The psychology of Labor Standard perspective The standard perspective is that employers buy the time and effort of the employees in exchange for money. What are some additional assumptions of this perspective? What are some of its implications? Something seem to be missing What can psychology teach us about labor? The psychology of labor Relative vs absolute levels of compensations The relationship between payment and motivation, effort, performance Low payments, high payments Labor & meaning Sabotage Relative vs absolute levels of compensations Relative vs absolute levels of compensations Person A gets $80,000 in a company where the range is $80,000 - $100,000 Person B gets $70,000 in a company where the range is $50,000 - $70,000 Who will be happier? Who will work harder? Who will stay longer with the company? What job will you select? Salary & happiness So, happiness is at least partially determined by relative salary Relative to what? How would you order the different effects? What is the largest deraminant of them Implications How would you keep your employees happy with their salary How would you compensate them? What structural changes could you take How would you deal with salary decreases and retirement? The relationship between payment and motivation, effort, performance At low levels of payment A tale of 2-markets Imagine you are about to move to a new apartment and you need some help in packing and carrying your stuff to your new home. Who to ask for help? What to promise them as compensation? Imagine that you start a new company. How would you want to pay your employees? Hourly? Monthly? With cash or with cash & gifts? A few theories 1.2 Altruism Effort level 1.0 0.8 o 0.6 r ip c e R 0.4 0.2 0.0 ty i c Selfish maximization No Payment Low Payment High Payment Level of payment Fiske’s Rational theory (1992) Four basic types of social relationships: Communal Sharing (CS) high-level of cooperation, equal treatment of all, and “we-ness.” Authority Ranking (AR) A clear superior-subordinate relationship. Equality Matching (EM) Combine features of CS and AR relationships – they are very structured but with perfect equality. Market Pricing (MP) generally consist of on-going cost/beneﬁt analysis and participants are paid for their labor via a wage rate that reﬂects the amount and quality of the work performed Fiske’s Rational theory (1992) Four basic types of social relationships: Communal Sharing (CS) high-level of cooperation, equal treatment of all, and “we-ness.” Authority Ranking (AR) A clear superior-subordinate relationship. Equality Matching (EM) Combine features of CS and AR relationships – they are very structured but with perfect equality. Market Pricing (MP) generally consist of on-going cost/beneﬁt analysis and participants are paid for their labor via a wage rate that reﬂects the amount and quality of the work performed Hypotheses The relationship between payment and effort will depend on the type of exchange (money vs. social markets). In Money-Market relationships effort will be exerted according to the reciprocity theory. In Social-Market relationships, effort will be shaped by the altruism theory and will not be sensitive to the level of payment. What about not paying? Rich background in social psychology Dissonance / intrinsic & extrinsic motivation Plus some interest in economics (e.g Gneezy & Rustichini 2000 a & b) 3 level of payments for (0, low, high): Math tasks Collecting donations The results are V shaped and these were interpreted as incomplete contracts Hypotheses 1.2 Social-Market Effort level 1.0 0.8 0.6 0.4 Money-Market 0.2 0.0 No Payment Low Payment High Payment Level of payment Two-markets What can shift people from one market to the other? In cases when both social and money aspects are present, which will “win”? Hypothesis: Introducing monetary payments into a social exchange will cause individuals to shift from perceiving the exchange as a Social-Market to a Money-Market, and effort patterns will follow. Experiment 1 Hypothetical survey about helping to move a sofa Asking for willingness of other students to help on a11 point scale Design Form of payment Cash Low Middle $Candy -------- No Level of payment Candy $0.5 $5 $0.5 Candy bar candy bar Godiva box $5 Godiva box Results Exp1 Expected WTH 8 Money Candy $Candy Control 7 6 5 4 No Payment Low Payment High Payment Level of payment Conclusion Exp 1 All main predictions held in this hypothetical surveys Will they hold with real effort? Experiment 2 In the greatest tradition of social psychology using a mind numbingly boring task for 3 minutes The task 15 Design Form of payment Cash -------- No Level of payment Candy Low $0.1 5 JB Middle $4 1/2 lbs JB Results Exp2 250 Money Effort level (balls) Candy 200 Control 150 100 50 No Payment Low Payment High Payment Level of payment Conclusion Exp2 The cash and candy conditions support the “two-markets” hypothesis Candy Cash conditions Effort under no payment is above low cash payment but not above low candy payment Experiment 3 Testing the Cash vs. $Candy conditions In a domain of mental effort The task (1-4) Select a set of numbers that adds up to 100 Ӂ 19 Ӂ 20 Ӂ 26 Ӂ 27 Ӂ5 Ӂ 10 Ӂ 13 Ӂ 38 Ӂ 17 Ӂ 40 Ӂ 34 Ӂ 31 Current total =52 I give up The task (5) Select a set of numbers that adds up to 100 Ӂ 11 Ӂ 15 Ӂ 61 Ӂ 27 Ӂ 18 Ӂ42 Ӂ 57 Ӂ 3 Ӂ 30 Ӂ8 Ӂ 19 Ӂ 69 Current total =72 I give up Design Form of payment Cash -------- No Level of payment Low Middle $Candy $0.5 $0.5 candy bar $5 $5 Godiva box Results Exp3 Effort level (seconds) 450 400 Money $Candy Control 350 300 250 200 150 100 No Payment Low Payment High Payment Level of payment Conclusion Exp3 The similarity between the cash and $candy conditions suggest that mean mentioning $ is sufﬁcient to change the type of relationship from social to money markets No payment is a social market and thus higher in effort General discussion Mark Twain ends chapter 2 of Tom Sawyer by noting that: “If he (Tom) had been a great and wise philosopher, like the writer of this book, he would now have comprehended that work consists of whatever a body is obliged to do, and that play consists of whatever a body is not obliged to do.” He than continues and adds that “There are wealthy gentleman in England who drive fourhorse passenger-coaches twenty or thirty miles on a daily line in the summer, because the privilege cost them considerable money; but if they were offered wages for the service, that would turn it into work, and then they would resign.” Other examples Paying for help seem to dramatically change the nature of the help “I am not chagrinning you because if I would, you would not be able afford me” Paying for sex seem to dramatically change the nature of sex “The most expensive sex is free sex” -- Woody Allen “The big difference between sex for money and sex for free is that sex for money usually costs a lot less” -- Brendan Behan Summary Paying changes the nature of labor The currency of payment (and the link to effort) also inﬂuences the nature of labor Companies can strive to have a mix of social and money markets in their relationships The relationship between payment and motivation, effort, performance At high levels of payment Large Stakes & Big Mistakes Incentives Incentives are an important part of the labor market The basic assumption is that increased (decreased) payment for performance will cause individuals to work more (less) Non-performance based, and long term payments are more complex ... High incentives @ work Stock brokers: commission compensation + bonus P&G partner advertising agencies: payment-by-results National federations in soccer World Cup: payment-by-round Students at school: “payment”-by-evaluation Farm labor, Sales peoples, etc. Incentives & performance Incentives do not always behave as we would expect Decreasing incentives to 0 can increase effort (Lepper, Green & Nisbett 1973; Gneezy & Rustichini 2000; Heyman& Ariely 2004) What about increasing incentives? Can they be counterproductive? Under what conditions? What do people predict? Packing-quarters predictions Simon predictions 80% Earnings 80% Earnings 70% Very-Good 70% Very-Good 60% 60% 50% 50% 40% 40% 30% 30% 20% 20% 10% 10% 0% Low Mid High 0% Low Mid High of high incentives The “Yerkes-Dodson law” Experiment: rats had to learn to discriminate safe from unsafe areas in a cage. Performance showed an inverted U-shape relation between arousal (size of electronic shock) and learning Task X Task Y Arousal Arousal Performance Performance The “Yerkes-Dodson law” Higher stakes of high incentives Chocking under pressure Taking an exam Giving a talk Home teams: championship in baseball and basketball (Baumeister & Steinhilber 1984) Roll-up game (Baumeister 1984) All of these suggest a possible decrease in performance of high incentives Increased incentives 1 Effort 2 We assume that link 1 is correct, but question link 2 for very high incentives Performance Predictions High, but not moderate, incentives can be counterproductive and can produce a reduced level of performance Experiment setup A place that we could pay a substantial sum given our research budget ‫؞‬rural India Payment for performance on 7 tasks The population Average all-India monthly per capita consumer expenditure (MPCE) in rural areas: Rs 495 (approx. $10) TV: 49.4%; Telephone: 6.9% Transportation: 51.7% bicycle, no cars Education: 5.6 years, 26% no formal education Religion: 90.8% Hindu, 5.7% Christians, 3.4% Muslims Gender: 26.4% female, 73.6% male (87 people) Payment levels Performance VGood 4 rs Medium 0 rs 20 rs 40 rs High 0 rs 200 rs 400 rs DV1 = % of people with P>(VGood) DV2 = % of max possible payment Game Types Creativity Concentration Motor skills Bluffing ability Game 1 - creativity Packing Quarters fit 9 metal pieces into black frame as fast as possible Scoring Rule: 1 trial good: <= 240 sec very good: <= 120 sec Game 1 - results 80% Earnings 70% Very-Good 60% Packing-quarters 50% 40% 30% p=.337 p<.001, p=.020 p=.493 20% 10% 0% Low p<.001, p=.006 Mid High Game 2 - concentration Simon repeat sequence by pushing corresponding light-buttons in same order Scoring Rule: 10 trials good: 1 repetition of >= 6 consec. lights very good: 1 repetition of >= 8 consec. lights Game 2 - results 80% Earnings 70% Very-Good Simon 60% 50% p=.865 p=.014, p=.010 40% 30% 20% 10% 0% Low p=.178 p=.095, p=.005 Mid High Game 3 - concentration Recall last 3-digits experimenter reads sequences of digits, stops at an unannounced point. Participant has to recall the last 3-digits (e.g., 8,7,8,2,5,9,7,3) Scoring Rule: 14 trials good: >= 4 correct trials very good: >= 6 correct trials Game 3 - results 80% Earnings 70% Very-Good 60% Recall last 3 digits 50% p=.896 40% p=.157, p=.207 p=.638 30% 20% p=.180, p=.075 10% 0% Low Mid High Game 4 - bluffing Lying about a coin-flip flip a coin and send a signal to research assistant about the state of the coin. Research assistant has to guess the true state. If s/he doesn’t, participant gets 1 point. Scoring Rule: 10 trials good: >= 6 points very good: >= 8 points Game 4 - results 80% 70% 60% p=.629 p=.103, p=.022 p=.986 50% 40% 30% 20% 10% 0% Low p=.003, p=.004 Earnings Very-Good Lying about coin flip Mid High Game 5 - motor skills Labyrinth Pass the ball along the pathway avoiding the holes in the board from “start” to “finish” Scoring Rule: 10 trials good: 1 trial >= 7th hole very good: 1 trial >= 9th hole Game 5 - results 80% Earnings 70% Very-Good Labyrinth 60% 50% 40% p=.053 30% p=.176, p=.001 20% 10% 0% Low p=.042 p=.981, p=.044 Mid High Game 6 - motor skills Dart Ball Throw a velcro ball at the inflated target Scoring Rule: 20 trials good: >= 5 balls hitting the center very good: >= 8 balls hitting the center Game 6 - results 80% Earnings 70% Very-Good Dart-Ball 60% 50% 40% 30% p=.346, p=.602 p=.176 20% 10% 0% Low p=.206 p=.080, p=.619 Mid High Game 7 - motor skills Roll-Up Attempt to drop the ball into the highest possible slot by deftly spreading apart then pushing together the two rods Scoring Rule: 20 trials good: >= 4 balls hitting the furthest hole very good: >= 6 balls hitting the furthest hole Game 7 - results 80% Earnings 70% Very-Good Roll-Up 60% 50% 40% 30% 20% 10% 0% Low p=.887 p=.202, p=.166 p=.896 p<.050, p=.044 Mid High Results all (I) creativity concentration 80% Earnings 80% 70% Very-Good 70% 60% Packing-quarters Earnings Very-Good Simon 60% motor skill 80% Earnings 70% Very-Good 80% 70% Dart-Ball 60% blufﬁng 60% 50% 50% 50% 50% 40% 40% 40% 40% 30% 30% 30% 30% 20% 20% 20% 20% 10% 10% 10% 10% 0% Low Mid High 0% Low Mid High 0% Low Mid 80% Earnings 80% Earnings 70% Very-Good 70% Very-Good 60% Recall last 3 digits 60% 50% 50% 40% 40% 30% 30% 20% 20% 10% 10% 0% Low Mid 80% Earnings 70% Very-Good 60% High Labyrinth 50% 40% 30% 20% 10% 0% Low Mid High 0% Low High Roll-Up Mid High 0% Low Earnings Very-Good Lying about coin flip Mid High Results all (II) 80% Earnings 70% Very-Good All 60% 50% p=.919 p<.001, p<.001 40% 30% 20% 10% 0% Low p=.450 p<.001, p<.001 Mid Incentive level High Summary No obvious difference in pattern of performance across the different game types Except for 1 case (i.e. Labyrinth) there was no (marginally) significant difference in performance between low and mid payment conditions Performance always lowest in high payment condition when compared with low and mid payment conditions together Predictions? Can people predict this? The effect of very high incentives would be of no consequence if people know about it and avoid incentives that are too high for particular tasks & individuals The prediction study (I) We described the India study to 60 students Students had to predict the results for Simon & Packing Quarters: Fraction of participants who would reach P(Good) & P(VGood) in each of the 2 games & each of the 3 payment conditions Incentive: students were paid by accuracy of their prediction (max of $10) The prediction study (II) Payment method per set: set = game & performance level max. $2.50 per set Total Difference Payment 0-2 $2.50 3-5 $2.25 6-8 $2.00 9-11 $1.75 12-14 $1.50 15-17 $1.25 18-20 $1.00 21-23 $0.75 24-26 $0.50 27-29 $0.25 >=30 $0.00 Predictions: results (I) Students predicted that as reward increased participants in the original experiment would on average... be more likely to achieve P(VGood) receive a higher % of max possible payment Predictions: results (II) Packing-quarters predictions Simon predictions 80% Earnings 80% Earnings 70% Very-Good 70% Very-Good 60% 60% 50% 50% 40% 40% 30% 30% 20% 20% 10% 10% 0% Low Mid High 0% Low Mid Repeated Measure ANOVA results: Earnings: F(2, 42) = 51.328, p < .001; all 3 pairwise comparisons p < .001 Very-Good: F(2, 42) = 64.336, p < .001; all 3 pairwise comparison p < .001 High Predictions: implications Students do not seem to have an intuitive understanding of the possible negative effects of very high incentives Do others? Do HR experts understand this? Do companies set very high incentives for other purposes? Implications Providing incentives are generally costly for those providing them, raising contingent incentives beyond a certain point may be a losing proposition Other questions Can people get used to high incentives? How high is too high? Would breaking bonuses into many smaller bonuses help? What kind of tasks are more likely to have negative effects of very large incentives? Summary We often assume that higher incentives increase performance (perhaps in a diminishing returns) These results show that higher incentives can decrease performance Labor & meaning Why do people work? The standard view is that people exchange leisure for labor in order to get $. What are the implications of this view? Nothing else matters People should stop working once they have reached their level of optimal returns The meaning of labor Build Lego for pay ($3, $2.70 etc.) 2 conditions Sabotage Overall summary: Topics Relative vs absolute levels of compensations The relationship between payment and motivation, effort, performance Low payments, high payments Labor & meaning Sabotage Overall summary Labor is complex People work for many different motives and incentives Figuring out these motives can help making emplyees happier and more productive Labor is not more rational than other aspects of our life -- and it is important to ﬁgure it out