American Journal of Sociology

Date: 1954

Show Summary


Co-operation and Competition in a Bureaucracy1

. . . . . . . . . . . . . .

This paper discusses performance and variations in competitiveness among twelve interviewers in two small sections of a public employment agency. The duties of the interviewers in both sections were essentially alike. They received requests for workers over the phone. The order forms on which job openings were described were filed in a common pool in each section. Most of the official’s time was spent interviewing applicants for jobs. After ascertaining the client’s qualifications, the interviewer searched the sectional files for suitable vacancies. If an acceptable job was found, he referred the client to it and later phoned the employer to determine whether the client had been hired.

"The statistics which show how many interviews and how many placements each person in the section did are passed around to all interviewers. Of course, you look at them and see how you compare with others. This creates a competitive spirit," said one of the interviewers, voicing the sentiments of most of his fellows. In a period of job shortages, competition took the form of trying to utilize job openings before anybody else did. Interviewers were so anxious to make placements that they even resorted to illicit methods. Said one:

When you take an order, instead of putting it in the box, you leave it on your desk. There was so much hiding of orders under the blotter that we used to ask, "Do you have anything under your rug?" when we looked for an order. You might leave an order you took on the desk, or you might leave it on the desk after you made no referral.… Or, you might take an order only partially; you write the firm’s name, and a few things; the others you remember. And you leave it on the pad [of order blanks]. You keep on doing this, and all these orders are not in the box.

You can do some wrong filling out. For instance, for a rather low-salary job, you fill out "experience required." Nobody can make a placement on that except you, because you, alone, know that experience isn’t required. Or, if there are several openings [on one order], you put the order into "referrals" [file category for filled job openings] after you make one placement. You’re supposed to put it into "referrals" but stand it up, so that the others can see it. If you don’t, you have a better chance of making the next placement than somebody else. And time and again you see four, five openings on one order filled by the same person. [In one case on file eight out of nine openings on one order had been filled by the same interviewer.]

The major opportunity for competitive monopolization of job openings occurred when they were received from employers. Since illicit practices were concealed from the observer, the extent of competition could not be determined through questioning or direct observation2 but was betrayed by the record of official transactions. The extent to which an interviewer filled the vacancies he had received over the phone with his own clients in excess of chance expectations furnishes an index of competitiveness.…


The members of Section A were more competitive than those of Section B.… The interviewer’s competitiveness was related to his productivity in Section A (Pearsonian

), but this was not the case in Section B
. In other words, hoarding of jobs was an effective way to improve an interviewer’s placement record only in one of these two groups.

The members of Section B were more co-operative: they discouraged competitive practices by making them ineffective. When they learned about interesting vacancies, they often told one another, but an interviewer who manifested competitive tendencies was excluded from the network of reciprocal information and lost the respect of his co-workers. Any advantage of hoarding jobs was, at least, neutralized by such lack of co-operation, as is indicated by the absence of a relation between competitiveness and productivity in this group. Since competitive practices made an interviewer unpopular and failed to raise his productivity, they were infrequent.

These officials themselves attributed the greater competitiveness in Section A to the ambitiousness of several members: "There is usually one individual who starts it, who becomes a pace-setter. Once it has started, it is too late." The others, so interviewers claimed, have to follow suit. However, the most competitive member of Section A in recounting her reactions when production records were first introduced made it clear that this explanation of competition on the basis of personality characteristics is inadequate:

When they introduced statistics, I realized how fast I worked. I even wanted to drop lower. I didn’t mind working fast as long as it didn’t show, but when it showed up like that on the record, I wanted to work less. But you know what happened? Some of the others started to compete with each other and produced more than I did. Then I thought to myself, "Since I can do it, it’s silly to let them get ahead of me." I’m only human. So I worked as fast as before.

When statistical records made the superior performance of this interviewer public knowledge, she decided to work less, possibly in response to pressures the others had brought to bear upon her. While complaining about her unfair standards, however, the other members of the section also improved their own performance. Consequently, this interviewer, just like the others, felt constrained by colleagues to compete for an outstanding record. One or two members of Section B, on the other hand, were also accused of competitive tendencies, but their colleagues successfully discouraged their expression in monopolistic practices. It is in this sense that the competitive practices of one group and the co-operative practices of the other were social factors, calling for explanation in sociological rather than psychological terms…

Differential conditions affected the development of these two groups. First, the supervisor in Section A relied heavily on performance records in evaluating interviewers: "And here, in the production figures, is the answer to the question: How good are you? Here you see exactly how good the work you did was." Interviewers often mentioned the pressure thus exerted: "[Especially] around rating time, you get this competition. You don’t care whether the best person gets the job, but you try to make the placement yourself." In contrast, the new supervisor in Section B surprised his subordinates by rating them more leniently than they had expected, and not primarily on the basis of production records. Consequently, as one interviewer reported, "we became less anxious about statistics; another experience like that, and we might forget all about placement credit."

Second, a common professional orientation existed only in Section B. While the members of Section A had been assigned, and had received their training, at different times, the majority of those in Section B received their training together after World War II, at a time when intensive counseling had been stressed, since many returning veterans needed occupational advice. One official said of this period:

When I first came here, in May, 1946, we had a very nice bunch. It was like an all-day consultation; we discussed placements with each other all day long. At that time, the veterans came back, and there was a lot of emphasis on counseling. Nobody asked you how many placements you made, then. The emphasis was on quality, and we consulted with each other all day.

In this situation, the group developed a common professional code, which discouraged speedy placement as constituting defective employment service. In effect, this orientation transformed competitive practices from illegitimate means for desirable ends into illegitimate means for worthless ends. If such practices did occur, they were vigorously opposed on moral grounds as violating the interest of clients. Nevertheless, as will be shown presently, competition could not have been effectively curbed if the supervisor’s evaluation practice had engendered acute anxiety over productivity. However, the existence of this code would have made it difficult for the supervisor to judge performance mainly by productivity, since doing so would have stamped him as ignorant of the essentials of good employment service.

No opportunity for the development of a common professional code had existed in Section A. Since competitiveness prevailed in this group, the individual whose personal professional standards made him reluctant to compete either became the deviant whose productivity suffered or modified his standards and entered the race with the others.

Third, most members of Section A had been appointed to temporary civil service positions during World War II. They were on probation pending permanent appointments when production records were originally introduced and even afterward remained subject to layoffs due to reductions in staff. Their insecurity led them to strive to impress superiors with outstanding performance. In contrast, all but one of the members of Section B were veterans, whose employment could not be terminated except for cause. As one envious colleague put it, "They felt that nothing could happen to them, because they were veterans, and had super-seniority."

Differences in these three conditions—security of employment, opportunity for the development of a common professional orientation, and the evaluation practice of the supervisor—gave rise to two dissimilar social structures. Productivity was highly valued in Section A and became associated with the individual’s standing in the group, while striving for sheer productivity was disparaged in Section B. Thus, whereas the most productive and most competitive member of Section A was considered the best interviewer by her co-workers and was most popular with them, the most productive member of Section B was least respected and least popular. As a result of these structural differences, co-operative norms prevailed only in Section B.

The interviewers in both sections disliked working in a competitive atmosphere. A member of Section A said: "If I see that an interviewer keeps orders on her desk, I take them and put them in the box.… Of course, you don’t make friends that way." Since the majority in this section, including its most popular members, were highly competitive, to antagonize them was to threaten one’s own standing in the group. This deterred interviewers from discouraging competitive practices. Antagonizing a deviant, however, does not endanger one’s status. Consequently, since a striver was unpopular in Section B, its members could use sanctions freely to combat competitive practices and enforce co-operative norms.


… The group most concerned with productivity was less productive than the other group. Fifty-nine per cent of the job openings received in Section A were filled, in contrast to 67 per cent in Section B. (The 8 per cent difference is significant on the .01 level.) Another implicit paradox is that competitiveness and productivity were directly related for individuals in Section A but inversely related for the two groups.

Anxious concern with productivity induced interviewers in Section A to concentrate blindly upon it at the expense of other considerations. In their eagerness to make many placements they often ignored their relationships with others as well as official rules. Competitiveness in this group weakened social cohesion, while co-operativeness in Section B strengthened it. This difference is further shown by the fact that usually none of the members of Section A spent their rest periods together, whereas all but one of those of Section B, a newcomer when this study was being made, did. Social cohesion enhanced operating efficiency by facilitating co-operation and by reducing status anxiety.

Although the members of both groups had occasion to assist one another, greater effort was required to elicit such co-operation in Section A. The social interaction that occurred in the office during the twenty-four busiest hours of one week was recorded and classified as official and private contacts, that is, those directly concerned with a specific job or client, and all others. The frequency of an interviewer’s official contacts with colleagues was related to his productivity in Section A (rank correlation

) but not in Section B (rank correlation
). This suggests that only interviewers who kept, as one put it, "hopping around all the time" to retrieve job orders that others kept on their desks were able to make many placements in the competitive section. In the cohesive group, on the other hand, the co-operation needed for making placements occurred as a matter of course, and not only in response to special requests. This effort was not required for high productivity.

To maximize his placements, the interviewer in Section A hoarded jobs and simultaneously tried to prevent others from doing so, thereby antagonizing his co-workers, whose co-operation he needed if he was to do well. The members of this section therefore attempted to conciliate colleagues whom their competitive practices had alienated. Often, shortly after having interfered with her operations, an interviewer paid another a compliment about her work or her apparel. The most competitive interviewer was in the habit of taking time out to joke with her co-workers and was proud of making more placements than anybody else, "nevertheless." Actually, this compensating friendliness, which made her popular despite her competitiveness, helped her to be productive.

In Section A, interviewers had to make special efforts at conciliation in order to make placements, but this was not necessary in Section B. At least, this impression is corroborated by the finding that frequency of private contacts with others was also related to productivity in Section A (rank correlation

) but not in Section B (rank correlation
). The members of the cohesive group, whose operating practices did not put colleagues at a disadvantage, did not have to devote time and energy to solicit and encourage co-operation, since it was not extended reluctantly. Their spontaneous co-operation improved operating efficiency.

Social cohesion also lessened the status anxiety generated by the evaluation system. Such anxiety is most acute in the individual who does not feel integrated in his work group and therefore seeks to derive social recognition from excelling at his task and from approval of superiors. Friendly relations with co-workers made the standing of the individual in the cohesive group independent of his productivity, particularly since fast work was disparaged as a sign of superficial service. The consequent reduction of anxiety in the antiproductivity-oriented group actually raised its productivity.

Fluctuations in productivity illustrate the dysfunction of status anxiety. Section B had not always operated more efficiently than Section A. Its productivity had been lower during the two months preceding the last rating but had abruptly increased then, while that of Section A had declined…

The two groups found themselves in different situations before and after they were rated. The members of Section A were familiar with the rating standards of their supervisor, for she had rated them in previous years. Their anxiety led them to work especially hard immediately before the annual rating. The members of Section B, on the other hand, had never before been rated by their new supervisor. They were also concerned about their record but could not calm their anxiety by concentrating upon certain tasks, because they did not know what the supervisor would stress; the explanation he gave to his subordinates was too vague and adhered too strictly to official procedures to help them to foresee his actual practices. This unfocused anxiety was particularly detrimental to efficient performance. Later, when the interviewers found out that they were not rated primarily on the basis of statistical records, their anxiety largely subsided and their productivity increased. In contrast, the experience of the members of Section A, whose rating was strongly influenced by their production records, intensified their status anxiety, but, when the rating was over, anxiety was no longer channeled into exceptionally hard work, with the result that their productivity declined below that of Section B.

Social cohesion is no guaranty against anxiety in a bureaucracy. Civil service status is too important to officials for them to remain immune to the threat of losing it. But when no such threat is felt, social cohesion reduces anxiety by divesting productivity of its significance as a symbol of status in the work group. Diminished anxiety as well as smoother co-operation then enables those in the cohesive group to perform their tasks more efficiently than the others.

In the absence of social cohesion, competitive striving for an outstanding performance record became a substitute means for relieving status anxiety in Section A. This psychological function of competition is illustrated by the following incident: The interviewers in this section became very irritable, and one of them even became physically ill, when a temporary supervisor, who tried to prevent competitive practices, interfered with their method of allaying anxiety. Status anxiety reduced operating efficiency. Even in the cohesive group, productivity was low when the unknown rating standards of a new supervisor produced acute and diffuse anxiety. Otherwise, however, the cohesive group was more productive, because social cohesion relieved status anxiety by making the individual’s standing in the group independent of his productivity. The very competitive striving that undermined the group’s cohesiveness also served to lessen the individual’s status anxiety in a noncohesive situation. The hypothesis that the cohesiveness of the group and the competitiveness of the individual in the less cohesive group both reduce status anxiety explains the paradox that the less competitive group as well as the more competitive individual in the competitive group each was particularly productive.

1 From , 1954, 59:530–535. By permission of The University of Chicago Press.

2 This is clearly indicated by the comment of one of a group of special interviewers, who were expected to use the job openings of the regular interviewers but usually had great difficulty in doing so: "Oh, they hide everything from us. We got more orders when you [the observer] sat in the middle of that section than ever before. We laughed about it. Interviewers would hand us orders asking whether we could use them—when you were looking. That had never happened before."


Related Resources

None available for this document.

Download Options

Title: American Journal of Sociology

Select an option:

*Note: A download may not start for up to 60 seconds.

Email Options

Title: American Journal of Sociology

Select an option:

Email addres:

*Note: It may take up to 60 seconds for for the email to be generated.

Chicago: "Co-Operation and Competition in a Bureaucracy1," American Journal of Sociology in Principles of Sociology: A Reader in Theory and Research, ed. Young, Kimball, and Mack, Raymond W. (New York: American Book Company, 1962), Original Sources, accessed May 27, 2022,

MLA: . "Co-Operation and Competition in a Bureaucracy1." American Journal of Sociology, Vol. 59, in Principles of Sociology: A Reader in Theory and Research, edited by Young, Kimball, and Mack, Raymond W., New York, American Book Company, 1962, Original Sources. 27 May. 2022.

Harvard: , 'Co-Operation and Competition in a Bureaucracy1' in American Journal of Sociology. cited in 1962, Principles of Sociology: A Reader in Theory and Research, ed. , American Book Company, New York. Original Sources, retrieved 27 May 2022, from