What this covers
Russ Roberts hosts Daniel Pink, author of Drive, in a wide-ranging conversation about what motivates people in work and education. The two examine behavioral-psychology research challenging the economist's traditional faith in monetary incentives, with Pink arguing that the carrot-and-stick model works for routine tasks but fails or backfires when creative thinking is required, while Roberts counters that money still matters at the margin and that the supporting studies are often overclaimed.
The conversation spans multiple domains: the psychology of praise and learning, the neuroscience of arousal and performance, how introducing money into moral contexts can perversely change behavior (the Israeli daycare fine study), and why workplace fairness operates more like an invisible market than a stated formula. They examine concrete evidence — experiments by Edward Deci on intrinsic motivation, Teresa Amabile on progress as a motivator, and Ariely's India study showing that large bonuses improved mechanical work but damaged conceptual problem-solving. They also touch on education, where Pink argues that homework often misses the "sweet spot" of challenge, and on sports and business, noting that high performers often describe money as scorekeeping rather than their driving force. Roberts presses throughout: Do one-shot lab tasks actually predict how incentives work in sustained, real-world effort? Can overpayment be read as loss aversion rather than disengagement from money? The tone remains civil and substantive even where the two most strongly disagree — particularly on whether the research truly shows that incentives backfire, or whether the effect size has been exaggerated.
Pink argues that the 'if-then' carrot-and-stick model of motivation works for simple algorithmic work but fails or backfires for creative, conceptual work, where autonomy, mastery, and purpose are more robust drivers; Roberts pushes back that incentives still matter at the margin and that the supporting research is interpreted too strongly.
- If-then rewards suit routine work but can extinguish intrinsic motivation in complex conceptual tasks
- Beyond a market-based baseline pay, additional money has little marginal effect on performance and can be counterproductive
- Introducing money into a non-monetary moral context (e.g., daycare fines) can change behavior in unexpected, sometimes perverse ways
This asset isn't compiled yet
You're seeing its claims, ranked. Compile it to build the argument threads, weight them, and check each claim against your library — the full view.
Per Carol Dweck's research, praising children for effort and strategy (rather than for innate qualities like being 'smart' or 'a mathematician') is motivating and healthy, because praising innate ability fosters the belief that effort doesn't matter and that ability is fixed, making children less willing to exert effort.
“if you... praise for effort and strategy and not for innate qualities... what Carol Dweck's research shows is that the kid ends up... having a kind of this idea that effort doesn't matter... it's all about an innate quality... and as a result you're less likely to make an effort”
The key economic question about money and motivation is at the margin: given a certain level of autonomy, mastery, and purpose, adding money may either enhance effort slightly or become counterproductive — and good management is a delicate balance of all four, not money alone.
“the question is at the margin that is given a certain level of autonomy mastery and purpose if you then add money to it does that enhance it at the margin... how do you balance that delicate mix of autonomy mastery purpose and money”
Publication bias (a form of confirmation bias) is a pervasive problem across all research fields — economics, medicine, biology, chemistry — because non-significant or non-confirming results are discarded or never published, and researchers may run multiple trials and keep only the wanted result.
“I always wonder whether these results to be replicated if you did it again... if they suppose they'd found they tried three different tasks and... they all work the way quote conventional thinking would have suggested... you can't publish that paper so they throw that result out”
There is an inverted-U relationship (Yerkes-Dodson) between stimulation/arousal and performance: too little stimulation produces no action, too much overwhelms and freezes people, and there is a 'sweet spot' of optimal arousal in between.
“a curve that shows that certain kinds of stimulation if there's not enough stimulation people don't do anything and if there's too much people get freaked out... and that there's a sweet spot along the curve... Yerkes doc Dodson”
In the Israeli daycare study, imposing a small fine (about $3) for late pickup actually roughly doubled the incidence of lateness, because it converted a moral/social obligation into a market transaction — once a price was attached, parents felt free to 'buy' extra time and the shame-based deterrent was extinguished.
“they imposed a fine... it was a very modest fine it was three dollars... after the imposition of the fine they noticed... an increase and a number of parents coming late... it actually ended up doubling the incidence of lateness”
In the ultimatum/dictator game, a responder may reject an unequal split (e.g., $2 of $10) even though rejecting costs them their own share, for the thrill of giving the proposer nothing in response to perceived arrogance — a result that surprises economists assuming pure self-interest.
“if you and I are dividing $10... I give you two and I keep eight well you might say no even though you're gonna lose... for the thrill of giving me zero for my arrogance... that surprise some economist”
Fairness operates like a market mechanism in the workplace: people judge fair pay by what comparable people of similar experience, contribution, and ability earn elsewhere, and violating that norm causes serious problems — so top producers should be paid more, but ad hoc '$5,000 for a great idea' schemes won't reliably improve performance.
“our norm of fairness is determined by you know what someone in a similar situation... my level experience my level of contribution my level of ability would be making at a comparable place... if you violate that norm... you're in big trouble”
Education has a 'sweet spot': if a task seems impossible the student shuts down, and if it is too easy they also disengage because there is no fun in it; effective learning (and most homework) should be challenging but achievable, whereas most homework is inane and misses this spot.
“the sweet spot of Education which is if it's a if it seems impossible the kid shuts down... if it's really easy they shut down because that's no fun either you got to find the it's challenging but you can do it”
Teresa Amabile's research found that the single most motivating thing at work is a sense of making progress, supporting the idea that mastery and visible progress are inherently motivating.
“some very interesting research by Teresa Amabile... showed that the single most motivating thing at work was a sense of making progress”
Harry Harlow's 1940s experiments with monkeys solving puzzles showed primates would play with and solve puzzles in the absence of any external reward, leading him to propose a drive to do inherently interesting things alongside the reward-and-punishment drive — a controversial claim in the behaviorist, Skinner-box-dominated psychology of that era.
“Harry Harlow who essentially did an experiment where he was training monkeys to figure out how to solve a puzzle... they start playing with the puzzles... in the absence of any kind of outside reward... maybe we also have this other drive”
The classic 'if-then' motivators (if you do this, then you get that) work reasonably well for simple, straightforward, algorithmic work, but for work requiring greater complexity, creativity, and conceptual thinking those motivators often don't work and can sometimes backfire.
“they work pretty well for relatively simple simple straightforward algorithmic sort of of work... The trouble is is that for work that requires greater complexity greater creativity greater conceptual thinking there's a fair amount of evidence that says that those kinds of motivators... often don't work very well and can sometimes backfire”
High-performing companies pay above the market-clearing wage, which can be read as paying people enough that they stop focusing on money and focus on the work — but an alternative reading is that overpayment uses the threat/loss of the job (loss aversion) as the real motivator.
“these high performing companies are paying more than the market clearing wage... they are paying people enough so that they're not focused on the money and they focus on the work”
The Ariely-type result may be misinterpreted because the experimental tasks (one-shot tasks like memorizing a number or throwing a ball) don't allow effort to matter; very large stakes induce fear and anxiety that impair performance, whereas real workplace tasks unfold over time and reward sustained effort, so incentives would likely help there.
“those large rewards or punishments would induce fear and fear anxiety would make it more difficult for me to do the task... the right experiment would be something where you've got three months to work on something and you've got a trade-off working hard versus not working hard I think rewards are really important”
Management is not a natural or divine institution but an invented technology (devised in the 1850s) designed to extract compliance; in modern knowledge work what is wanted is engagement, and engagement comes through self-direction (autonomy), not coercion or monitoring.
“we have this sort of misplaced notion of management we think of it as something that emanated from nature... when in fact it's just something that some guy invented in the eighth in 50s... it's a technology really for getting compliance... today you don't want compliance you want engagement and the way that people engage... they engage through self-direction”
Unexpected, non-contingent rewards (given after the fact) are far less disruptive to creative conceptual work than purely contingent 'carrot and stick' rewards announced in advance.
“reward that is a little bit more unexpected a little less much less contingent and I think that those kinds of war those kinds of rewards are less disruptive of a lot of the complicated conceptual creative work that I think that the purely contingent carrot and sticks can disrupt”
Large contingent bonuses are highly motivating for simple/short-term tasks (people will do anything to hit their number, sometimes taking the low road), but for purely conceptual creative tasks they generate activity and long hours without necessarily producing more creative or conceptually brilliant output.
“if you say to someone if you hit your numbers... I'll give you a hundred thousand dollar bonus I think that's very motivating... for simple tasks people are going to... figure out any way possible to hit their number even if it requires taking the low road”
Edward Deci's experiments showed the counterintuitive result that introducing external contingent rewards onto an activity people already enjoy can extinguish their intrinsic interest in it, rather than adding to it as an 'additive' assumption would predict.
“Edward DC at the University of Rochester... did a series of experiments that showed... that introducing these kinds of external contingent rewards on something that people actually enjoy doing can actually extinguish their interest in doing it”
For successful entrepreneurs and great athletes, money functions less as the prime driver and more as a form of feedback and scorekeeping; the deep motivation is intrinsic (love of the work, desire to do something extraordinary), with besting opponents or earning money being a consequence rather than the goal.
“money operates in a slightly different way that is money operates as a form of feedback”
Most work in advanced economies is becoming less routine and algorithmic (because such work can be offshored or automated), so businesses are mismatching 20th-century if-then motivators against 21st-century conceptual work, creating compatibility problems.
“most work in advanced economies is becoming less routine and algorithmic because that kind of work you can send overseas or automated and that as a result we have this kind of mismatch that we're using essentially a motivational operating system... made for twentieth-century work”
Work is becoming 'more human': people increasingly seek meaning, inherent satisfaction, and self-direction from work, and work increasingly calls on higher-level skills — so treating work as inherent disutility (and long hours as inherently bad) is mistaken, since many people work long hours because they like what they do.
“the nature of work both particularly what people do and how they do it is in its own way becoming more human... this idea that work is somehow inherently a kind of disutility is mistaken”
Unexpected ex-post rewards have the advantage of being adjustable and keeping the work effort open-ended; firms like Apple likely reward top engineers with bonuses that are not specified in advance, which tends to encourage productivity better than predetermined contingent bonuses.
“rewarding an employee unexpected with any kind of whether it's praise or compensation... after the fact has the advantage of letting the bonus be adjustable which also makes the work effort more open-ended”
In the Ariely et al. India study, large performance bonuses (scaled up to roughly five months' salary) improved performance on mechanical tasks but often led to worse performance on conceptual tasks, with the most heavily incentivized people performing the worst of all — dramatically worse, not just marginally.
“when the tasks required... mechanical skill bonuses work more or less as they expected... but once you've got to any kind of conceptual skill a higher reward often led to worse performance and... the most incentivize people perform the worst of all”
Netflix has no formal vacation policy — employees can take as much vacation as they want anytime — and abuse does not occur because an emergent culture, values, and norms (Hayekian spontaneous order) regulate behavior without money or rules.
“Netflix and inside of its corporate headquarters it doesn't have a vacation policy you can take as much vacation if you want anytime you want... there isn't abuse within that culture... it works because there is this emergent order... that has nothing to do with money”
Wikipedia outperforming Encarta is a case that would have been deemed nearly theoretically impossible by conventional economics — that unpaid volunteers who don't know each other could triumph over a well-funded, profit-maximizing corporation — and is best explained by the power of 'play' (autonomy mixed with mastery and purpose).
“a bunch of people who don't know each other get together and decide to do... work for free and it will... triumph over something created with the proper incentives by a well-funded publicly held corporation profit maximizing she would say no”
Baseball players post measurably better individual numbers in the year before free agency (controlling for other factors), yet their teams don't win more — suggesting incentives elicit individual effort but possibly toward flashier individual stats rather than team success.
“in the year before they become free agents... they actually post better numbers... they try harder... but what's interesting is that... controlling for other things their teams didn't do any better”
In Roland Fryer's study of paying for academic outcomes across four districts, the two most expensive programs (Chicago and New York, paying for grades/test scores) had no effect, whereas paying for behaviors (Washington, Dallas) fared better — with the strongest, most robust effect in Dallas, where second-graders were paid $1 per book read (verified by quiz), at a fraction of the cost.
“in the two most expensive ones... Chicago and New York they were paying kids... if you do well on standardized test you get... a hundred dollars... these did not work there was no effect”
People are intensely attuned to fairness, which is why public anger focuses not on CEOs who earn a lot while their companies succeed (seen as fair) but on CEOs who earn a lot while running companies into the ground (seen as unfair); the prevailing principle in reactions to compensation is the fairness norm, with likely cultural variation in what counts as fair.
“if you're the CEO of a large public company and your company does really well... you get a lot of money I don't think most people have a problem with that... the problem is is that if you... run your company into the ground you make a lot of money”
Alfie Kohn's strong philosophical claim that any reward from another person (monetary or praise) inherently dehumanizes and replaces intrinsic with extrinsic motivation is not empirically true; rewards from others are not inherently dehumanizing.
“that doesn't mean that... any kind of reward from an external source is inherently dehumanizing and that's really his point... he has essentially a philosophical point... that rewards from other people by their very nature dehumanize... and I just don't think that's empirically true”
Paying kids cash per book ('reading is something only a chump would do for free') risks signaling that reading is undesirable work and may produce short, easy reading in the short term; Pink rejects both per-book pay and rewarding children with prizes like an iPod for reading, as a parenting and education-policy matter.
“what you've done is you said reading is sort of like working at a fast-food restaurant it's something that only a chump would do for free... I think the kid's gonna read short books... not gonna read particularly challenging”
Once you pay people enough, additional units of money have relatively little impact on additional units of performance or satisfaction; what matters more are autonomy, mastery, and purpose.
“once you pay people enough and I would argue pay people more than enough additional units of money have relatively little impact on additional units of performance or additional units of satisfaction what seems to matter more are these other elements autonomy”
Mastery — the desire to get better at things for its own sake — is a deep innate source of satisfaction that lacks an obvious economic or survival benefit (e.g., amateurs playing instruments or chess on weekends earn little), suggesting an evolutionary rather than economic explanation.
“mastery is our desire to get better at stuff because we like to get better at stuff this is you know why people play musical instruments on the weekend... it's just an undeniable deep source of satisfaction”
Some motivators are 'hygienic' (Herzberg): if you don't pay people enough you won't get motivation, but paying enough only sets a baseline standard — and fairness functions as a form of hygiene, since people will reject even beneficial offers they perceive as unfair (as in the ultimatum/dictator game).
“this harkens back a little bit to the work of Hertzberg 50 years ago and the idea that some kinds of motivators are hygienic... if you don't pay people enough you're not going to get motivation that is you have to get it has to sort of set a standard of Hygiene”
New research indicates people are happier and experience the highest states when active and engaged rather than idle, contradicting the common intuition that being completely free of obligations is the ideal state.
“new research today that shows how much happier people are it when they're active rather than when they're idle... being active and engaged and involved in something is... the sources of highest human experience”
People perform better when they do work in the service of something larger than themselves; Adam Grant's research shows that introducing purpose and context to a task makes people perform better even in the absence of economic incentives.
“there's a sense of purpose that is people tend to perform better when they do what they do in the service of something larger than themselves... research... by a fellow at Wharton named Adam grant showing that if you introduce purpose and context to a task two people perform better on it”
Effective businesses and families typically handle norm violations (lingering diners, late pickups) not with explicit textbook monetary penalties but with social signals — the upraised eyebrow, the disdainful look, shame — because monetary penalties high enough to deter create litigation, measurement disputes, and customer exit.
“the textbook economics incentives of three dollars a minute or a meter on the restaurant table for lingering over your belt those are not the ways that effective businesses handle these problems they use the upraised eyebrow the disdainful look and they use shame”
People who left or were pushed out of large organizations to work for themselves repeatedly cited feeling they 'weren't making a contribution'; the desire for one's work to have a visible impact is a deep human yearning, and workplaces that recognize people's mixed drives (to contribute, to improve, to control their lives) are better for both business and humans.
“one word that people kept using... dan I felt like I wasn't making a contribution... I think people are willing to work hard and expend effort... but they have to see some outcome... the more that workplaces become in some ways more human... it's better for business and it's better for human beings”
Human beings are not as uniformly rational as conventional economic theory assumes; they care about much more than money (pride, reputation, autonomy, mastery, purpose), as Adam Smith recognized in The Theory of Moral Sentiments.
“the human beings aren't as uniformly rational as conventional economic theory would make us out to be I agree with that... we care about a lot more than just money”
Differentiating homework by student and reframing it as 'home learning' (rather than uniform 'homework') is a better practice because most homework is inane, done out of inertia, and too easy for some and too hard for others, so no one learns.
“he gives kids different homework... my daughter isn't getting the same homework as her classmate Fred... he doesn't call it homework he called it home learning... I even think that sort of that semantics move is is very helpful”
There is a curious question of whether the pursuit of mastery (e.g., playing the bassoon or chess on weekends) has any biological or survival benefit, since it doesn't obviously help one eat, slake thirst, or reproduce, suggesting it may have had some now-unknown past survival benefit.
“playing a musical instrument playing the bassoon on the weekend or playing chess on the weekend isn't necessarily the optimal way to save your hunger... get your teams into the next generation”
Russ Roberts and Daniel Pink both blog for free, which under a purely rational net-present-value calculation looks like a loser, illustrating that people do valuable work for fun and intrinsic reasons rather than direct monetary return.
“that's why you know we you and I blog for free yeah it's fun... if we were purely rational calculators we could... say okay what is the net present value of this blog... looks like a loser”