I might have the goal of running a mile in less than four minutes, or proving a mathematical theorem important enough to make me famous, or becoming President of the United States. Other people have achieved those goals. But it's at least possible that they are out of my reach.
But it is not out of your reach to train to run a four minute mile, or to campaign to be the President. It is this type of behavioral goal that is under discussion; we are not talking about generalized goal-reaching ability, but establishing desired habits.
My points are that:
all of us have successfully established other habits besides the ones under discussion, without using workarounds of the type I mentioned earlier in this thread, and
beginning from a presupposition that we lack control over our behavior, irrationally limits the number of options available for a solution.
I'd go further to say that I expect most adults have had the experience of making at least one "life-changing decision" that they then executed without further external support. Actually, no, scratch that... Most people in my experience have made lots of "never again" or "always" decisions by the time they grow up, that they've flawlessly executed, without needing to think about it. (It's just that many if not most of those decisions will have been really bad.)
So what I'm saying is, if you want to make some type of behavioral change, it's more helpful to begin with the presupposition that at some point in your life you have changed at least one thing by fiat before, and then try to repeat that process. And if that doesn't work out, to assume you've changed at least one thing by willpower, and apply those lessons, etc.
There has to be something you've done that's worked well, IOW.
So, it's established that human capabilities vary greatly: some people can run faster, or prove harder theorems, than others. It would therefore not be surprising if the same were true for behavioral goals, given how difficult they can (1) feel and (2) be, as measured by how many people achieve them. In fact, I think it's obvious that there are such differences; consider, e.g., the goal "get up every morning at 6am", and the well-established fact that there are substantial and stable differences in people's circadian rhythms.
If all you're arguing...
Related To: Incremental Progress and the Valley, Silver Chairs, Paternalism, and Akrasia, How a pathological procrastinator can lose weight
Akrasia can strike anywhere, but one place it doesn't seem to strike too often or too severely, assuming you are employed, is in the work place. You may not want to do something, and it might take considerable willpower to perform a task, but unless you want to get fired you can't always play Solitaire. The reason is clear to most working folks: you have to do your job to keep it, and not keeping your job is often worse than performing an undesirable task, so you suck it up and find the willpower to make it through the day. So one question we might ask is, how can we take this motivational method and put it to our own use?
First, let's look at the mechanics of the method. You have to perform a task and some exterior entity will pay you unless you fail utterly to perform the task. Notice that this is quite different from working for prizes, where you receive pay in exchange for performing a particular task. Financially they may appear the same, but from the inside of the human mind they are quite different. In the former case you are motivated by a potential loss, whereas in the later you are motivated by a potential gain. Since losses carry more weight than gains, in general the former model will provide more motivation than the latter, keeping in mind that loss aversion is a statistical property of human thought and there may be exceptions.
This suggests that certain techniques will work better more often than others. For example, if you run a website about rationality and need programming work done for it, you have a couple of options. You can wait for someone to volunteer their time, you can offer a prize for implementing certain features for the site, or you can offer to pay someone to do it on the condition that if they don't meet certain deadlines they won't get paid or will be paid a lesser amount. If you aren't so lucky as to have someone come along who will volunteer their time and do a fantastic job for free, you are faced with accepting mediocre free work, offering prizes, or paying someone. Since prizes are usually inefficient, it appears that offering to pay someone is the best option, so long as you are able to stipulate that there will be no or reduced pay if the work is not done on time and to specification.
It's also important that the entity with the authority to implement the loss reside outside the self. This is why, for example, a swear box works best if others are around to keep you honest (or if you're religious, believe that god is watching you): the temptation to let yourself slide just-this-one-time is too great. And this really comes back to an issue of akrasia: you don't have to expend any willpower for someone else to implement a loss on your part, whereas you do to make yourself take a loss.
In what other ways can this method work? Post in the comments further examples of applying loss aversion to overcome akrasia, with particular attention to details of the methods that can make or break them.