Interesting new study out on moral behavior. The one sentence summary of the most interesting part is that people who did one good deed were less likely to do another good deed in the near future. They had, quite literally, done their good deed for the day.

In the first part of the study, they showed that people exposed to environmentally friendly, "green" products were more likely to behave nicely. Subjects were asked to rate products in an online store; unbeknownst to them, half were in a condition where the products were environmentally friendly, and the other half in a condition where the products were not. Then they played a Dictator Game. Subjects who had seen environmentally friendly products shared more of their money.

In the second part, instead of just rating the products, they were told to select $25 worth of products to buy from the store. One in twenty five subjects would actually receive the products they'd purchased. Then they, too, played the Dictator Game. Subjects who had bought environmentally friendly products shared less of their money.

In the third part, subjects bought products as before. Then, they participated in a "separate, completely unrelated" experiment "on perception" in which they earned money by identifying dot patterns. The experiment was designed such that participants could lie about their perceptions to earn more. People who purchased the green products were more likely to do so.

This does not prove that environmentalists are actually bad people - remember that whether a subject purchased green products or normal products was completely randomized. It does suggest that people who have done one nice thing feel less of an obligation to do another.

This meshes nicely with a self-signalling conception of morality. If part of the point of behaving morally is to convince yourself that you're a good person, then once you're convinced, behaving morally loses a lot of its value.

By coincidence, a few days after reading this study, I found this article by Dr. Beck, a theologian, complaining about the behavior of churchgoers on Sunday afternoon lunches. He says that in his circles, it's well known that people having lunch after church tend to abuse the waitstaff and tip poorly. And he blames the same mechanism identified by Mazar and Zhong in their Dictator Game. He says that, having proven to their own satisfaction that they are godly and holy people, doing something else godly and holy like being nice to others would be overkill.

It sounds...strangely plausible.

If this is true, then anything that makes people feel moral without actually doing good is no longer a harmless distraction. All those biases that lead people to give time and money and thought to causes that don't really merit them waste not only time and money, but an exhaustible supply of moral fiber (compare to Baumeister's idea of willpower as a limited resource).

People here probably don't have to worry about church. But some of the other activities Dr. Beck mentions as morality sinkholes seem appropriate, with a few of the words changed:

Bible study
Voting Republican
Going on spiritual retreats
Reading religious books
Arguing with evolutionists
Sending your child to a Christian school or providing education at home
Using religious language
Avoiding R-rated movies
Not reading Harry Potter.

Let's not get too carried away with the evils of spiritual behavior - after all, data do show that religious people still give more to non-religious charities than the nonreligious do. But the points in and of themselves are valid. I've seen Michael Keenan and Patri Friedman say exactly the same thing regarding voting, and I would add to the less religion-o-centric list:

Joining "1000000 STRONG AGAINST WORLD HUNGER" type Facebook groups
Reading a book about the struggles faced by poor people, and telling people how emotional it made you
"Raising awareness of problems" without raising awareness of any practical solution
Taking (or teaching) college courses about the struggles of the less fortunate
Many forms of political, religious, and philosophical arguments

My preferred solution to this problem is to consciously try not to count anything I do as charitable or morally relevant except actually donating money to organizations. It is a bit extreme, but, like Eliezer's utilitarian foundation for deontological ethics, sometimes to escape the problems inherent in running on corrupted hardware you have to jettison all the bathwater, even knowing it contains a certain number of babies. A lot probably slips by subconsciously, but I find it better than nothing (at least, I did when I was actually making money; it hasn't worked since I went back to school. Your mileage may vary.

It may be tempting to go from here to a society where we talk much less about morality, especially little bits of morality that have no importance on their own. That might have unintended consequences. Remember that the participants in the study who saw lots of environmentally friendly products but couldn't buy any ended up nicer. The urge to be moral seems to build up by anything priming us with thoughts of morality.

But to prevent that urge from being discharged, we need to plug up the moral sinkholes Dr. Beck mentions, and any other moral sinkholes we can find. We need to give people less moral recognition and acclaim for performing only slightly moral acts. Only then can we concentrate our limited moral fiber on truly improving the world.

And by, "we", I mean "you". I've done my part just by writing this essay.

New Comment
57 comments, sorted by Click to highlight new comments since:

From Beck's article:

If you have ever worked in the restaurant industry you know the reputation of the Sunday morning lunch crowd. Millions of Christians go to lunch after church on Sundays and their behavior is abysmal...

Yikes. I worked at a small breakfast & lunch restaurant for 4 years of Sundays and never noticed this. We had a noticeable rush after the 8AM, 10AM, and 11:30 masses let out so we had a sizable Catholic crowd and I never took note of their behavior.

Not trusting my own memory, I called my mother who's been a waitress at various restaurants for 30 years, most recently including 9 years serving the after-church crowd every Sunday. I asked her whether the churchgoers behaved any differently than the other customers. She said no. I told her about Beck's article. She said they didn't act any worse or better than other customers.

Just a little anecdotal counter-evidence to reduce your possible confirmation bias the next time you go out to lunch on a Sunday. I live in Massachusetts. My mother is Catholic. Obviously the overarching point of the post may still be valid.

Upvoted. A claim like this needs more evidence than "if you have ever worked in the restaurant industry you know..."

Especially if we have people working in the restaurant industry saying otherwise.

Extra points for not trusting your own memory.

Note that there are probably group effects at work: There may be communities where Becks effect adds up and other where it doesn't (lots of group dynamics possible) and is probably discharged elsewhere.

There's some research showing the opposite effect: escalation of good-deed-doing as one good deed leads to another. For instance, a classic study on the foot-in-the-door technique found that people who were asked to put a small "keep California beautiful" sign in their window were later more likely to agree to put a huge "drive carefully" sign in their yard (as an apparently unrelated request). This escalation could also be due to a self-signaling process, as people come to believe that they're the type of person who does this sort of thing.

Part of the difference is the time scale: the self-satisfaction of doing a good deed may fade relatively quickly, while the strengthened commitment to do-gooding persists for longer. That actually fits with Baumeister's view of willpower. He's argued that willpower is like a muscle: when used it tires in the short term but is strengthened for the the long term.

So it's hard to say whether involvement in symbolic do-gooding like church, Facebook groups, or political arguments helps or hinders the pursuit of genuinely important moral causes. They're not necessarily sinkholes - they could be useful practice, building moral fiber instead of wasting it. If you let them take over and you never do anything besides "practicing," though, then you may have a problem.


These two effects can both be going on at the same time, and together the story makes sense. Suppose that any given person has a perception of themselves - do I eat a good diet, give to charity, or what not - and also keeps track of whether they have lived up to that person.

The foot-in-the-door trick works because while the requests are apparently unrelated, of course they are not: You've decided you are the type of person who puts signs up for a good cause, so this is something you would agree to again. This effect is in that context sufficient to overwhelm the feeling that they've "put up enough signs today," but as more signs were added this could change. They also probably want to be consistent with themselves.

However, the very point of the new study is that you now act less moral in an unrelated context. If I think of myself as someone who buys environmental products, that doesn't create a self-image of someone who is more generous in the dictator game or more honest.

This would suggest once again that the solution is to do only moral things that are worth doing, but also to especially focus on those that will change our moral view of ourselves, either overall or towards more productive actions.

There have been other studies showing that the self-image escalation effect (found in the foot-in-the-door study) extends beyond specific behaviors (like putting up signs) to more general types of behavior (like supporting good causes in the community) or personality traits (like honesty). In the study I described, signing a petition (instead of putting a sign in your window) also works at increasing your chances of agreeing to put a big sign in your yard. In another study, second graders were asked not to play with a desirable toy, using a mild request that made it seem (to the child) that it was their own choice not to play with the toy. The prediction was that this would lead the kids to see themselves as good boys/girls for not playing with the forbidden toy. Compared to a control group (or a group of children who were given a sterner warning, so that they saw the adult as requiring them not to play with the toy), those kids described themselves as more honest, and a couple weeks later they were less likely to cheat at a game in order to win a prize.

the self-satisfaction of doing a good deed may fade relatively quickly, while the strengthened commitment to do-gooding persists for longer. That actually fits with Baumeister's view of willpower. He's argued that willpower is like a muscle: when used it tires in the short term but is strengthened for the the long term.

This point significantly undermines the conclusion of the original post.

This doesn't sound like quite the same sort of thing.

If you put a sign up, you see the sign, and it is present as a physical reminder. With the intangible good deeds, there isn't such a token. I suspect that actually making tokens (gold stars, candies, money, Whatever.) for the good deeds in the style of the original post would not have a foot-in-the-door effect.

Can we check this somehow?

Pick a cause (say, VillageReach). In two similar locations, run a donation drive for it; in the first, have a bake sale (or other quickly consumed goods), in the second, sell those little cause-promoting bracelets (or other lasting tokens). Test altruism in both an hour later (purchase of moral satisfaction) and a week later (foot-in-the-door).

Also, a study on smokers showed that asking people to stop smoking for a short time makes them more likely to accept stopping smoking for several weeks, so tokens seem unecessary for foot-in-the-door.

Yep, exactly what I was going to say. Advocates of easy & useless do-gooding claim that such activities are a slippery-slope towards more difficult & more impactful do-gooding. I am somewhat skeptical, but this research does not contradict it.


I'm seeing tons of this on Facebook regarding Haiti relief. A proliferation of groups and events like "Wear Red for Haiti" and "Pray for Haiti" and "For every person who joins this group, I'll give $1 for Haiti, because I'm a millionaire attention whore, and hey look someone wrote 'gullible' on the ceiling" (paraphrasing, granted) and "Sending Reiki Energy Healing to Haiti" (*RAGE*). I feel like they could all have the same title: "Join here to feel better about not donating actual money to actual people doing actual helpful work in Haiti."

Cute ending. If course donating money to orgs should not be enough - they should be orgs that actually do something helpful, not just orgs whose heart is in the right place.


This reminds me of an article I read about taxi drivers who would go home early after making their daily quota instead of staying on the road longer when business was good. Everyone (the drivers, their company, the potential customers) would be better off if they kept working.

Similarly, when people meet their daily quota of good deeds, they stop being good, even though it would be better for everyone (the do-gooder, their society, and the people they counterfactually could have been good to) if they kept doing good.

Is there a general quota-bias at work here? Or is my pattern-finding algorithym misfiring?

Mike Caro, a poker player, writes about this sort of behavior. The idea here is that people psychologically want to do a little above the median each day. They work late to get up to normal money, and quit early when they do well. Whereas optimal behavior is the opposite.

Everyone (the drivers, their company, the potential customers) would be better off if they kept working.

Well, the drivers would be better off in terms of money, but this way they get to enjoy the rest of the day more than if they'd keep working. Don't confuse being better off in terms of money, and being better off in general.

(Of course, it may be that they'd also be better off in general if they kept working, but that conclusion doesn't seem certain based on what you've said so far.)


Getting off early is a good thing, but see Mycroft65536 regarding poker: they would do better by knocking off early on bad days and working longer when business is good. Either they would (a) get more money working the same hours or (b) work fewer hours for the same money. Either way, it works.

I didn't mention it, but other replies did: They'd stop working earlier on bad days instead, so the leisure time total would not decrease (well, assuming roughly equal numbers of good and bad days, with median days being the most common.) Also, its quality would likely improve as well given the financial boost of making more profits on the good days.

They'd stop working earlier on bad days instead, so the leisure time total would not decrease (well, assuming roughly equal numbers of good and bad days, with median days being the most common.)

This assumes that the effect of leisure time on well-being is merely a function of the amount of leisure time you have. That seems unwarranted: you can't simply take a long break and then spend the next year working with no leisure time at all, even if that gave you an equal amount of leisure time as a scenario where you distributed it more evenly. Likewise, it seems intuitively plausible to me that if you

a) take the day off early as a result of having made a lot of money

b) take the day off late after you've worked through a bad day and gotten a feeling of deserving that time off instead of quitting early and not accomplishing anything

your subjectively experienced leisure time quality ends up being higher than if you'd distribute the leisure time otherwise. (At least for some people - admittedly, I think I'd myself feel better if I did things the way you propose.)

I don't have any examples, but I think your hypothetical is a common problem and that a lot of leisure is degraded by guilt, and a lot of fake work is done out of guilt. But I think another effect is working with the cab drivers, possibly in addition.

Subjectively, that's likely, but there's a different way to look at it: if you think of driving the cab as providing a service, then a few people knocking off early on slow days isn't hurting the customers (after all, the remaining cabbies can handle the traffic), and working extra on busy days helps (more cabbies, less wait).

Less cab driving doesn't hurt but more cab driving helps? Either that's a weird margin we're sitting on or your rationalisations are inconsistent.

Ignoring logistical complications:

  • When there are more cabs than people who want to ride, changing the number of cabs will not change the number of cab trips, as all people will immediately take their trip in whatever cab comes up. These are the slow days.

  • When there are fewer cabs than people who want to ride, changing the number of cabs will change the number of cab trips, as all new cabs will immediately find passengers. These are the busy days.

The cab drivers can't change the number of passengers available on any given day, but they can influence the number of cabs.

That does make sense. If the 'on slow days' and 'on busy days' qualifiers were in that post when I read it then I clearly missed them.

This reminds me of an article I read about taxi drivers who would go home early after making their daily quota instead of staying on the road longer when business was good.

Labor Supply of New York City Cabdrivers: One Day at a Time

That's an interesting effect, but satiety isn't permanent. Indulging an appetite may increase it over a longer time scale (e.g. food and sex). I hope to see similar experiments that cover a significant time (at least a week).

Consciously try not to count anything [you] do as charitable or morally relevant except actually donating money to organizations.

This can be as wrong as what you're trying to counteract. The ideal is to credit work you do, and dollars you donate, only on the actual effect.

Voting Republican

"Republican" here must just stand for "your favored party", if what's important is the feeling of doing the virtuous thing (some mercenary votes are excused, e.g. choosing a president from your home state so it will get more pork).

This does not prove that environmentalists are actually bad people - remember that whether a subject purchased green products or normal products was completely randomized. It does suggest that people who have done one nice thing feel less of an obligation to do another.

No, I think it suggests that people who feel they were coerced into "being good" by people with a different notion of "good" are in a foul mood for some time thereafter.


I wonder how long-lasting this "quota" effect is. The study only looked at the immediate effects of moral behavior, not the more important long-term effects.

To make an analogy with physical exercise, maybe flexing your moral muscles exhausts your ability to be moral for the rest of the day, but when you wake up tomorrow your moral strength will be not only restored but actually strengthened. Most forms of exertion I can think of (e.g. learning, writing, working) work like this, so I wouldn't be surprised if the same held for doing good deeds.

I wonder if the is opposite true. Steal a movie. Watch some twisted porn. Post a shocker in a kids' forum. Become ridden with guilt. Be motivated to do more charity.

So you steal a movie, which means the next homeless guy you see gets change in his cup, which lets you slam the front door in a girl scout's face, the memory of which drives you to volunteer at a soup kitchen, which in turn assuages your conscience when you buy incandescent light bulbs because they look better than CFLs, so you help an old lady across the street, which relieves you of all responsibility for the other old lady who just got hit by a truck, who haunts you in your dreams, so you adopt a child, who grows up to become a mad scientist who destroys the world, thus ending the vicious cycle once and for all.

And that's why piracy is wrong.

Not that I've noticed. *cough*

It's certainly a cliche - someone being nice and moral outwardly because of a secret, shameful vice which is harmless to others.


Probably it worked. But if you start to use this as an excuse, then it will not continue to work.

Gosh, if that were true, it could be an effective group cohesion building mechanism!

Just establish some arbitrary restrictions to extremely common things, or demand the performance of really boring duties, with the requisite restitution for failing to fulfill them being a series of group identify-affirming activities!

This is exactly the type of research which I would like to see reproduced thoroughly before drawing any conclusions.

I think it's especially interesting that we "buy" a moral reputation through very public, symbolic actions and "spend" this moral reputation by being an asshole to people. I would like to see the opposite study done - if people who perform less public moral actions then choose to forgo the more public moral actions.

Hypothesis: The opposite effect would happen - people would become more willing to publicly signal after doing something genuinely nice. Why? Because they just went through all the effort to be moral, and nobody knows! So they feel entitled to a bit of moral grandstanding.

I don't think that you have reason to assume that the short term effects are the same effects as the long term effects.

We behave on average like the person we think we are. A single good action doesn't result in updating our self identity and it's easier to just update the other actions to be the person you think you are in the short term. In the long term however, doing good things changes your self image in a way where you think that you are the kind of person who does good things.

Note: The original study link is now broken. Here's an version:

Great example of ways to feel good without actually doing anything right here:

Even growing up in the Boy Scouts, I've been suspicious of the "do a good deed daily", and been trying to do good deeds at every opportunity -- but this is important to know. Thank you.

I wonder if being a "professional" environmentalist would not level this effect (in this particular setting), because once you start identifying something as A Job, it stops being quite so much a part of You. Or, rather, it becomes a different part.

Similarly, do people who routinely help at soup kitchens get desensitized to helping-poor-people-derived warm fuzzies?

I think that moral balancing is probably a subtype of what we might call "willpower balancing," because I've noticed that I engage in this sort of behavior in situations where the only person who suffers from it is me. For instance, I've gone off my diet after I've had a good job interview. I explicitly remember thinking "I've done enough for the day, I deserve to indulge."

People tend to work on the basis of checking off each of their must-do satisfaction actions for a day: 1) Catch/gather food, or do something to pay for food. Check. 2) Eat/stockpile food or stuff to barter for food. Check. 3) Do I have kids? Make sure they're still around and fed. Check. 4) Kid's other parent around? Have sex, if we're not too tired. Check?

(Note this is not a caveman kind of thing. This is my basic day, and I suspect the basic day of many others.)

Most religions are constructed around lists of things you're not supposed to do, because these lists are the only way to get the attention of things-to-do off-checking people. "I could get my food ration for the day by bonking that guy over there. But, I'm not supposed to steal, so, won't do that." We absorb those "Thou shalt nots" pretty early, for the most part, and then we move on to efficiently guiding ourselves and our families through the day. Most people get pretty good at this.

So, once we're good at checking off our satisfaction list without breaking the rules, what do we do with our left-over intellectual capacity/stockpiled food-barter? The traditional expectation in charity from every "good Christian" is "10% of your income." (Most think you can substitute volunteer time for some of that.) Also, traditionally, all that money and time went to or through the Church, but still, it's a good jumping off point. The 10% tithing, with volunteer work as acceptable substitute, rule is right there, waiting for you. But we, here, are not interested in blindly following rules that were made long, long ago in a completely different era. We want optimization.

(aside: if even a slim majority of well-off people followed tithing rules, whether to churches or to SIAI or anywhere in between, it might be horribly inefficient, but I've a feeling it would also be adequate to address most charitable needs. Still, in discussions here we aren't just interested in solving the problems we've got now, but also in having approaches that would theoretically solve whatever problems arose. /aside)

The next easiest thing for us to do is to add to our satisfaction checklist: 5) Be nice to someone. This is perhaps what most people do. The studies described fit with this, and though that doesn't prove the validity of the hypothesis, it does describe how I feel, myself. So, I offer it for consideration.

But I think the checklist has an internal daily quality to it--that is, the checklist resets daily, because the things we always have to get done, like eating and feeding the kids, usually need to get done daily or close to daily. (I think this has implications for those pondering the value of knocking off work on good days vs bad days. I find it is difficult to plan satisfaction over multiple days. Not impossible, but strategies that do not require adherence to multiple-day [as opposed to single-day or lifetime] pre-commitments may do demonstrably better.)

So, I think your chip-in-the-jar method has real merit, because it bypasses the difficulty of planning to do good things over multiple days. However, it may cause you to donate your time less than might be optimal, or to donate to new causes that might be worthy, because you daily see the jars and the good work you're planning, so beware.

Fundamentally, if my hypothesis is correct, I'm saying the gratification from doing good things may come from our own self-programming. So, A) reducing acclaim for certain acts may not make a difference, B) talking about morality often is still important, to keep being nice on the checklist C) consciously withholding self-reward for small acts may work, but it may also serve to extinct the conditioning you've already built up. Be careful when self-programming! D) The religious self-program much less than we do, and they are more charitable. If we are rational, we should do better than they. (Ack, Can't find the link where I read that exact thing here...have to go!)

Interesting, and I can think of my own examples easily enough. I would like it tested over time as a previous commenter mentioned. I wonder has science isolated the "sharing / good deed" segments of the brain, as they have for certain other moral decisions? (see here: , in the interests of your time skip ahead to 10:00 in to bypass Dawkin's introduction and subsequent thanks segment)

This reminds me of the whole carbon neutral thing. Donate money until it counters out the bad you do, and then stop.

I figure if you're going to counter out your carbon footprint, you might as well counter out a few other people while you're at it.

Very interesting.


If this is true, then I need to spam all my friends to join some Facebook charity groups.

We know that most charities basically throw away all their money. If we can get people to do less charity, world will be a better place. Paradoxically by Facebook charity groups, which hopefully can get viral and stop more and more people from money wasting.

I suppose I might be biased from ignoring all but the best charities, but I find that kind of hard to believe. Even if they throw away most of their money, if they're a reasonably good charity, what's left will still be worthwhile. Also, if you're donating at random, you have a significant chance of happening on one of the really good charities. For example, according to the Disease Control Priorities Project ( , there's one for $3 per Disability-Adjusted Life Year. In other words, you can give someone a year's worth of pleasure for three dollars.

Of corse, the best thing would probably to be just to point out those really good charities, like I did just now.