tl:dr; Dynamic consistency is a better term for "willpower" because its meaning is robust to changes in how we think constistent behavior actually manages to happen. One can boost consistency by fostering interactions between mutually inconsistent sub-agents to help them better empathize with each other.

Despite the common use of the term, I don't think of my "willpower" as an expendable resource, and mostly it just doesn't feel like one. Let's imagine Bob, who is somewhat overweight, likes to eat cake, and wants to lose weight to be more generically attractive and healthy. Bob often plans not to eat cake, but changes his mind, and then regrets it, and then decides he should indulge himself sometimes, and then decides that's just an excuse-meme, etc. Economists and veteran LessWrong readers know this oscillation between value systems is called dynamic inconsistency (q.v. Wikipedia). We can think of Bob as oscillating between being two different idealized agents living in the same body: a WorthIt agent, and a NotWorthIt agent.

The feeling of NotWorthIt-Bob's (in)ability to control WorthIt-Bob is likely to be called "(lack of) willpower", at least by NotWorthIt-Bob, and maybe even by WorthIt-Bob. But I find the framing and langauge of "willpower" fairly unhelpful. Instead, I think NotWorthIt-Bob and WorthIt-Bob just aren't communicating well enough. They try to ignore each other's relevance, but if they could both be present at the same time and actually talk about it, like two people in a healthy relationship, maybe they'd figure something out. I'm talking about self-empathy here, which is opposite to self-sympathy: relating to emotions of yours that you are not immediately feeling. Haven't you noticed you're better at convincing people to change their minds when you actually empathize with their position during the conversation? The same applies to convincing yourself.

Don't ask "Do I have willpower?", but "Am I a dynamically consistent team?"

Stanford Report has a university public press release about a recent paper [subscription required] in Psychological Science.  The paper is available for free from a website of one of the authors.

The gist is that they find evidence against the (currently fashionable) hypothesis that willpower is an expendable resource.  Here is the leader:

Veronika Job, Carol S. Dweck, and Gregory M. Walton
Stanford University

Abstract:

Much recent research suggests that willpower—the capacity to exert self-control—is a limited resource that is depleted after exertion. We propose that whether depletion takes place or not depends on a person’s belief about whether willpower is a limited resource. Study 1 found that individual differences in lay theories about willpower moderate ego-depletion effects: People who viewed the capacity for self-control as not limited did not show diminished self-control after a depleting experience. Study 2 replicated the effect, manipulating lay theories about willpower. Study 3 addressed questions about the mechanism underlying the effect. Study 4, a longitudinal field study, found that theories about willpower predict change in eating behavior, procrastination, and self-regulated goal striving in depleting circumstances. Taken together, the findings suggest that reduced self-control after a depleting task or during demanding periods may reflect people’s beliefs about the availability of willpower rather than true resource depletion.

(HT: Brashman, as posted on HackerNews.)

Related to: Alien Parasite Technical Guy, A Master-Slave Model of Human Preferences

In Alien Parasite Technical Guy, Phil Goetz argues that mental conflicts can be explained as a conscious mind (the "alien parasite”) trying to take over from an unsuspecting unconscious.

Last year, Wei Dai presented a model (the master-slave model) with some major points of departure from Phil's: in particular, the conscious mind was a special-purpose subroutine and the unconscious had a pretty good idea what it was doing¹. But Wei said at the beginning that his model ignored akrasia.

I want to propose an expansion and slight amendment of Wei's model so it includes akrasia and some other features of human behavior. Starting with the signaling theory implicit in Wei's writing, I'll move on to show why optimizing for signaling ability would produce behaviors like self-signaling and akrasia, speculate on why the same model would also promote some of the cognitive biases discussed here, and finish with even more speculative links between a wide range of conscious-unconscious conflicts.

The Signaling Theory of Consciousness

This model begins with the signaling theory of consciousness. In the signaling theory, the conscious mind is the psychological equivalent of a public relations agency. The mind-at-large (hereafter called U for “unconscious” and similar to Wei's “master”) has socially unacceptable primate drives you would expect of a fitness-maximizing agent like sex, status, and survival. These are unsuitable for polite society, where only socially admirable values like true love, compassion, and honor are likely to win you friends and supporters. U could lie and claim to support the admirable values, but most people are terrible liars and society would probably notice.

So you wall off a little area of your mind (hereafter called C for “conscious” and similar to Wei's “slave”) and convince it that it has only admirable goals. C is allowed access to the speech centers. Now if anyone asks you what you value, C answers "Only admirable things like compassion and honor, of course!" and no one detects a lie because the part of the mind that's moving your mouth isn't lying.

This is a useful model because it replicates three observed features of the real world: people say they have admirable goals, they honestly believe on introspection that they have admirable goals, but they tend to pursue more selfish goals. But so far, it doesn't explain the most important question: why do people sometimes pursue their admirable goals and sometimes not?

It is obvious that many people find cryonics threatening. Most of the arguments encountered in debates on the topic are not calculated to persuade on objective grounds, but function as curiosity-stoppers. Here are some common examples:

• Elevated burden of proof. As if cryonics demands more than a small amount of evidence to be worth trying.
• Elevated cost expectation. Thinking that cryonics is (and could only ever be) affordable only for the very rich.
• Unresearched suspicions regarding the ethics and business practices of cryonics organizations.
• Sudden certainty that earth-shattering catastrophes are just around the corner.
• Assuming the worst about the moral attitudes of humanity's descendants towards cryonics patients.
• Associations with prescientific mummification, or sci-fi that handwaves the technical difficulties.

The question is what causes this sensation that cryonics is a threat? What does it specifically threaten?

Custers & Aarts have a paper in the July 2 Science called "The Unconscious Will: How the pursuit of goals operates outside of conscious awareness".  It reviews work indicating that people's brains make decisions and set goals without the brains' "owners" ever being consciously aware of them.

A famous early study is Libet et al. 1983, which claimed to find signals being sent to the fingers before people were aware of deciding to move them.  This is a dubious study; it assumes that our perception of time is accurate, whereas in fact our brains shuffle our percept timeline around in our heads before presenting it to us, in order to provide us with a sequence of events that is useful to us (see Dennett's Consciousness Explained).  Also, Trevina & Miller repeated the test, and also looked at cases where people did not move their fingers; and found that the signal measured by Libet et al. could not predict whether the fingers would move.

Fortunately, the flaws of Libet et al. were not discovered before it spawned many studies showing that unconscious priming of concepts related to goals causes people to spend more effort pursuing those goals; and those are what Custers & Aarts review.  In brief:  If you expose someone, even using subliminal messages, to pictures, words, etc., closely-connected to some goals and not to others, people will work harder towards those goals without being aware of it.

Related to: The Conjunction Fallacy, Conjunction Controversy

The heuristics and biases research program in psychology has discovered many different ways that humans fail to reason correctly under uncertainty.  In experiment after experiment, they show that we use heuristics to approximate probabilities rather than making the appropriate calculation, and that these heuristics are systematically biased. However, a tweak in the experiment protocols seems to remove the biases altogether and shed doubt on whether we are actually using heuristics. Instead, it appears that the errors are simply an artifact of how our brains internally store information about uncertainty. Theoretical considerations support this view.

EDIT: The view presented here is controversial in the heuristics and biases literature; see Unnamed's comment on this post below.

EDIT 2: The author no longer holds the views presented in this post. See this comment.

A common example of the failure of humans to reason correctly under uncertainty is the conjunction fallacy. Consider the following question:

Linda is 31 years old, single, outspoken and very bright. She majored in philosophy. As a student, she was deeply concerned with issues of discrimination and social justice, and also participated in antinuclear demonstrations.

What is the probability that Linda is:

(a) a bank teller

(b) a bank teller and active in the feminist movement

In a replication by Gigerenzer, 91% of subjects rank (b) as more probable than (a), saying that it is more likely that Linda is active in the feminist movement AND a bank teller than that Linda is simply a bank teller (1993). The conjunction rule of probability states that the probability of two things being true is less than or equal to the probability of one of those things being true. Formally, P(A & B) ≤ P(A). So this experiment shows that people violate the conjunction rule, and thus fail to reason correctly under uncertainty. The representative heuristic has been proposed as an explanation for this phenomenon. To use this heuristic, you evaluate the probability of a hypothesis by comparing how "alike" it is to the data. Someone using the representative heuristic looks at the Linda question and sees that Linda's characteristics resemble those of a feminist bank teller much more closely than that of just a bank teller, and so they conclude that Linda is more likely to be a feminist bank teller than a bank teller.

This is the standard story, but are people really using the representative heuristic in the Linda problem? Consider the following rewording of the question:

Linda is 31 years old, single, outspoken and very bright. She majored in philosophy. As a student, she was deeply concerned with issues of discrimination and social justice, and also participated in antinuclear demonstrations.

There are 100 people who fit the description above. How many of them are:

(a) bank tellers

(b) bank tellers and active in the feminist movement

Notice that the question is now strictly in terms of frequencies. Under this version, only 22% of subjects rank (b) as more probable than (a) (Gigerenzer, 1993). The only thing that changed is the question that is asked; the description of Linda (and the 100 people) remains unchanged, so the representativeness of the description for the two groups should remain unchanged. Thus people are not using the representative heuristic - at least not in general.