still represents the lesser path, the choice favoring inclination (to delay performing the task) over duty (to honor one’s plans and not to procrastinate), and therefore the perfectly autonomous agent would never choose to procrastinate. Even if circumstances change to make procrastination more attractive, the ideal autonomous agent would not be tempted, although the heteronomous agent might be. (This assumes that those incentives do not change enough to make delay the overall best choice, both prudentially and ethically; if so, the nature of higher path itself would change, and choosing delay would no longer qualify as procrastination but, rather, would be prudent delay.)
And indeed, as Kant was well aware, none of us is a perfectly autonomous agent, for we are all at least weak to some degree, if not impure of will. All of us at times succumb to momentary impulse—an interesting television show is on, the task at hand just seems unbearable at the moment, and so forth. If our pH is high, this will occur only occasionally, and it will happen more often if the will is weaker. And if the “rational” incentives to procrastinate do not enter into our decision making formally, simple weakness is all there is to it. The agent still has a duty to strengthen her will, character, or virtue, but she is not a bad person for her weakness.
However, deliberate rational consideration of the incentives to procrastinate—the factors identified in so much detail by economists—would signal an impure will, for these elements would never be considered by an autonomous agent (even a merely weak one) in determining one’s maxim.46 Admittedly, if an agent with an impure will can manipulate her environment in such a way as to render procrastination less attractive, she will procrastinate less, and no one can deny that this is a good outcome. (p.228) Kant’s description of an impure will focuses on one whose inclinations are oriented toward good, in which case inclination and duty lead to the same action. But the danger remains, as with any instance of mixed motivation, that inclination will sometimes dominate choice and will not always correspond to duty. So, ethically speaking, it is better in the long run to resist procrastination through an act of will than to rely on one’s inclination based on deliberate manipulation of the choice environment. A person’s character, will, or virtue (as fortitude) is more essential to her “self” and less contingent on the details of a particular situation or environment, and exerting her strength of character will further develop it, so future exertions will come more naturally (though taking the same effort).47
So far, we have only discussed single incidents of procrastination, which are troubling enough. But of even greater concern is persistent procrastination, the type that we find so hard to combat and to dig out from under. There are two ways to explain persistent procrastination within the context of the Kantian-economic model. The simpler theory is to posit a simply weak will. Someone with a pH of 75 percent will procrastinate in one-quarter of the relevant choice situations, independent of whether she procrastinated previously (or how often). Suppose such an agent faces the possibility of procrastination today and succumbs to it with 25 percent probability. If the decision arises again tomorrow, she will procrastinate then, also with 25 percent probability, regardless of her decision the day before. Looked at from the beginning, the likelihood of a longer procrastination diminishes, of course—two subsequent days of procrastination occur 25 percent × 25 percent, or 6.25 percent, of the time. But the important thing is that persistent procrastination is not necessarily symptomatic of anything more elaborate than a simply weak will that occasionally manifests itself in “runs” of repeated procrastination. Nonetheless, the longer the procrastination keeps up, the less likely simple weakness is the culprit—there is probably something else going on.
It strikes us as very intuitive that procrastination one day increases the chance of procrastination in the future; in fact, this is a standard finding in the psychological literature.48 In the context of the Kantian-economic model, this would be the case if succumbing to procrastination in one situation would lower one’s pH before encountering the next similar situation (for example, putting off grading exams to watch one TV show leads to higher chances of doing it with the next show, and possibly even the next time exams need to be graded). For the merely weak person, it seems unlikely that this would occur. Such a person may repeatedly exhibit weakness, as described above, and may even experience a “slide” in strength of will or character as a result of an increase in the incidence of procrastination. (p.229) But I have argued that insofar as pH is caused by weakness, it should be fairly robust, since it is less corruptible by outside factors. Therefore, it is less likely to be affected by a single incident of procrastination in such a way as to lead to increased likelihood of procrastination the following times.
However, if the procrastination is based on the influence of an impure will on pH, and an impure will corrupts itself, then we would expect pH to fall as the agent continues to procrastinate, leading to a procrastination trap. Recall that the impure will represents laxity in resolve; the will (which is to say, the agent herself) simply gives up trying to resist the pull of inclination and instead admits its influence into her decision making at the level of maxim formation. In other words, impurity of the will implies that the influences identified by economists hold sway—influences that, as described above, may lead to less procrastination but for reasons that will disappear if the incentives change for the worse.
In either case, in the context of the Kantian-economic model, the agent still has a way out of the procrastination trap, no matter how long she has been in it: she can choose to break it through an act of will or volition. She can exercise her autonomy, her “inner freedom” or virtue, and choose to resist the increasingly strong temptation to continue to procrastinate. This is bound to happen eventually, since as long as pH does not diminish completely, there is always some willpower left, some reserve of strength the agent can summon up to resist inclination and follow the dictates of duty. But obviously, the sooner she does this, the less time she will spend in a procrastination trap, and the easier it will be to dig out of it.
There are several scholars whose analyses of procrastination (or weakness of will in general) have much in common with the ideas presented herein. Psychiatrist George Ainslie argues that willpower properly refers to the ability of agents to link their present actions to their future ones, so that procrastination today will be seen to lead to repeated procrastination in the future through a process of recursive self-prediction.49 Rather than rely on external manipulation of the choice situation, Ainslie advocates an internal restructuring of the relevant costs and benefits of acting at the prudent time rather than procrastinating.50 The formation of personal rules—preferably with bright lines precisely demarcating approved and disapproved behavior—is one example: a person establishes a rule for herself, which motivates her to resist the temptation to violate it if she believes that failing to resist this time will make future resistance less likely. If I believe that a doughnut eaten today, in violation of a personal (p.230) rule against eating doughnuts, will lead to eating doughnuts every day thereafter, I vest today’s decision with the enormous consequences of perpetual failure and will more likely pass up the doughnut today.
Insomuch as Ainslie recommends that agents reassess their incentives internally, rather than relying on manipulation of external cues, in order to achieve some degree of self-control, he is in agreement with the model presented in this chapter, as well as the general thesis regarding self-control. However, it falls short of the Kantian ideal, because the agent is still making decisions based on her incentives or inclinations, albeit manipulated, rather than making choices based on duty alone. Granted, the internal reconceptualization of incentives may have been conducted out of duty—as, too, may be external manipulations—but this is an indirect way to act out of respect for the moral law. Such effort would not have to be invested in manipulation of incentives, whether internal or external, if agents would increase their strength of will or volition and thereby their capacity to transcend the impact of incentives and inclinations altogether.
Among economists, Jeong-Yoo Kim is unique in that he does employ a version of willpower to explain resistance to procrastination, as well as the persistence of procrastination when resistance fails.51 He posits an “unconscious working of will, an automatic process of pre-programmed mechanism (will) that tends to resist yielding to temptation,” a mechanism that “is like a machine that only responds stochastically” to incentives to procrastinate.52 This fixed measure of willpower, together with a person’s perception of her own willpower, contributes to a probability function, the random draw from which determines the actual choice (what I would call her “operative willpower”). Because of this understanding, “the actual choice cannot be viewed as the result of a conscious mental process” but is instead simply stochastic.53 Therefore, whether one procrastinates or resists is just a matter of luck, and also has an effect on the probability of success in the future through a change in her perception of her strength of will.
While the model presented in this chapter does present the actual choice as a conscious, free one—a difference of more interest to philosophers than to economists—Kim’s explanation of perpetual procrastination bears significant similarity to mine, in that both of our models depict a progressive decline in willpower. Where I differ from Kim is in the nature of willpower; he regards the basic measure of willpower as fixed but the agent’s operative willpower as influenced also by her self-perception. As a result, her operative willpower can be stronger or weaker than her actual willpower, which never changes. While I recognize (though have (p.231) not yet emphasized) the possible relevance of self-perception in the operation of the will, I would argue that any influence of it would be on a person’s willpower itself; in other words, I see no relevant distinction between actual and operative willpower. If a person loses faith in her willpower, her willpower declines, period. As she succeeds in exerting her willpower, her belief in herself grows, and her willpower grows as well. But I see no reason Kim’s model could not be modified to make the effect in willpower more direct in this way.
Conclusion and a Remaining Question
It has been my contention that procrastination, like weakness of will in general, results from a lack or insufficient exertion of willpower. Procrastination can be avoided, therefore, not only by indirect measures, such as externally manipulating the choice environment or internally reconceptualizing the costs and benefits of acting now, but also directly by exerting one’s willpower. Not only is one’s willpower an omnipresent force, but its exercise will strengthen it for future use and will make us less reliant on costly external coping mechanisms.
But a practical question remains: Can we reasonably do this? Even if we accept the existence of an independent faculty of choice, surely the willpower behind it is not unlimited; it may be omnipresent, but surely it is not omnipotent. Among those scholars who acknowledge the existence of the will, some compare willpower to a muscle, which can be strengthened over time through repeated use but has limited efficacy at any one given time and can therefore be exhausted (until replenished with time). Richard Holton is the leading advocate of this view among philosophers, repudiating the Humean model for phenomenological as well as empirical reasons, arguing that explaining strength of will by strength of desire does not correspond to our experiences during periods of resoluteness: “It certainly doesn’t feel as though in employing will-power one is simply letting whichever is the stronger of one’s desires or intentions have its way. It rather feels as though one is actively doing something, something that requires effort.”54
The empirical aspect of Holton’s claim is based in large part on the work of psychologist Roy Baumeister, who, with various colleagues, has studied the exertion of self-control in a number of laboratory situations and is credited with developing the concept of “willpower as muscle.” In a survey of the psychological research, Baumeister and Mark Muraven write that “controlling one’s own behavior requires the expenditure of some inner, limited resource that is depleted afterward” but also “shows long-term improvement, just as a muscle gets stronger through exercise … (p.232) gaining strength with practice,”55 to which Heatherton and Baumeister add, “which if left alone becomes flaccid.”56 This theory complements the Kantian analysis above in that it provides psychological support for the dynamics of strength of will, in particular that the will can be strengthened with continued use.
Other aspects of this literature resemble the analysis of the will herein. For instance, Baumeister and Heatherton write of “acquiescence,” of persons giving in to undesirable impulses rather than fighting them, similar to the actions of an impure will versus simple weakness (using procrastination as one example, based on the observation that procrastinators are normally not “compelled” to put tasks off but instead actively choose to do so).57 In the same paper, the authors also write of “transcendence … a matter of focusing awareness beyond the immediate stimuli (i.e., transcending the immediate situation),” which may be considered analogous to the Kantian/Stoic concept of moral apathy (or, more generally, autonomy).58 These parallels suggest that the Kantian model of will can fruitfully be integrated with the research deriving from Baumeister’s work.
Regardless of the precise dynamics or nature of this type of strength or resolve, it seems undeniable that there are limits to the amount of willpower we can exert in any given situation or over time. In such cases of insufficient resolve, we must rely on clever coping mechanisms and manipulation of the choice environment, such as most economists, philosophers, and psychologists recommend. But we must also recognize that sometimes these extrapsychic tools are not available, feasible, or cost-effective. It is then that we find we must rely on our willpower, but if we have let our “muscles” wither through neglect, we will find them lacking when we need them most.
For helpful comments and criticism, I would like to thank the participants in the CSMN workshop, particularly Chrisoula Andreou, George Ainslie, Christine Tappolet, Sergio Tenenbaum, and Frank Wieber, as well as those who attended the Philosophy Forum at the College of Staten Island, where an early version of this chapter was discussed.
(1) . As George Ainslie states in chapter 1 of this book, procrastination is the “basic impulse” and is “as fundamental as the shape of time.” However, the identification of procrastination with weakness of will must be qualified in light of Sarah Stroud’s chapter 3, which is a detailed conceptual analysis of procrastination and how it does and does not fit into various philosophical understandings of weakness of will and akrasia.
(2) . One prominent exception is Don Ross (“Introduction”), who, like Ryle, disputes the existence of an independent will. Other treatments of the will in economics include Brennan, “Voluntary Exchange”; Cooter, “Lapses, Conflict, and Akrasia”; and Kim, “Hyperbolic Discounting” (discussed below).
(4) White, “Does Homo Economicus Have a Will?”Homo Economicus
(6) Akerlof, “Procrastination and Obedience,” 1
(7) O’Donoghue and Rabin, “Doing It Now or Later,” 103n. 2
(8) O’Donoghue and Rabin, “Doing It Now or Later,” 103
(9) O’Donoghue and Rabin, “Incentives for Procrastinators.”
(10) O’Donoghue and Rabin, “Choice and Procrastination.”
(11) O’Donoghue and Rabin, “Procrastination on Long-Term Projects.”
(12) Fischer, “Read This Paper Later.”
(13) Fischer, “Read This Paper Even Later.”
(14) Velleman, “What Happens,” 461. (In its modern form, the Humean model is normally attributed to Donald Davidson; see his Essays on Actions and Events.)
(15) Wallace, Normativity and the Will, 174
(16) Velleman, “What Happens,” 461
(17) Holton, “How Is Strength of Will Possible?” 48–49
(18) Searle, Rationality in Action, 5
(19) Searle, Rationality in Action, 14
(20) Searle, Rationality in Action, 62
(21) . Searle reminds us of Velleman when he supposes a world in which our actions were determined by our intentions: “if that were how the world worked in fact, we would not have to act on our intentions; we could, so to speak, wait for them to act by themselves. We could sit back and see how things turned out. But we can’t do that, we always have to act” (Rationality in Action, 232–233).
(22) Searle, Rationality in Action, 17
(23) Searle, Rationality in Action, 74, 83, 95
(24) . Free choice, as I use the term, should not be confused with free will in the metaphysical sense; I am rejecting psychological determinism only, not (necessarily) physical determinism. In a sense, the agent is determining her choice through her will, but there is no antecedent psychological cause (such as desires or preferences) affecting the working of her will; the agent is the determinant through her volition. (I thank Chrisoula Andreou for emphasizing the last point.)
(25) Searle, Rationality in Action, 73
(26) . Kant, Metaphysics of Morals, 397. (All citations from Kant’s works will list the Academy pagination.) On Kant’s theory of virtue as strength, see Gregor, Laws of Freedom, 70–75; Guyer, Kant on Freedom, 306–311; and Engstrom, “Inner Freedom” (and references therein, especially 290, n. 5).
(27) Kant, Metaphysics of Morals, 394
(28) Kant, Religion, 29
(29) . See below on how sticking to one’s plans, such as dieting, may be seen as following the duty of self-respect.
(30) Kant, Religion, 29
(31) Kant, Metaphysics of Morals, 408
(32) Kant, Religion, 29–30
(33) Kant, Metaphysics of Morals, 408
(34) Kant, Metaphysics of Morals, 407–408
(35) Kant, Metaphysics of Morals, 408
(36) Engstrom, “Inner Freedom,” 309n. 22)
(37) . Kant, Metaphysics of Morals, 408. To Kant, the ideal “state of health in the moral life,” deriving from the ancient Stoics, is “moral apathy,” which disregards affect (and rejects extreme passions) to achieve a “tranquil mind,” which is thereby the “true strength of virtue” (Metaphysics of Morals, 409). See also Kant, Religion, 253–254; Gregor, Laws of Freedom, 72–73; Engstrom, “Inner Freedom,” 307–308, 310n. 24; and Seidler, “Kant and the Stoics.”
(38) . However, the strength of the temptation should not matter, as this is a sensuous matter. But insofar as weakness implies allowing inclination into decision making, even temporarily, it may trigger greater weakness, and resistance may, in turn, signal a strengthening of character.
(39) Kant, Anthropology, 266
(40) Kant, Anthropology, 267
(41) Kant, Metaphysics of Morals, 394
(42) Kant, Metaphysics of Morals, 409
(43) Kant, Metaphysics of Morals, 397
(44) . I follow Andreou’s definition: “those cases of delaying in which one leaves too late or puts off indefinitely what one should—relative to one’s ends and information—have done sooner” (“Understanding Procrastination,” 183). But again, see Stroud’s chapter 3 in this volume for a discussion of the temporal nature of procrastination and the problems this poses for an interpretation of procrastination as weakness of will.
(45) Hill, “Weakness of Will.”
(46) Holton, “How Is Strength of Will Possible?”
(47) . This is similar to strengthening a muscle; see the discussion of willpower as muscle below.
(48) Kim, “Hyperbolic Discounting.”
(50) Bénabou and Tirole, “Willpower and Personal Rules.”
(51) Kim, “Hyperbolic Discounting.”
(52) Kim, “Hyperbolic Discounting,” 346, 349
(53) Kim, “Hyperbolic Discounting,” 350
(54) Holton, “How Is Strength of Will Possible?” 49
(55) . Muraven and Baumeister, “Self-Regulation,” 247, 254. A more recent paper proposes that general decision making (not just self-control situations) draws on this same limited psychological resource, resulting in “decision fatigue”; see Vohs et al., “Making Choices.”
(56) Heatherton and Baumeister, “Self-Regulation Failure,” 93
(57) Baumeister and Heatherton, “Self-Regulation Failure: Overview,” 6–9
(58) Baumeister and Heatherton, “Self-Regulation Failure: Overview,” 4