7 ms·
On the Folly of Rewarding A, While Hoping for B (1975) [pdf]
- artsyca 6y agoTotally happening in the corporate world. Rewarding employee mentality while hoping for not employee mentality.
- RcouF1uZ4gsC 6y agoAlong the same lines, one of my biggest pet peeves is rewarding using Metric C, which is used not because it actually correlates with your goals, but because it is easy to measure and get a concrete number. An example would be using lines of code written per day as a measure of developer productivity.
- artsyca 6y agoNot to mention once metrics become targets, they cease being useful as metrics because they're apt to be gamed. It seems you simply can't win when you need to guard against mindlessly selfish behaviour.
- rightbyte 6y agoProtip: Instead of saying which metrics you use you could use a opaque evaluation system that seems arbitrary for the subjects.
- artsyca 6y agoSure! And have them all report each other for thought crimes and maintain secret files on everyone. Either that or let AI sort it all out.
- RichardHeart 6y agohttps://en.wikipedia.org/wiki/Goodhart%27s_law https://en.wikipedia.org/wiki/Goodhart%27s_law
- bonoboTP 6y agoOr of course Measure D, which just so happens to be high for yourself and people you like (or is easy to optimize for you / them).
- 082349872349872 6y agoMy current favourite Metric C: "If you feel like showing off, average everything into everything else and call it the Gross Index of Total Enemy Morale. This won't fool anyone who knows the propaganda business, and you won't be able to do anything with or about it, but you can hang it on a month-by-month chart in the front office, where visitors can be impressed at getting in on a military secret. (Incidentally, if some smart enemy agent sees it and reports it back, enemy intelligence experts will go mad trying to figure out just how you got that figure. It's like the old joke that the average American is ten-elevenths White, 52% female, and always slightly pregnant.)"
- renewiltord 6y agoI enjoyed this read. It reminds me of a few things I've observed recently: * If you give a little, you will be subject to increased disapproval and if you don't care at all, you'll be subject to distant tut-tutting. Examples: the tech industry faces widespread condemnation for its gender ratio while hedge funds face little - only 1 in every 9 senior hedgies is female[0]. I'm not complaining about the gender ratio condemnation, merely commenting on the difference. Another example is that Bernie Sanders's speech was interrupted by Black Lives Matter protestors. Donald Trump's wasn't.[1] * Internet commenters frequently remark on how they would pay for good news, yet they prefer reading free low-quality news sources over paid high-quality news sources. Similarly with clickbait vs. informative writing. They share the former, they comment on the former, and they enjoy the former. Effectively, they only pay the former. * Voters and plans (brought up in the article) are my favourite. You cannot attack someone without a plan but every plan has flaws. Therefore, offering a plan is disincentivized while offering mere ideas is incentivized by people who claim the opposite. I'm glad this was example 1 in the article. * Internet commenters will frequently be unforgiving of errors corrected. They reward hiding errors and punish honest corrections. No apology is sufficient. Often this manifests as upvotes or likes for those who stick brutishly to their positions while merely abandoning to indifference those honestly convinced. Strictly this is not really only Internet commenters. The fact that Red Cross having to not give away free doughnuts resulted in animosity towards them is a lesson for future players: do not give away anything. * YC's ill-fated attempt at crowd-sourcing a funding target was another amusing result. They intended to support something that they wouldn't have heard of, but they rewarded popularity, which resulted in their having to fund something they'd heard of a million times. 0: https://docs.preqin.com/reports/Preqin-Special-Report-Women-in-Alternative-Assets-October-2017.pdf https://docs.preqin.com/reports/Preqin-Special-Report-Women-... 1: https://time.com/3989917/black-lives-matter-protest-bernie-sanders-seattle/ https://time.com/3989917/black-lives-matter-protest-bernie-s...
- artsyca 6y agoYou know, everyone's favorite stalinist republic in eastasia also operates under similar meritocratic auspices. You'll probably feel like downvoting this because you have no idea what I'm talking about.
- xdavidliu 6y agoThere's a strange point that is made on page 771. Toward the bottom of the page, it says: Such a conclusion would be wrong.² The "conclusion" being that doctors want to minimize both false positives and false negatives. The footnote is: ²In one study (4) of 14,867 films for signs of tuberculosis, 1,216 positive readings turned out to be clinically negative; only 24 negative readings proved clinically active, a ratio of 50 to 1. Can someone explain to me why the ratio matters? Surely the ratio would also depend on the actual incidence of tuberculosis in the population, and would not be determined solely by doctors' choices, right?
- antishatter 6y agoThe ratio matters because the second type of error is significantly worse ie: it is much worse to have tuberculosis and be told you do not have tuberculosis. They are looking at type 1 vs type 2 error. I think that they are saying three things 1) real tests aren't actually trying to minimize both types error of which is the opposite of the hypothesis "It might be natural to conclude that physicians seek to minimize both types of error" 2) illustrating with a very dangerous disease (to give the reader an idea of risk) the difference in types of error. 3) Demonstrating potential trade offs that occur, Type 1 error is so much less risky they are ok with 50x as much error.
- jellicle 6y agoCynically, if doctors are financially rewarded for performing treatments, doctors may prefer a test with significant false positives over one with minimal false positives. (If your knee jerks and you start saying things like 'doctors never allow financial incentives to change their medical decisions!', let me respond by saying there are many studies that say that they do, and also you can replace 'doctors' with 'health-adjacent corporations' if you like.)
- kazinator 6y agoIf nobody at all has tuberculosis, then even a test heavily biased toward false negative will still produce nothing but false positives. In that situation, only two kinds of test will avoid false positives: (1) a perfectly reliable test, or else (2) a broken tests that never reports a positive.
- asdfman123 6y agoSurely making everyone come back into the office will boost productivity.
- impendia 6y agoI find this article... strange. Near the beginning, it talks about widespread mutiny in Vietnam, and attributes it to the reward system: "What did the man at the bottom want? To go home. And when did he get to go home? When his tour of duty was over! This was the case whether or not the war was won." So, in other words, the war was lost because management did not properly align incentives for soldiers with desired outcomes? I find the following article more persuasive: https://libcom.org/history/1961-1973-gi-resistance-in-the-vietnam-war https://libcom.org/history/1961-1973-gi-resistance-in-the-vi... It identifies three important factors for the same phenomenon: -- First, that new recruits discovered immediately that their recruiters had lied to them: "Guarantees of special training and choice assignments were simply swept away." -- Secondly, for the many Black GI's, a growing consciousness of their own oppression within American society. -- And, finally, the sheer futility and meaninglessness of the war itself: "a seemingly endless ground war against an often invisible enemy, with the mass of people often openly hostile, in support of a government both unpopular and corrupt." In my opinion, the article overstates the importance of management. I work as a tenured professor at a research university -- another profession discussed in the article. Frankly, there are few rewards or disincentives for behavior of any sort -- at least not of the sort that come from my bosses. And yet I do a good job anyway, for the simple reason that I believe my job is worth doing well.
- jacobr1 6y ago> I work as a tenured professor at a research university Isn't the reward tenure itself? And it requires you to put in enough effort for your department to consider to approve you. Behavior that would prevent approval is implicitly discouraged (might be certain views held academic or not, or pedagogical approaches) and requirements (publishing, funding brought in, student feedback, favor your peers) depending on the department and field. > And yet I do a good job anyway, for the simple reason that I believe my job is worth doing well. Once tenured, you've already been selected for the traits that are desired in the role and proven your ability to execute them well.
- stepstop 6y ago> Once tenured, you've already been selected for the traits that are desired in the role and proven your ability to execute them well. I agree that the OP has a bit of self-selection bias. Similar to the “why don’t you go to college and get a real job. I did that, so can you” that internalizes the success while ignoring externalities
- motohagiography 6y agoMore recently, this concept was linked from Lesswrong/SlateStarCodex, https://www.lesswrong.com/posts/n86TNJS8k2Shd84kn/the-asshole-filter https://www.lesswrong.com/posts/n86TNJS8k2Shd84kn/the-asshol... , which I would recommend skipping in favor of summaries given it wanders and you can only read the same word over again so many times. The basic dynamic is: if you advertise a rule, and then don't enforce it, you in-effect reward people who transgress the rule and directly discourage people who obey it - and then wonder why you are "surrounded by assholes," when what you've done is set up a filter that only assholes manage to permeate. I think of this as a variant of, the "Hock Principle: Simple, clear purpose and principles give rise to complex and intelligent behavior. Complex rules and regulations give rise to simple and stupid behavior." More than post-hoc explanation, it can predict a pattern for how large organizations and bureaucracies have clusters of terrible people at the top, how an individual may develop a pattern of being subject to abusive relationships, how some social programs will just sustain the problem they are mandated to mitigate, how corruption spreads in institutions, how some people always seem to end up in terrible jobs, and even how discussion forums can turn toxic. It comes down to the maxim, "what you reward, you get." If you do not set boundaries and then enforce the boundaries you set, the only people who will pass through them will necessarily be the kind of people who don't respect you or your boundaries. Set boundaries for yourself and how you relate to the world and others, and be absolutely aware that when you choose to relate without them, you are rewarding people for not respecting you. Do the same in your companies, products, services, and teams. If you have ever worked with someone who was an asshole and wondered why and how they were able to succeed, it is because the people around the person let it work for them.
- jmcqk6 6y agoCharles Munger says it really well: "Show me the incentive, and I'll show you the outcome."
- michaelt 6y agoIMHO your summary obscured the link's point as much as it clarified it, because when you said "if you advertise a rule, and then don't enforce it" the context I thought of was "Employee car park speed limit 10mph" and the rest of the post made no sense in that light. The example given in the link - asking people to e-mail a slow-to-respond team instead of their fast-to-respond boss - makes the idea of a filter that means the boss only talks to rulebreakers a lot more logical. I do agree the article uses far too many words to say this though...
- john-radio 6y ago> ... since aspirants to public office need acceptance (from at least 50.1 percent of the people) ... cries in American
- _dwt 6y agoI stumbled on this paper very early on in my career and it significantly shifted my attitudes toward work, politics, and society. In short: people follow incentives. Of course, the suggestion of "altering the reward system" breaks down a bit in the presence of Goodhart's law, but it's still very common to see (especially in large or unusually bureaucratic organizations) policies which create incentives completely at odds with the stated goal or metric. One of my favorite examples is the "budget game" many companies play, where each department/division/etc. puts forward a proposed set of projects and budget; the corporate planners then allocate the overall corporate budget. Departments obviously "sand-bag" and ask for more than they think they need (knowing they won't get everything they ask for); there's often a "use it or lose it" policy too, which leads to what I can only describe as an orgy of spending at the end of the fiscal year. This can be great fun for vendors, consultants, and so on!
- HappyDreamer 6y ago> people follow incentives Maybe the less they care about the company's overall goals and vision, the more they follow KPI:s and bonus incentives. When there seems to be no overall purpose and meaning anyway (except for making rich people richer?), then why not game the system? But there can also be intrinsic motivation, just "unselfishlessly" doing what seems like good for the organization, if one likes its goal and reasons for existing. Edit: You wrote: "significantly shifted my attitudes toward work, politics" I wonder, how did this change how you thereafter did things at the workplace? (And politics somehow?)
- _dwt 6y agoIt just made me more cognizant that people respond to incentives rather than intentions. It's not enough to propose new "corporate values" or state that a policy proposal must be passed because it "addresses" some problem - show me how the proposed change incentivizes the desired behavior. Are there perverse incentives? Principal-agent problems? And so on. Make no mistake, I think people are "intrinsically motivated" too, at least to a point. I'd like to believe I am. But it's hard to deny that people (myself included) do things for some reason, some gain - even if that gain is something intangible like "a feeling that I'm doing good" or "perceived status in my organization".