Fighting Akrasia: Incentivising Action

post by Gordon Seidoh Worley (gworley) · 2009-04-29T13:48:56.070Z · LW · GW · Legacy · 58 comments

Contents

58 comments

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.

58 comments

Comments sorted by top scores.

comment by AnnaSalamon · 2009-04-29T21:21:56.200Z · LW(p) · GW(p)

In addition to incentives, workplaces harness, among other things:

  1. Reciprocity, and commitment/consistency effects. You accepted the job; they did their part by giving you money; you feel now like you have to work. (As an example of the power here: I recall a study in which individuals who were sent a survey, with some money inside as a "thank you for filling this survey out", had a much higher survey-completion rate than people who were credibly promised that those who returned the surveys would receive a larger sum of money after their surveys were received. Even though the first group had no incentive-structure (since they already had the money), only reciprocity.)

  2. Work-conducive contexts. Workplaces try to set up a context in which the pieces needed to work are all available, distractions are relatively unavailable, and working is the expected activity for a certain set-aside portion of one's week.

comment by whpearson · 2009-04-29T20:15:29.073Z · LW(p) · GW(p)

What I'd like is not so much a boss hanging over me, but a fellow colleague that needs my work. I think that triggers my "do stuff to avoid social status loss" behaviour.

Replies from: Cosmos
comment by Cosmos · 2009-04-30T17:39:23.455Z · LW(p) · GW(p)

This just reminded me of microcredit in developing countries. The loans are often given to a group of several people, rather than one individual, and the repayments have to be made in public in the village. The group self-selects individuals likely to repay (so they don't have to suffer for the non-repayment of others), and the public nature of the credit put their reputation on the line.

This suggests a method of implementing "do stuff to avoid social status loss" - when you wish to commit yourself to doing something, you should make a public announcement to a group whose image of you matters. That way if you back down, you lose credibility in your social group.

comment by Psychohistorian · 2009-04-29T18:32:16.881Z · LW(p) · GW(p)

There's one thing that really fights akrasia, and that's being both observed and painfully accountable.

Akrasia typically acts via future discounting. Sure, it's worth doing, but it's not worth doing now; it can be done later. Having a boss who will yell at you (or fire you) if you miss a deadline or a supervisor who will chew you out for playing minesweeper takes away the future discounting problem, because the bad event is something that happens with high probability right now. There's nothing special about employment in this context; anything that makes the pain of procrastination immediate will fight akrasia effectively.

I'm just pointing out that taking employment as your example is overly narrow. There's a very clear mechanism (present infliction) by which employment fights akrasia, and it's how to use that mechanism that really deserves attention, in my opinion.

Replies from: Cosmos, Vladimir_Nesov
comment by Cosmos · 2009-04-29T19:38:24.882Z · LW(p) · GW(p)

I agree that discounting plays a huge role in akrasia. There are immediate costs, but longer term gains. This implies that the solution to akrasia is to lower your discount rate. Problem solved.

...now, how does one go about doing that?

Replies from: pjeby, ciphergoth
comment by pjeby · 2009-04-29T22:01:39.583Z · LW(p) · GW(p)

...now, how does one go about doing that?

Move the reward closer perceptually . i.e., imagine already having the rewards you will get, viewing them as imminent and certain.

This is not a panacea, of course; there are dozens of things that can interfere with you actually doing it, as I mentioned in my video on this. And one I didn't mention is that if there is anything you perceive as bad about reaching your goal, then that is also going to be magnified by bringing it perceptually closer to you.

That is, changing your discount rate for an item applies to the bad as well as the good. (So if you have a "fear of success", this may actually enhance it.)

comment by Paul Crowley (ciphergoth) · 2009-04-29T19:53:48.759Z · LW(p) · GW(p)

"discount rate" only applies to exponential discounting. Hyperbolic discounting is parameterized by k, which is the reciprocal of how far away the reward have to be for its perceived value to be halved. The fix isn't to try to reduce k, which is obviously not directly possible, but to try to ensure the intertemporal bargaining is more often won by your longer-term interests.

Replies from: Cosmos
comment by Cosmos · 2009-04-29T22:40:31.452Z · LW(p) · GW(p)

The fix isn't to try to reduce k, which is obviously not directly possible[...]

Why is this obvious?

And could indirect measures do it?

comment by Vladimir_Nesov · 2009-04-29T19:47:57.991Z · LW(p) · GW(p)

That's interesting. This leads to a known method of setting up intermediate deadlines. Maybe a clever scheme that absorbs the risks of not meeting a few of the intermediate deadlines, but still threatens with big penalties can be effective.

comment by MrHen · 2009-04-29T18:19:13.156Z · LW(p) · GW(p)

Akrasia can strike anywhere, but one place it doesn't seem to strike too often or too severly, 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.

Akrasia struck all the time back when I was working at (gasp!) Burger King. There were employees and managers that would follow this behavior. "Well, I should do this because it is what my employer wants, but..." The trick is to do as little as possible without getting fired. Lying helps.

This is the biggest issue I have with some of the punishment/penalty systems suggested in the comments is that they teach people not to get caught instead of doing whatever it was they were supposed to be done. The clock Alicorn linked to is a minor example, but the workaround is to turn the clock off and sleep anyway. Most punishment systems have similar problems.

Replies from: billswift, dclayh
comment by billswift · 2009-04-29T22:16:04.661Z · LW(p) · GW(p)

In fact, it's probably better not to set goals that you are not going to keep in the first place, then to get in the habit of ignoring them. I make it a point now not to set my alarm clock, if I don't have to be up by particular time. Similarly, I try not to set a schedule for things I don't need to do at a particular time, I make a list and work on something from the list (unless something gets near a deadline). I found that having the list also helps me to avoid spinning my wheels thinking about what to work on.

Replies from: MrHen, gworley
comment by MrHen · 2009-04-30T01:41:03.436Z · LW(p) · GW(p)

In fact, it's probably better not to set goals that you are not going to keep in the first place, then to get in the habit of ignoring them.

True, but when something like exercise is important to your mental and physical health you can only ignore it for so long before things start mattering. Something like an alarm clock is there to make me get to work on time but my laziness kicks in anyway. No tricky alarm clock can overcome that. (Except maybe that one that turns on and starts rolling around underneath the bed.)

Personally, I find it easier to set the alarm clock on Saturday so my body/mind begins forming the habit of getting up at the same time. If I sleep in on Saturday it knows and on Monday will be whining, "But on Saturday I slept until noon!"

I think this is another area where we need to be careful of other-optimizing. I say what works for me, you can say what works for you, and between all of us hopefully the next person in line gets just the right tip.

Replies from: Cameron_Taylor
comment by Cameron_Taylor · 2009-04-30T15:14:56.892Z · LW(p) · GW(p)

True, but when something like exercise is important to your mental and physical health you can only ignore it for so long before things start mattering. Something like an alarm clock is there to make me get to work on time but my laziness kicks in anyway. No tricky alarm clock can overcome that. (Except maybe that one that turns on and starts rolling around underneath the bed.)

At my last job my trick was to place two tablets of modafinil and a bottle of energy drink on top of the alarm clock, out of reach of the bed. Since my barely awake self knows it doesn't need to get up just yet, taking the modafinil and energy drink is easy on even the laziest of days. Then I slept for another 45 minutes and woke up alert and rearing to go!

comment by Gordon Seidoh Worley (gworley) · 2009-05-01T13:34:41.779Z · LW(p) · GW(p)

This is something I've noticed, too. In my GTD system for a long time I let past due actions sit, on the theory that eventually I would do all these things that went past due because I would feel the pressure. Didn't work. What does seem to work, though, is negotiating realistic due dates with myself. If I set a date by which I want to do something and it doesn't get done by that date, then either I need to do it right now or it was something that really could have waited until later, so I renegotiate the due date with myself. I'm not a total expert at this yet, since I tend to do a lot of renegotiation, a task which should in theory be rare, but so far it seems to provide a real help in that I'm not constantly looking at things that don't actually need to be done right away.

comment by dclayh · 2009-04-29T19:21:56.474Z · LW(p) · GW(p)

In grad school also it's perfectly possible to play solitaire all the time (depending on your particular field/situation, anyway).

comment by Scott Alexander (Yvain) · 2009-04-29T15:50:48.426Z · LW(p) · GW(p)

This post made me think of this website.

Replies from: gworley
comment by Gordon Seidoh Worley (gworley) · 2009-04-29T17:19:24.940Z · LW(p) · GW(p)

Thanks for the link. I think this service is a step in the right direction, but I find I have the following problem with services with this model. When I make the commitment and set the penalties, I then tend to immediately convert the penalties I will pay into a sunk cost, so that I already account for losing that amount (this may just be the good budgeter in me, but it defeats these sorts of systems). How might we change the structure of such a system to at least prevent this type of self defeat?

Some ideas that come to my mind are:

1) Pay for meeting goals, funded by ad revenue or subscriptions to a related service (such as a discussion forum)

2) Randomly fluctuating penalties (although this will in practice require setting an acceptable range, and then you can just budget for the max)

3) Increasing penalties (especially if they increase exponentially; after I pay $1024 for not meeting the goal, then I might do something about it so I don't pay $2048 the next time)

Any other ideas on how we might modify such a system to be more successful?

Replies from: pjeby, Psychohistorian, Cosmos
comment by pjeby · 2009-04-29T17:51:18.865Z · LW(p) · GW(p)

Any other ideas on how we might modify such a system to be more successful?

The problem with this whole approach is that it presupposes an external system under your control can defeat your own desire to evade that system. (It also presupposes that you are broken and unfixable except via workarounds... which is an even worse presupposition!)

Replies from: gworley
comment by Gordon Seidoh Worley (gworley) · 2009-04-29T19:52:45.667Z · LW(p) · GW(p)

You are broken and unfixable except via workaround ... except via subjects which shall not be named but one might abbreviate IA and AI). That's the consequences of having a brain designed by evolution.

Replies from: pjeby
comment by pjeby · 2009-04-29T21:21:43.519Z · LW(p) · GW(p)

You are broken and unfixable except via workaround ... except via subjects which shall not be named but one might abbreviate IA and AI). That's the consequences of having a brain designed by evolution.

For any achievable goal you have and are failing to achieve, other people exist who are achieving it, or have already done so, without the workaround. Therefore, it's not evolution or the design of brains that's at fault.

Replies from: gworley
comment by Gordon Seidoh Worley (gworley) · 2009-04-30T13:43:20.185Z · LW(p) · GW(p)

Here we get into rather muddy water, but to be brief, although other people may be so lucky as to not need a workaround to do something, remember that individual brains in humans are not the same thing as the statistical regularity that we call a human brain. While some individual may not have a particular psychological trait, other people do, and overall we can assess whether a particular trait is common to most people (there will always be exceptions due to random variation). Further, although some person may have a brain that makes it easy to achieve some goal, this doesn't mean that every person has a brain that can be tuned to function in the same way.

You would be right if the human brain were a computer produced by a nice, low-noise process, but it's not. Evolution contains a lot of noise, and the brain works because it's designed to function even when several of its systems don't work exactly the way an engineer might have liked them to. When I say that we are broken and unfixable, I mean that each individual brain has systems that, based on centuries of attempts to change them through mere thinking, cannot be fixed without the medical technology to physical restructure brain systems.

Replies from: pjeby
comment by pjeby · 2009-04-30T14:48:01.038Z · LW(p) · GW(p)

Here we get into rather muddy water, but to be brief, although other people may be so lucky as to not need a workaround to do something, remember that individual brains in humans are not the same thing as the statistical regularity that we call a human brain. While some individual may not have a particular psychological trait, other people do, and overall we can assess whether a particular trait is common to most people (there will always be exceptions due to random variation). Further, although some person may have a brain that makes it easy to achieve some goal, this doesn't mean that every person has a brain that can be tuned to function in the same way.

Are you claiming, then, to have been born without the ability to achieve any goal? If so, the fact that you managed to post this comment would appear to falsify that claim.

Also, you seem to be using all the power of your intelligence to defend the position that you are helpless; this seems like a poor use of that intelligence. As Eliezer says, the problem with selective application of arguments is that the smarter you get, the stupider you become.

It might be more useful for you to look at how you successfully achieve certain goals, so you can find out what you're doing differently in those contexts.

Replies from: gjm, gworley
comment by gjm · 2009-05-01T14:29:50.861Z · LW(p) · GW(p)

Are you claiming, then, to have been born without the ability to achieve any goal?

It looks to me as if pjeby and gworley are miscommunicating somehow, and this looks like where the miscommunication started. gworley, as I understand him/her, is saying not "we can't do anything" but "there are lots of things we can't do, even though it seems like we ought to be able to"; different things for different people.

It is true that for every achievable (i.e., achievable-by-someone) goal you have and aren't achieving, there are other people who are achieving it or have done so. I don't know how you think you know that those people are doing it without a workaround (not least because I don't know exactly what you consider a "workaround"). But, in any case: it doesn't follow that every achievable-by-someone goal you have is achievable by you, with or without workarounds. 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.

Replies from: pjeby
comment by pjeby · 2009-05-01T15:05:47.841Z · LW(p) · GW(p)

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:

  1. 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

  2. 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.

Replies from: gjm
comment by gjm · 2009-05-01T21:00:32.178Z · LW(p) · GW(p)

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 is that most people are somewhat capable of achieving some behavioral goals, I have no disagreement with that; nor do I see anything in what gworley has written to suggest that s/he does.

And, since (as you point out) all of us frequently achieve behavioral goals, no one is suggesting that workarounds like the ones discussed here should be, or are likely to be, applied to all such goals. Only to the ones that feel really difficult, or that people have tried and failed to achieve before -- by fiat, willpower, whatever. And for those goals, the belief "I can't control my behavior in this way without hacks" is not a "presupposition". (It might none the less be a mistake, of course.)

Replies from: pjeby
comment by pjeby · 2009-05-01T22:08:52.568Z · LW(p) · GW(p)

And for those goals, the belief "I can't control my behavior in this way without hacks" is not a "presupposition". (It might none the less be a mistake, of course.)

It's a presupposition if you're assuming it's true, without having tested it.

And it's a dangerous presupposition because it rules out options that lead to an increased sense of control -- and sense of control over one's life is consistently shown to be a VERY important psychological variable for happiness, peace of mind, and motivation.

Self-fulfilling prophecies, again: presupposing lack of control leads to reduced sense of control leads to less actual control, and down the spiral goes. Presupposing greater control -- or the ability to learn control -- leads to increased sense of control leads to more actual control, and up the spiral goes.

This isn't about facts, IOW, it's about frames. I used to make the sort of arguments gworley is making, about nearly everything in my life. My wife and I would talk about trying to arrange the furniture in such a way as to thwart our tendencies to pile stuff up in high-traffic areas... instead of learning to just pick shit up.

I've spent most of my life trying to make things work more easily, instead of cultivating the ability to do difficult things. The problem is that it's effectively sending a subliminal message to yourself that you can't handle anything difficult. I was making myself weaker and weaker, when I could've been getting stronger.

That is my point: that the specific arguments gworley is using, even if "true", are not useful. I was trying to point out that there are other true things that are more useful to put into your brain, if you want it to produce good results.

This is one of those glass half empty/half full things; from a purely logical perspective there's no difference, but emotionally -- and therefore behaviorally -- there is.

Replies from: taryneast
comment by taryneast · 2011-07-25T09:39:16.743Z · LW(p) · GW(p)

My wife and I would talk about trying to arrange the furniture in such a way as to thwart our tendencies to pile stuff up in high-traffic areas... instead of learning to just pick shit up.

Cool! OOC - what strategies/heuristics did you implement to learn this ability?

comment by Gordon Seidoh Worley (gworley) · 2009-04-30T16:19:38.254Z · LW(p) · GW(p)

It might be more useful for you to look at how you successfully achieve certain goals, so you can find out what you're doing differently in those contexts.

And that's exactly what we're doing here. We're not helpless, just helpless to change certain things in certain ways.

Replies from: pjeby
comment by pjeby · 2009-04-30T17:46:54.308Z · LW(p) · GW(p)

We're not helpless, just helpless to change certain things in certain ways.

So, you've never achieved anything in your life without some kind of external carrot or stick, then? Is that what you're saying?

Replies from: gworley
comment by Gordon Seidoh Worley (gworley) · 2009-05-01T13:42:36.226Z · LW(p) · GW(p)

Of course not. I do things all the time without the reward existing outside myself. I eat because I'm hungry, and to my mind, although I can imagine all kinds of things that benefit from my not being hungry, I think it's safe to say that I eat primarily because I don't want to be hungry (I assume the brain's goal system is this shallow when it comes to hunger, though it may be deeper and I just don't know it).

comment by Psychohistorian · 2009-04-29T18:22:59.825Z · LW(p) · GW(p)

You could set up Stickk or something by making a large down payment, and then having it pay you out of this downpayment at regular intervals unless you break the rules. It's basically the same thing, but it may feel less like a sunk cost.

comment by Cosmos · 2009-04-29T17:47:58.961Z · LW(p) · GW(p)

I agree that I also convert these contracts to sunk costs. That service works for people who use the standard heuristics, however I doubt it would help most of us very much.

There are variants on this model you might find more attractive, however. Some services will set up a contract to donate your money to a particular charity that you strongly disagree with, if you don't meet your goals. That gets around the sunk cost problem, since you will still take a contingent personal loss if you fail to meet your goals.

Replies from: gworley
comment by Gordon Seidoh Worley (gworley) · 2009-05-01T13:43:16.012Z · LW(p) · GW(p)

I can still see myself accepting this as a sunk cost, but it might work for some people.

Replies from: gjm
comment by gjm · 2009-05-01T14:32:26.630Z · LW(p) · GW(p)

You should not accept it as a sunk cost, unless you are absolutely certain that meeting your goals is absolutely outside your power. (What makes a sunk cost sunk is the fact that it's too late to avoid it.) If you are certain of that, then I think you should change your goals.

(Instead of not accepting it as a sunk cost, you could take it as such and then treat meeting your goals as something that earns you a reward. That doesn't seem particularly demotivating, though.)

comment by apophenia · 2011-08-07T09:20:51.580Z · LW(p) · GW(p)

A year ago, I hired Alicorn as a manger, to tell me to do the things I want to do. I am still employing her. I am externally motivated--I don't think we've yet tried giving her authority to pay my "salary" yet. This is mostly because I'm not sure what a reasonable motivational system would look like in that case. If anyone has a suggestion I consider reasonable, I'll give it a try.

comment by Cosmos · 2009-04-29T17:35:28.851Z · LW(p) · GW(p)

I think the issue you raise here boils down to that of having a credible commitment to punish in the event of defection. It is very credible that the company you work for will indeed fire you if you don't do your job. Roland in the comments suggests "Find a romantic partner who is willing to give you a kick in the ass if you procrastinate." This is another external source of commitment, although it might or might not be credible - will your romantic partner want to inflict enough harm to get you motivated?

The key is to be able to generate a credible commitment with yourself, and from my own personal experience I think this is possible. I take promises very seriously - if I make an agreement with someone, I will fulfill it, which means I don't make a promise unless I am serious about keeping it. So far, anyway, I have yet to break a promise.

In order to overcome akrasia, I hijack this obligation system for my own use. I can make a promise to myself to do something. But where is the loss aversion? After all, can I credibly punish myself if I decide to break the promise? The reason it works is because if I make this explicit promise to myself and then subsequently break it, I can no longer say I have never broken a promise.

Unfortunately, because the stakes are so high, I only rarely make these kinds of commitments. I still struggle daily with akrasia for small decisions.

comment by Douglas_Knight · 2009-04-29T17:14:52.108Z · LW(p) · GW(p)

I agree that most people don't experience so much akrasia at work (though I have some nagging doubts in the back of my head), but I don't think this is the mechanism. I don't have much of an explanation, but I think that it is more about social expectations. or maybe attaching activities to places.

So I don't think these methods are as effective as workplaces. But that doesn't say how effective they are, or how to harness the other effects of workplaces.

Replies from: Douglas_Knight
comment by Douglas_Knight · 2009-04-30T06:50:00.624Z · LW(p) · GW(p)

Let me be a little more specific about other mechanisms. Joel Spolsky's post on akrasia, the subject of an LW post had this striking line:

eliminating overtime and working strictly 40 hour weeks...won't reduce a team's output

I think that is more suggestive of the ways in which workplaces fight akrasia, but I only say "suggestive" since I don't know what's actually going on there.

comment by Vladimir_Nesov · 2009-04-29T17:21:29.151Z · LW(p) · GW(p)

This is a good specific distinction, but I'm afraid at least partially it's circular: many of the people who inquire for the techniques to fight procrastinations, talk about their work. They manage to do the bare minimum, but they don't advance as much as they could if they found a way to effectively precommit to doing more.

comment by gwern · 2009-05-01T18:56:38.550Z · LW(p) · GW(p)

"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."

Are they really so inefficient? Worth noting that in at least a few ways, Robin Hanson thinks they're much more efficient than other methods; see http://hanson.gmu.edu/whygrant.pdf and http://www.overcomingbias.com/2007/01/prizes_versus_g.html

comment by Nwallins · 2009-05-01T02:03:12.370Z · LW(p) · GW(p)

sware box

Do you mean "swear box"?. In any case, I think you should explain this term.

Replies from: gworley
comment by Gordon Seidoh Worley (gworley) · 2009-05-01T13:26:28.891Z · LW(p) · GW(p)

Thanks, fixed. I've been having a hard time getting spell check to work right in the article editor.

A swear box is a box into which you place money as the cost for swearing, i.e. saying taboo words. The idea is that you will swear less by increasing the cost of swearing. For example, in America people who have a swear box might place a quarter or a dollar into it each time they swear. Naturally, it works best when there are others around to help keep you honest.

comment by MorganHouse · 2009-04-30T07:31:30.701Z · LW(p) · GW(p)

For most of the past three years, I have been without a regular job, mostly doing contract work. Ever since I quit my job, my goal has been to be part of a successful software startup. This requires a lot of work, and I still have not spent any significant effort on it, all the time believing that I some day would have the industry to see it through.

Last Saturday, after reading Fix it and tell us what you did, I decided I would once and for all determine whether or not I am able to work on a startup project full-time. If I am unable to spend 40 hours the first week, and 160 hours each of the following two months, I will inform my business partners that I am not sufficiently industrious to carry out this kind of work, and return to salary based work.

This is an efficient motivator, since it leaves zero time for procrastination. And should I fail to put in the required amount of work, I still won't be procrastinating, since I will cancel all my projects (giving up my dreams in the face of reality).

Replies from: rhollerith
comment by rhollerith · 2009-04-30T07:35:26.507Z · LW(p) · GW(p)

If I were you, I would not cancel your projects till you have tried having your business partners in the room with you when you are working. (Maybe you have.)

comment by CronoDAS · 2009-04-29T21:41:09.818Z · LW(p) · GW(p)

The last time I had a job, I spent most of my time "not working" and feeling guilty about it. I decided to solve the problem by leaving the job, and I've been job-free since September 2006.

For some reason, I almost never end up regretting putting things off or simply choosing not to do them at all. Don't want to do homework? There's an easy solution to that problem. Drop the course!

(Advice to anyone reading that's currently in college: take as few courses as possible each semester, so you have more time to focus on each one that you do take. Graduating in five or six years is just as good as graduating in four.)

As it turns out, the only reason I managed to get through college was that my parents were very determined to see me graduate and were willing to micromanage my life in order to make sure that I did. As far as I'm concerned, the only significant benefit I've gotten from not having dropped out of Rutgers after my sophomore year was continued financial support from my parents...

Replies from: badger, gjm
comment by badger · 2009-04-29T23:31:41.292Z · LW(p) · GW(p)

Anecdotal counter-advice: Taking fewer credits each semester was counter-productive for me. When I took 4 classes a semester, the homework and study was too spaced out. I can't maintain a regular schedule, and end up procastinating more because I have more time available. This can go too far though. Taking 7 classes, including a thesis course, and working part-time was a recipe for disaster. There is a sweet-spot where the work is steady, but not overwhelming.

Graduating in five or six years is just as good as graduating in four.

Except when you consider tuition and the opportunity cost of not starting work earlier. At my school tuition was the same between 12 and 21 credits a semester, so my wife completing her degree in three years made us better off by at least $40,000.

Replies from: Jordan
comment by Jordan · 2009-05-01T08:19:59.126Z · LW(p) · GW(p)

I second your anecdotal account. I'll procrastinate forever if I have little to do, and will become a zealous workaholic when I have endless work piled up. I call it hard-core mode.

I wonder to what extent this relates to Yvain's post on Typical Psyche Fallacy?

Replies from: gwern, gworley
comment by gwern · 2009-05-01T19:01:40.584Z · LW(p) · GW(p)

I'll procrastinate forever if I have little to do, and will become a zealous workaholic when I have endless work piled up.

Reminds me a lot of Structured Procrastination:

"All procrastinators put off things they have to do. Structured procrastination is the art of making this bad trait work for you. The key idea is that procrastinating does not mean doing absolutely nothing. Procrastinators seldom do absolutely nothing; they do marginally useful things, like gardening or sharpening pencils or making a diagram of how they will reorganize their files when they get around to it. Why does the procrastinator do these things? Because they are a way of not doing something more important. If all the procrastinator had left to do was to sharpen some pencils, no force on earth could get him do it. However, the procrastinator can be motivated to do difficult, timely and important tasks, as long as these tasks are a way of not doing something more important."

comment by Gordon Seidoh Worley (gworley) · 2009-05-01T13:27:32.018Z · LW(p) · GW(p)

Quite a lot, which I plan to use to help inform my next post on akrasia.

comment by gjm · 2009-04-29T23:44:28.808Z · LW(p) · GW(p)

Alternative advice, perhaps more useful to people not yet in college than to people already there: When choosing any sort of discretionary education, make sure you're doing something that you find interesting enough that you want to do more of it rather than less, or useful enough to give you the same motivation by indirect means.

comment by roland · 2009-04-29T16:52:15.585Z · LW(p) · GW(p)

Find a romantic partner who is willing to give you a kick in the ass if you procrastinate.

Replies from: gworley, rhollerith, taryneast
comment by Gordon Seidoh Worley (gworley) · 2009-04-29T17:22:44.331Z · LW(p) · GW(p)

While it sounds good in theory, my experience is that this rarely works, because a romantic partner is usually the person who will be most forgiving towards you. But thinking about the stereotypes of men/women in some cultures, maybe I only think it won't work because of the American cultural expectations for the behavior of romantic partners.

Replies from: roland
comment by roland · 2009-04-29T21:59:13.736Z · LW(p) · GW(p)

What you said reflects my experience in past relations so I agree.

An idea I had the other day was to actively seek for the right partner through an internet posting. You could propose a deal like: "If you manage to make me overcome my procrastination we will share the financial benefits."

comment by rhollerith · 2009-04-29T23:04:23.071Z · LW(p) · GW(p)

Unlike roland and gworley, my experience is that my current romantic partner helps me substantially in my fight against procrastination.

Specifically, my diet is better than it would be if she did not express her opinions on my diet and if I were not motivated to avoid disappointing her. (Both of us have similar chronic health problems, including food allergies.)

Also, she regularly prods me to start a medical treatment that I have been putting off for the last couple of years. Although I have not yet started the treatment, it is pretty clear that I will start it sooner than I would have without her influence. In fact, I might have never gotten around to it without her influence.

For a time, there was an open wifi network available at her apartment, and I would bring a laptop with me during visits to her place to take advantage of it. By the time the open wifi network went away, she had gotten into the habit of monitoring what I was doing on the laptop to make sure I was not wasting time. I found this monitoring quite helpful, and wish that we lived together so she could monitor more of my internet usage or that she was confident enough with computers to use vnc or something to monitor my internet usage remotely when I am at my apartment and she is at hers.

She grew up in New York City, and I get the impression from what she says about her childhood friends that women of her generation who grew up there are more likely to prod their men like this than American women in general are.

comment by taryneast · 2011-07-25T09:50:01.625Z · LW(p) · GW(p)

I find this strategy to be non-optimal, and sometimes downright counter-effective.

Firstly because it's a PITA to be that romantic partner: having to not only motivate oneself but somebody else as well. If you care about your romantic partner, you'll be helping them have an easier life, not a more difficult one (though leave yourself open for mutual motivation agreements).

Secondly - because if you externalise your procrastination, you aren't "really" learning the habit that you're trying to instill... if the external motivation stops - so will your new-found habit.

Thirdly - because the pathological case can be dire. I've seen partnerships where, for example, one person has been asked to motivate the other into giving up cigarettes... then after a while, the smoker began to resent the "nagging" from the motivator (even though they were originally asked) and then, when the smoker fell off the wagon they actually blamed the motivator for failing to motivate them enough.

Bottom line: it's ok to ask somebody to help you out... but don't use them as a crutch, and recognise that it's not universally beneficial.

Replies from: roland
comment by roland · 2011-07-26T01:18:12.712Z · LW(p) · GW(p)

You raise good points and I think I mostly agree.

From my own experience I know how much a good social environment can help. In my current job the support and inspiration from other group members has greatly helped me with getting things done(as opposed to doing them on my own). I think a lot of the problems with procrastination is that many of the tasks we do nowadays are no longer group tasks, instead we do them on our own like:

  • writing a report
  • programming
  • researching something etc...
Replies from: taryneast
comment by taryneast · 2011-07-26T10:34:04.172Z · LW(p) · GW(p)

Agreed, a supportive social environment is definitely a good thing. Especially if you surround yourself with peers who think it's normal to do what you're trying to achieve, for example - eat healthily and play sports, or who think rationally and get things done.

I think there's a big difference between "supportive friends" and "a drill-instructor whose job it is to motivate you". I think the latter is purely external motivation, whereas the former lets you set up an environment conducive to intrinsic motivation.