閤 The Fair wage-Effort Hypothesis and Unemployment OR。 George A Akerlof, Janet L. Yellen The Quarterly Journal of Economics, Vol. 105, No. 2.(May, 1990), pp. 255-283 Stable url: http://inksistor.org/sici?sic0033-5533%028199005%29105%03a2%03c255%03atfwhau%3e2.0.co%3b2-q The Quarterly Journal of Economics is currently published by The MIT Press Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at http://www.jstor.org/about/terms.htmlJstOr'sTermsandConditionsofUseprovidesinpartthatunlessyouhaveobtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the jsTOR archive only for your personal, non-commercial use Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at Each copy of any part of a JSTOR transmission must contain the same copyright notice that ap on the screen or printed page of such transmission STOR is an independent not-for-profit organization dedicated to and preserving a digital archive of scholarly journals. For more information regarding JSTOR, please contact support @jstor. org Thu mar1522:21:552007
The Fair Wage-Effort Hypothesis and Unemployment George A. Akerlof; Janet L. Yellen The Quarterly Journal of Economics, Vol. 105, No. 2. (May, 1990), pp. 255-283. Stable URL: http://links.jstor.org/sici?sici=0033-5533%28199005%29105%3A2%3C255%3ATFWHAU%3E2.0.CO%3B2-Q The Quarterly Journal of Economics is currently published by The MIT Press. Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at http://www.jstor.org/about/terms.html. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use. Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at http://www.jstor.org/journals/mitpress.html. Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission. JSTOR is an independent not-for-profit organization dedicated to and preserving a digital archive of scholarly journals. For more information regarding JSTOR, please contact support@jstor.org. http://www.jstor.org Thu Mar 15 22:21:55 2007
THE QUARTERLY JOURNAL OF ECONOMICS Vol. CV May 1990 Issue 2 THE FAIR WAGE-EFFORT HYPOTHESIS AND UNEMPLOYMENT* GEORGE A. AKERLOF AND jANET L. YELLEN This paper introduces the fair wage-effort hypothesis and explores its implica tions. This hypothesis is motivated by equity theory in social psychology and social exchange theory in sociology. According to the fair wage-effort hypothesis, worker proportionately withdraw effort as their actual wage falls short of their fair wage. Such behavior causes unemployment and is also consistent with observed cross- section wage differentials and unemployment patterns . INtROdUCTION This paper explores the consequences of a hypothesis concern g worker behavior, which we shall call the fair wage-effort tion of a fair wage; insofar as the actual wage is less than the fair wage, workers supply a corresponding fraction of normal effort. If e denotes effort supplied, w the actua fair wage-effort hypothesis says that =min(u/u*,1), where effort is denoted in units such that i is normal effort This We would like to thank Samuel Bowles, Daniel Kahneman under grant numbers SES 86-005023 and sEs 88-07807 administered by the titute for Business and Economic Research at the University of califo Akerlof and Yellen [ 1988] contains a summary of the results obtained in this e 1990 by the President and Fellows of Harvard College and the Massachusetts Institute of Technology. The quarterly Journal of Economics, May 1990
256 QUARTERLY JOURNAL OF ECONOMICS hypothesis explains the existence of unemployment. Unemploy ment occurs when the fair wage w* exceeds the market-clearing wage. With natural specifications of the determination of w*, this hypothesis may explain why skill and unemployment are negatively correlated. In addition, it potentially explains wage differentials and labor market segmentation. 3 The motivation for the fair wage-effort hypothesis is a simple observation concerning human behavior: when people do not get what they deserve, they try to get even. The next section will present five types of evidence for the fair wage-effort hypothesis. First, it will draw on psychology, where the fair wage-effort hypothesis corresponds to Adams'[1963] theory of equity. Numer- ous empirical studies have tested this theory. They are, on balance, strongly supportive. Second, in sociology the fair wage-effort hypoth esis corresponds to the Blau-Homans [1955, 1961] theory of social exchange. Sociological studies, including studies of work situations, how that equity usually prevails in social exchange Third, the fair wage-effort hypothesis accords with common sense. It appears frequently in literature; it is considered obvious by personnel extbooks; and it explains commonly observed taboos regarding iscussion of wages and salaries. Fourth, the fair wage-effort hypothesis explains wage compression among individuals with different skills. Fifth, simple models of the fair wage-effort hypothe sis potentially explain empirically observed unemployment-skill correlations; they also explain why unemployment has not fallen with the rise in education despite lower unemployment of more educated workers Having reviewed the evidence for the fair wage-effort hypothe sis, Sections III and IV construct models using this hypothesis These models differ in the determination of the fair wage w*. I Section III w* is exogenous. In Section Iv w* depends on relative wages as well as on market forces. These models provide efficiency wage explanations for unemployment. Yet they are not subject to the criticism that bonding schemes or complicated contracts wi reduce or eliminate involuntary unemployment. If such bonds are considered unfair then they will not be optimal In relations where fairness is important, grudges due to past events lead to potential 2. For evidence of d 4. For reviews of this literature and the problems with efficiency wage models, [1987], and Yellen [198
THE FAIR WAGE-EFFORT HYPOTHESIS future reprisals. In the existing literature this model most closely resembles Summers'[1988] relative wage-based efficiency wage theory. In Summers' model workers compare their own compensa tion with that of comparable groups in other firms; in our model, in contrast, workers compare their pay with that of coworkers in the same hrm II. MOTIVATION FOR THE FAIR WAGE-EFFORT HYPOTHESIS A. Equity Theory dams [1963] hypothesized that in social exchange between two agents the ratio of the perceived value of the the perceived value of the "outcomes"would be equal. In a labor exchange the"input "of the employee is the perceived value of his labor, and the"outcome"is the perceived value of his remunera tion. On the firms side the input is the perceived value of th remuneration, and the outcome is the perceived value of the labor. In the context of a wage contract, Adams formula says that the perceived value of the labor input will equal the perceived value of the remuneration This formula can be translated into economic notation to say that the number of units of effective labor input (denoted e for effort)times the perceived value of a unit of effective labor(denoted w*)will equal the perceived value of remuneration (denoted w). In other words We wish to emphasize that w*, the perceived value of a unit of labor, will be the fair wage, and not the market-clearing wage According to psychologists, with both w and w* fixed, workers who do not receive a fair wage for input of effort e= 1 may change actual effort e, or they may change their perceived effort. similarly, they may change their perceived level of remuneration(by rede- fining the nonpecuniary terms of the job). In the theory below, we shall assume that when wages are underpaid workers adjust actual rather than perceived efforts or the perceived value of the nonpecu nary returns to the job Psychological experiments have mainly concentrated on disco ering whether individuals who are overpaid will increase their effort input since psychologists consider this the surprising prediction of Adams'theory. They consider it obvious that agents who feel underrewarded will supply correspondingly fewer inputs [Walster, Walster, and Berscheid, 1977 p 42]. As might be expected, overre-
QUARTERLY JOURNAL OF ECONOMICS ents yield ambiguous results. It has been (Walster, Walster, and Berscheid, 1977, p 124] that this ambiguity occurs because it is less costly for overpaid agents to increase the psychological evaluation of their labor inputs than to increase actual input. These experimental results are consistent with the hypothesis that overpayment does not increase input, and thus that While much less work has been done on underpaid subjects, several studies have obtained supportive results. In one revealing study Lawler and O' Gara [1967] compared the performance of workers who were paid the"going " rate of 25 cents per interview with the performance of interviewers who were seriously underpan at the rate of 10 cents per interview. The underpaid interviewers conducted far more interviews that were on average of significantly lower quality. Psychologically the lower paid interviewers also had reduced self-esteem--suggesting that workers adjust not only the amount of effort but also their perception of the quality of the labor input when equity is not realized. In a clever experiment Pritchard, Dunnette, and Jorgenson [1972] hired men to work for a fictitious Manpower firm they ealistically set up for their experiment. After the workers had been at work for three days, the firm announced a change in their method of pay. Subjects'earnings were variously adjusted upward or downward. Those subjects with downward adjustments expresse considerable job dissatisfaction on a questionnaire and also per formed less well in their work after the change In a similar experiment Valenzi and Andrews [1971] hired workers at $1.40 per hour, but then announced that, due to the budgetary process involving their grant from the National Institute of Mental Health orkers would receive more than the stipulated $1.40, and some would receive less. Twenty-seven percent of those who were given the lower wage of $1.20 quit immediately-a result consistent with an upward sloping labor supply curve but also explained by the workers anger at their unfair treatment In what is probably the most revealing experiment, Schmitt and Marwell [1972] gave workers a choice: whether to work coopera tively in pairs or to work alone. When pay was equal, workers chose to work in pairs. However, workers were willing to sacrifice signif cant earnings to work alone when the pay in pairs was unequa consider this implication of equity theory obvious; some expe alerta tive e planations tradodtmoans and friede ry, but in all cases there are easy man,1971]
THE FAIR WAGE-EFFORT HYPOTHESIS 259 B Relative Deprivation Theory The economic consequences of the fair wage-effort hypothesis depend on how the fair wage is determined According to relative deprivation theory, peoples'conceptions of fairness are based on comparisons with salient others. Psychological theory, however, offers little guide as to which reference groups will be salient. There are three natural possibilities: individuals may compare themselves with others in similar occupations in the same firm, with those in dissimilar occupations in the same firm, or with individuals in other firms. In the model constructed in Section iv below workers compare themselves with others in the same firm. If workers compare themselves with similar others who are"close substitutes e find that equilibrium will be segregated and workers of different abilities will work in different firms. Labor is allocated inefficiently but there is no unemployment. If workers, however, compare themselves with others who are“ dissimilar”or“ complements”in production, equilibrium is characterized by unemployment for low-skill workers or by dual labor markets with pay disparities for low-skill workers Although the behavioral consequences of relative deprivation have been hard to document(for natural reasons), there is very good evidence that relative deprivation generates feelings of dissatisfac tion. This corresponds exactly to the model proposed in Section iv) Martin [1981] has done an ingenious experiment in a near-field situation which shows that workers are likely to experience feelings of relative deprivation when there are unequal wages. Technicians at a factory were asked to imagine themselves in the position of a technician earning the average pay in a firm similar to their own They were first asked which pay level-highest or lowest pay of technicians; highest, average or lowest pay of supervisors-they would most like to know for comparison to their own technicians wanted to know the pay of the highest level of techni ians--which is consistent with our model that people work less hard if they are paid less than they deserve but not harder if they receive more than they deserve Those people who receive less are of comparatively little interest (and therefore have little positive fluence on work); whereas those people who are paid more are of considerable interest and, if the ratio is deemed inequitable, can have considerable negative impact. 6. Most experiments make an implicit assumption regarding the wage consid ered fair: either some stated wage, a previously received wage or wages
260 QUARTERLY JOURNAL OF ECONOMICS The second part of Martin,s experiment is of further impor- tance for our model. After workers had made their comparison choice, they were then given a pay plan and asked to rate it on the basis of being dissatisfying, expected, or just. When the difference in pay of the supervisors and technicians was large the technicians found the pay levels to be dissatisfying and unjust. This gives an empirical basis for the assumption in Section I that low paid workers will feel relatively deprived when workers of other groups C Social Exchange Theory Sociologists, as well as psychologists, have developed a versie of equity theory. Blau' s model of exchange [1955] hypothesizes that there will be equivalent rewards net of costs on both sides of exchange Blau's model was motivated by his empirical study [1955 of the helping behavior of agents in a government bureaucracy. The agents who did investigative work would consult with other agents concerning difficult problems. Although consultation with other agents, rather than with the supervisor, was against the official rules of the agency, and its existence was denied by the supervisor,on average, agents had five contacts with other agents per hour, most of which were consultations In this agency agents varied in expertise Blau noticed that agents of average expertise would consult agents with the greatest expertise only infrequently. In contrast, agents of equal ability consulted with each other frequently. This suggested puzzle to Blau: why did the average agents not ask for more help from the experts? According to his explanation, the average agents refrained from consulting the experts more because they found difficult to reciprocate. They were able to pay each expert with gratitude and respect; but there were diminishing returns to the experts from receiving gratitude. The exchanges between the average agents and the experts, blau concluded were not carried beyond the point where the two sides of the exchange were of equal value Homans 1961] has proposed a similar theory, based on his own observations, Blau s study, and on work on conformity by social psychologists led by Festinger. The Blau-Homans theory is a general theory of social exchange Homans develops a key proposi tion regarding social exchange when the subjective equalities are not met on the two sides of an exchange: "The more to a man's disadvantage the rule of distributive justice fails of realization, the more likely he is to display the emotional behavior we call anger" [Homans, 1961, p. 75]. In simple English, if people do not get what
THE FAIR WAGE-EFFORT HYPOTHESIS they think they deserve, they get angry. It is this simple proposition that underlies our model. workers whose wage is less than the fair wage w* will be angry. The consequence of this anger is to reduce their effective labor input below the level they would offer if fully satisfied. This relation is given the simple, natural, functional form Sociologists have documented the existence of output restric tion in the workplace In his classic study of 1930 Mathewson [1969 ecords 223 instances of restriction in 105 establishments in 47 different locations These observations were recorded from his work experiences as a participant observer, interviews with workers, and om of six colleagues, who were also participant observers. According to Mathewson, "occasionally workers have an idea that they are worth more than management is willing to pay them. when they are not receiving the wage they think fair, they adjust their production to the pay received. This is an exact statement of the fair wage-effort hypothesis. The following, from the bulletin board of a machine shop, expresses the fair wage-effort hypothesis poetically: t the company is stealing Fifty pennies from my pocket every day; They will lose ten times as many g, i dare say For it makes one so disgusted So that nevermore my brow will drip with sweat When they' re in an awful hurry Someone else can rush and worry Till an increase in my wages do i get No malicious thoughts I harbor For the butcher or the barber who get eighty cents an hour from the start. Nearly three years I've been working Like a fool, but now I' m shirking When i get what's fair, I'll always do my part. Someone else can run their races Till I'm on an equal basis with the ones who learned the trade by mining coal Though I can do the work, it's funny New men can get the money And i cannot get the same to save my soul [Mathewson, 1969, p 127
262 QUARTERLY JOURNAL OF ECONOMICS In the introduction to the reprinted edition of Mathews Donald Roy, a sociologist known for his own worker participant observations of restriction in a machine shop relates a story from his own experience [1952]. A machine crew were discontent because of what they considered an unfair ratio between wages and profits A laminating machine in this factory apparently had extremely odd performance: it would operate perfectly for a long time and then go mysteriously awry. Sheets of heavy paper in the process of lamina- tion would suddenly tear and stick to the machine' s rollers, necessi- tating difficult and sticky work to unwrap the material. The crew operating the machine was putting too much stress on it, causing the paper to tear and stick. despite the necessity of cleaning the rollers(an unpleasant job relative to tending the working machine they considered this operation worthwhile to redress their griev ances [Roy, 1969, p. xxiv]. The preceding story illustrates that workers reduce their effective labor power if they feel they getting less than they deserve. It also indicates that they may feel that they deserve a wage higher than that required to induce them to be physically present at their jobs; further, the remuneration of dissimilar agents-in this case the profit earners--enters their calculation of their fair wage Studies by Mathewson and Roy are examples of the work of the uman relations school of organization. according to this school of thought, workers have considerable control over their own effort and output. This ability of workers to exercise control over their effort, and their willingness to do so in response to grievances, underlies the fair wage-effort hypothesis. A recent report in The New York Times [Salpukas, 1987] concerns the problems generated by two-tier wage systems. Despite the considerable savings in labor costs, many of the companies that adopted such systems are now phasing them out due to the resentment of employees on the job as well as the high turnover generated by the low wages. These wage systems have"produced a resentful class of workers who in some cases are taking their hostility out on s""[Salpukas, 1987, p. 1] " The attitude on the airplane can be a big problem, "said Pat A. Gibbs, the head of the Association of Professional Flight Attendants, which represents the tenants at American [Airlines] "You can tell that the anger is there. " Robert L. Crandall, American's chairman and chief executive, acknowledged in a recent speech that quality of service has suffered because of the pressures that deregulation has The lower-paid workers often do just what is required and no more, and sometimes refuse to help the higher-paid workers.. 'Having people work side by
THE FAIR WAGE-EFFORT HYPOTHESIS side for different pay is difficult said Mr. Olson of Giant Foods, about half of the supermarket chains workers are in the lower pay tier [Salpukas, 1987, p D22] E Literature, Jealousy, and retribution Jealousy and retribution, the relation between equity and performance are not recent discoveries of psychologists and sociolo- gists: they are part of everyone's experience Literature offers many xcellent examples, such as the story of Joseph [Bible, Genesis, 37-50]. Joseph's father, Jacob loved him more than all his children and made him a coat of many colors. When Joseph,s brothers saw that their father loved him most of all, they hated him. One day when Joseph was in the countryside they threw him into a pit, from which he was fortuitously rescued and sold into slavery. when Jacob heard of Joseph s presumed death, he wept inconsolably This sad story of Jacob, Joseph, and his brothers is an example of management failure made worse by inequitable rewards F. Personnel management Texts Textbooks on personnel management regard the need for equitable treatment of workers as obvious. By way of illustration Dessler[1984, p. 223] writes: The need for equity is perhaps the mos ant factor in determining pay rates Externally, pay must compare fave ith those in other organizations Ill find it hard to attract and retain employees Pay rates must also be uitable internally in that each employee should view his or her pay as equitable ven other employees pay rates in the organization.(Emphasis in last sentence added .) Kochan and Barocci, who view equity as most important in experts,”” inions of compensation systems, quote approvingly from a War Labor Board project(by William H. Davis: "There is no single factor in the whole field of labor relations that does more to break down morale create individual dissatisfaction, encourage absenteeism, increase labor turnover and hamper production than obviously unjust inequalities in the wage rates paid to different individuals in the same labor group within the same plant"[Kochan and Barocci, 1985, p. 249 Carroll and Tosi [1977, p. 303] write"Pay satisfaction is influenced by what an individual gets as compared to what he wan and considers fair. The fairness of pay(perceived equity of pay) determined largely by an individuals comparison of himself and his pay to other reference persons and theirs [sic]
