Power isthe great motivator

David C. McClelland and David H, Burnham

The drive for achievement that typically motivates thefounder-entrepreneur can seriously handicap the manager.Their research has persuaded the authors of this McKinseyAward-winning article that the prize sought by the reallyeffective manager is much more likely to he power andinfluence. Properly oriented, the power-motivated executivewill delegate more effectively, command greater respect, andget far hetter performance from subordinates than theobsessive achievers on top management's list of potentialpromotion candidates.

What makes or motivates a good manager? The question is soenormous in scope that anyone trying to answer it has difficultyknowing where to begin. Some people might say that a good manageris one who is successful; and by now most business researchers andbusinessmen themselves know what motivates peoi)Ie who success-fully run their own small businesses. The key to their success hasturned out to be what psychologists call "the need for achieve-ment," the desire to do something better or more efficiently thanit has been done before.

Any number of books and articles summarize research studiesexplaining how the achievement motive is necessary for a personto attain success on his own."̂ But what has achievement motiva-tion got to do with good management? There is no reason ontheoretical grounds why a person who has a strong need to be moreefficient should make a good manager. While it sounds as if every-one ought to have the need to achieve, in fact, as psychologistsdefine and measure achievement motivation, it leads people tobehave in very special ways that do not necessarily lead to goodmanagement.

" Eor instance, see David C. McCiellimd, The Achieving Society, New York VanNostrand. 1961, and David C. McClelland and David Winter, Motivating EconomieAchievement. New York, Eree Press, 1969.

SUMMER 1977 27

For one thing, because they focus on personal improvement, ondoing things better by themselves, achievement-motivated peoplewant to do things themselves. For another, they want concreteshort-term feedback on their performance so that they can tell howwell they are doing. Yet a manager, particularly one of or in a largecomplex organization, cannot perform all the tasks necessary forsuccess by himself or herself. He must manage others so that theywin do things for the organization. Also, feedback on his subordi-nate's performance may be a lot vaguer and more delayed than itwould be if be were doing everything himself.

The manager's job seems to call more for someone who can in-fluence people than for someone who does things better on his own.ln motivational terms, then, we might expect the successfulmanager to have a greater ''need for power" than need to achieve.But there must be other qualities beside the need for power thatgo into the makeup of a good manager. Just what these qualitiesare and bow they interrelate is the subject of this article.

To measure the motivations of managers, good and bad, we studieda number of individual managers from different large US corpora-tions who were participating in management workshops designedto improve their managerial effectiveness. (The research methodsand workshop techniques are described on pages 32 33.)

The general conclusions of these studies is that the top manager ofa company must possess a high need for power, that is, a concernfor influencing people. However, this need must be disciplined andcontrolled so that it is directed toward the benefit of the institutionas a whole and not toward the manager's personal aggrandizement.Moreover, the top manager's need for power ought to be greaterthan his need for being liked by people.

Now let us look at wbat these ideas mean in the context of realindividuals in real situations and see what comprises the profile ofthe good manager. Later, we will look at the workshops themselvesto determine how they go about changing behavior.

The Peter principleFirst off, what does it mean when we say that a good manager hasa greater need for "power" than for "achievement?" To get a moreconcrete idea, let us consider the case of Ken Briggs, a salesmanager in a large US corporation who joined a managementworkshop. Some six or seven years ago. Ken Briggs was promoted

THK McKINSEY QUARTERLY

to a managerial position at corporate headquarters, where he hadresponsihility for the salesmen who service his company's largestaccounts.

In filling out his questionnaire. Ken showed that he correctly per-ceived what his job required of him, namely, that he should in-fiuence others' success more than achieve new goals himself orsocialize witb his subordinates. However, when asked to write astory depicting a managerial situation. Ken unwitf.ingly revealedthrough his fiction that he did not really share those concerns.Indeed, he discovered that his need for achievement was very high -in fact over the 90th percentil<; - and his need for power very low,in about the 15th percentile. Ken's high need to achieve was nosurprise - after all, he had been a very successful salesman - butobviously his motivation to influence others was much less thanhis job required. Ken was a little disturbed but thought that per-haps the measuring instruments were not too accurate and thatth(̂ gap between the ideal and his score was not as great as itseemed.

Then came the real shocker. Ken's subordinates confirmed whathis .stories revealed: he was a poor manager, having little positiveimpact on those who worked for him. Ken's subordinates felt thatthey had little responsibility delegated to them, that he never re-warded but only criticized them, and that the office was not wellorganized, but confused and chaotic. On all three of these scales,his office rated in the 10th to 15th national percentiles,

As Ken talked the results over privately with a workshop leader,he became more and more upset. He finally agreed, however, thatth(.' results of the survey confti-med feelings he had been afraid toadmit to himself or others. For years, he had been miserable m hismanagerial role,. He now knew the reason: he simply did not want,nor had he heen ahle, to influence or manage others. As he thoughtback, he realized that he had failed every time he had tried toinfluence his staff, and he felt worse than ever.

Ken bad responded to failure by setting very high standards - hisoffic(' scored in the 98th percentile on this scale - and by trying todo most things himself, which was close to impossible; his ownactivity and lack of delegation consequently left, his staff de-moralized. Ken's experience is typical of those who have a strongneed to achieve hut low power motivation. They ma}' become verysuccessful salesmen and, as a consequence, may be promoted intomanagerial jobs for which, ironically, they are unsuited.

SUMMER ]977 29

If achievement motivation does not make a good manager, whatmotive does? It is not enough to suspect that power motivationmay be important; one needs hard evidence that people who arebetter managers than Ken Briggs do in fact possess stronger powermotivation and perhaps score higher in other characteristics aswell. But how does one decide who is the better manager?

Managing for morale

Real-world performance measures are hard to come by if one istrying to rate managerial effectiveness in production, marketing,finance, or research and development. In trying to determine whothe better managers were in Ken Briggs's company, we did not wantto rely only on the opinions of their superiors. For a variety ofreasons, superiors' judgments of their subordinates' real-worldperformance may be inaccurate. In the absence of some standardmeasure of performance, we decided that the next best index of amanager's effectiveness would be the climate he or she creates inthe office, reflected in the morale of subordinates.

Almost by definition, a good manager is one who, among otherthings, helps subordinates feel strong and responsible, who re-wards them properly for good performance, and who sees thatthings are organized in such a way that subordinates feel theyknow what they should be doing. Above all, managers shouldfoster among subordinates a strong sense of team spirit, of pride inworking as part of a particular team. If a manager creates andencourages this spirit, his subordinates should perform better.

In the company Ken Briggs works for, we have direct evidence ofa connection between morale and performance in the one areawhere performance measures are easy to come by - namely, sales.In April 1973, at least three employees from each of this company's16 sales districts filled out questionnaires rating their office fororganizational clarity and team spirit (see pages 32-33). The scoreswere averaged and totaled to give an overall morale score for eachoffice. The percentage gains or losses in sales for each district in1973 were compared with those for 1972. The difference in salesflgures by district ranged from a gain of nearly 30 percent to a lossof 8 percent, with a median gain of around 14 percent. Exhibit Ishows the average gain in sales performance plotted against theincreasing averages in morale scores.

From Exhibit I we can see that the relationship between sales andmorale is surprisingly close. The six districts with the lowest

THE McKINSEY QUARTERLY

I xnibit I Correlation between morale and sales performancein one company

30% average gain in sales by district, 1972-3

2 districts!

25

20

15 4 districts)

10

4di5trictsf

BdistrictsT j

0

Morale 0 lOl 2ol 3ol 40 iscore perceived organizational clarity plus team spirit

morale early in the year showed an average sales gain of onlyaround 7 percent by year's end (although there was wide variationwithin this group), whereas the two districts with the highestmorale showed an average gain of 28 percent. When morale scoresrise above the 50th percentile in terms of national norms, theyseem to lead to better sales performance. In Ken Briggs's company,at least, a high level of morale at the beginning is a good indica-tion of how well the sales division actually performed during thecoming year.

And it seems very likely that the manager wbo can create highmorale among salesmen can also do the same for employees in otherareas (production, design, and so on), leading to better per-formance. Given that high morale in an office indicates that thereIS a good manager present, what general characteristics does hepossess?

SUMMER 1977

In examining the motive scores of over 50 managers of both highand low morale units in all sections of the same large company, wefound that most of the managers - over 70 percent - were high inpower motivation compared with men in general. This findingconfirms the fact that power motivation is important for manage-ment. (Remember that as we use the term "power motivation," itrefers not to dictatorial hehavior, but to a desire to have impact,to be strong and influential.) The hetter managers, as judged bytbe morale of those working for them, tended to score even higherin power motivation. But the most important determining factor ofhigh morale turned out to be not how their power motivationcompared to their need to achieve but whether it was high(;r thantheir need to be liked. This relationship existed for 80 percent ofthe better sales managers. And the same held true for othermanagers in nearly all parts of the company.

In the research, product development, and operations divisions,73 percent of the better managers had a stronger need for power

Research methods and workshop techniques

The case studies and data used in this article aro derived from a number ofworkshops attended by some 500 manaj^ers from over 25 US corporations whocame to learn ahout their managerial styles and abilities and how they mightbe improved. The workshops also provided us with an opporLuTiity to studywhich motivation pattern - a concern for achievement, people, or power, or acombination of these makes tho best managers. The illustrative Exhibits arebased on data drawn from just ono of tho companies reprtisentiKl.

Terminology. Tho motive labels used refer to well-dt^fined characteristicsmeasured by coding an individual's spontaneous reactions for the frequencywith which he or she is concerned to (1) d(J something bettc;r or more elKcienlly(need for achievement), (2) establish or maintain good relations with others(need for affiliation), or (3) have impact on others (need for power).

Motivation. When the managers first arrived at the workshops, they fill(̂ d outa questionnaire analyzing their job and explaining what they thouj^'ht itrequired of them, 'l^hey were also asked to write a numher of stories to picturesof various work situations, which were coded for the extent to which thewriter was concerned ahout achievement, affiliation, or power, as well as fortho amount of inhibition or self-control rovealed. The results were thenmatched against national norms. A comparison of job requirements andmotivational patterns can often help in assessing whetbor an individual is mthe right job, is capable of adjusting to that joh. or is a suitable candidate forpromotion.

THK MCKINS1':Y QUARTERLY

than a need to be liked (or wbat we term "affiliation motive") ascompared with only 22 percent of the poorer managers. Why sbouldthis be so? Sociologists have long argued that, for a bureaucracyto function effectively, those who manage it must be universalisticin applying rules. That is, if they make exceptions for the particularneeds of individuals, the whole system will break down.

The manager with a high need for heing liked is precisely the onewho wants to stay on good terms with everybody, and, therefore,is the one most likely to make exceptions in terms of particularneeds. If a male employee asks for time off to stay home with hissick wife to help look after her and the kids, the affiliative manageragrees almost without thinking, because he feels sorry for the manand agrees that his family needs him.

When President Ford remarked in pardoning ex-President Nixonthat he had "suffered enough," he was responding as an affiliativemanager would, because he was empathizing primarily with

Style. To ascertain participants' managerial style we asked tbem to respondto various realistic work situations in office settings. Their answers werecoded for six different managerial styles: democratic, affiliative, pace-setting,coaching, coercive and authoritarian. The managers were asked to commenton the effectiveness of each style and say which ono they preferred.

Effectiveness. To determine how effective managers were, we surveyed atloast three subordinates of each manager, using questions designed to revealtheir supervisors' characteristics along six dimensions: (1) degree of con-formity required; (2) amount of rosponsihility delegated; (3) standard ofperformance required; (4) degree to which rewards are given for good work asopposed to sanctions when things go wrong; (5) degree of organizationalclarity: (6) team spirit in the office.* The managers who received the highestmorale scores (organizational clarity plus team spirit) from their suhordinateswere deemed to bo the bost, possessing the most desirahle motive patterns.

Maturity. Participants were assessed for one further attribute deemed import-ant for good management: maturit;-. Scores were obtained for four stages ofmaturity by coding managers" stories for thoir attitudes toward authority andthe kinds of emotions they displayed over specific issues. People in stage 1depend on others; those in stage 2 are interested primarily in controllingthemselves: those in stage 3 like to manipulate others; and those in stage 4are altruistic.

re. Now

SUMMER 1977 33

Nixon's needs and feelings. Sociological theory and our data bothargue, however, that the person whose need for affiliation is highdoes not make a good manager. This kind of person creates poormorale because he or she does not understand that other people inthe office will tend to regard exceptions to the rules as unfair tothemselves, just as many US citizens felt it was unfair to letRichard Nixon off and punish others less involved than he was inthe Watergate scandal.

But so far our findings are a little alarming. Do they suggest thatthe good manager is one who cares for power and is not at allconcerned about the needs of other people? Not quite, for the goodmanager has other characteristics which must also be taken intoaccount.

Socialized power

Above all, the good manager's power motivation is not orientedtoward personal aggrandizement but toward the institution whichhe or she serves. In another major research study, we found thatthe signs of controlled action or inhibition that appear when aperson exercises his or her imagination in writing stories tell agreat deal about the kind of power that person needs.* We dis-covered that, if a high power motive score is balanced by highinhibition, stories about power tend to be altruistic. That is, theheroes in the story exercise power on behalf of someone else. Thisis the "socialized" face of power as distinguished from the concernfor personal power, which is characteristic of individuals whosestories are loaded with power imagery but which show no sign ofinhibition or self-control.

In our earlier study, we found ample evidence that these latterindividuals exercise their power impulsively. They are more rudeto other people, they drink too much, they try to exploit otherssexually, and they collect symbols of personal prestige such asfancy cars or big offices. Individuals high in power and in control,on the other hand, are more institution minded; they tend to getelected to more offices, to control their drinking, and to want toserve others. Not surprisingly, we found in the workshops that thebetter managers in the corporation also tend to score high on hothpower and inhibition.

* David C. McClelland, William N. Davies, Rudolf Kalin and Kric Warner, TheDrinking Man, New York, The Free Press, 1972.

THE McKINSEY QUARTERLY

Average scores of managers according to motive profile

Percenlile ranking of average scores (national norms)0 10 20 30 40 50 60

Sense of responsibility

Scoresfrom at leastthree subordinates oft ft '{I Affiliaiive managersL k J (affiliation greater llian power, high inhibition)

•Personai power managers(power greater tiian affiliation, low mhibition)

[_^3| Institutional managersIZUI (power greater than affilialion, high inhibition)

Let us recapitulate what we have discussed so far and have illus-trated with data from one company. The better managers we studiedare high in power motivation, low in affiliation motivation andhigh m inhibition. They care about institutional power and use itto stimulate their employees to be more productive. Now let uscompare them with affiliative managers - those in whom tbe needior affiliation is higher than the need for power - and with thepersonal power managers - those in whom the need for power ishigher than for affiliation but whose inhibition score is low.

Motivational permutations

In the sales division of our illustrative company, tbere weremanagers who matched all three types fairly closely. Exhibit IIshows how their subordinates rated the offices they worked in onresponsibility, organizational clarity and team spirit. There are

SUMMER 197735

scores from at least three subordinates for each manager, andseveral managers are represented for each type, so that the averagesshown in the Exhibit are quite stable. Note tbat the manager whois concerned about being liked by people tends to have sub-ordinates who feel that they have very little personal responsibility,that organizational procedures are not clear, and tbat they havelittle pride in their work group.

In short, as we expected, affiliative managers make so many adhominem and ad hoc decisions that they almost totally abandonorderly procedures. Their disregard for procedure leaves em-ployees feeling weak, irresponsible, and without a sense of whatmight happen next, of where they stand in relation to theirmanager, or even of what they ought to be doing. In this company,the group of affiliative managers portrayed in Exhibit II werebelow the 30th percentile in morale scores.

The managers who are motivated by a need for personal power aresomewhat more effective. They are able to create a greater senseof responsibility in their divisions and, above all, a greater teamspirit. They can be thought of as the managerial equivalents ofsuccessful tank commanders such as General Patton, whose owndaring inspired admiration in his troops. But notice how in ExhibitII these men are still only in the 40th percentile in the amount oforganizational clarity they create, as compared to the high power,low affiliation, high inhibition managers, whom we shall term''institutional."

Managers motivated by personal power are not disciplined enoughto be good institution builders, and often their subordinates areloyal to them as individuals rather than to the institution they bothserve. When a personal power manager leaves, disorganizationoften follows. The subordinates' strong group spirit, which themanager has personally inspired, deflates; they do not know whatto do for themselves.

Of all the managerial types, the ''institutional" manager is themost successful in creating an effective work climate. Exhibit IIshows that the subordinates of an institutional manager feel thatthey have more responsibility. Also, this kind of manager createshigh morale because he or she produces the greatest sense oforganizational clarity and team spirit. If such a manager leaves,he or she can be more readily replaced by another manager, be-cause the employees have been encouraged to be loyal to the insti-tution rather than to a particular person.

THE McKINSEY QUARTERLYJO

Since it seems undeniable from Exhibit II that either kind of powerorientation creates better morale in subordinates than a peopleorientation, we must consider that a concern for power is essentialto good management. Our findings seem to fly in tbe face of a longand influential tradition of organizational psychology, which in-sists that authoritarian management is wbat is wrong with mostbusinesses in this country. Let us say frankly that we think thebogeyman of authoritarianism has in fact been wrongly used toplay down the importance of power in management. After all,management is an influence game. Some proponents of demo-cratic management seem to have forgotten this fact, urgingmanagers to be primarily concerned with people's human needsrather than with helping them to get things done.

Managerial styles

But a good deal of the apparent confiict between our findings andthose of other behavioral scientists in this area arises from the factthat we are talking about motives, and behaviorists are often talk-ing about actions. What we are saying is that managers must beinterested in playing the influence game in a controlled way. Thatdoes not necessarily mean that they are or should be authoritarianm action. On the contrary, it appears that power-motivatedmanagers make their suhordinates feel strong rather than weak.The true authoritarian in action would have the reverse effect, mak-ing people feel weak and powerless.

Thus another important ingredient in the profile of a manager ishis or her managerial style. In the illustrative company, 63 percentof the better managers (those whose subordinates had highermorale) scored higher on the democratic or coaching styles ofmanagement as compared with only 22 percent of the poorermanagers, a statistically significant difference. By contrast thelatter scored higher on authoritarian or coercive, managementstyles. Smce tbe better managers were also higher in power moti-vation. It seems that, m at^tion, they express tbeir power motivationm a democratic way, which is more likely to he effective.

To see how motivation and style interact, let us consider the caseol George Prentice, a manager in the sales division of anothercompany. George had exactly the right motive combination to bean institutional manager. He was high in the need for power lowm tbe need for affiliation, and high in inhibition. He exercised hispower m a controlled, organized way. His stories reflected thistact, in one, for instance, he wrote: ''The men sitting around the

SUMMER 197737

table were feeling pretty good; they had just finished plans forreorganizing the company; the company has been beset with anumber of organizational problems. This group, headed by ahard-driving, brilliant young executive, has completely reorganizedthe company structurally with new jobs and responsibilities . . . "

This described how George himself was perceived by the company,and shortly after attending the workshop he was promoted to vicepresident in charge of all sales. But George was also known to hiscolleagues as a monster, a tough guy who would ''walk over hisgrandmother" if she stood in the way of his advancement. He hadthe right motive combination and, in fact, was more interested ininstitutional growth than in personal power. But his managerialstyle was all wrong. Taking his cue from some of the top executivesin the corporation, he told people what they had to do andthreatened them with dire consequences if they didn't do it.

When George was confronted with his authoritarianism he recog-nized that this style was counterproductive - in fact, in anotherpart of the study we found that it was associated with low morale -and he subsequently changed to acting more like a coach, whichwas the scale on which he scored the lowest initially. George sawmore clearly that his job was not to force other people to do thingsbut to help them to figure out ways of getting their job done hetterfor the company.

The institutional manager

One reason it was easy for George Prentice to change his man-agerial style was that in his imaginative stories he was alreadyhaving thoughts about helping others, characteristic of men withthe institution-building motivational pattern. In further examininginstitution builders' thoughts and actions, we found they havefour major characteristics:

•1 They are more organization-minded. That is, they tend to joinmore organizations and to feel responsible for huilding up theseorganizations. Furthermore, they believe strongly in the import-ance of centralized authority.

II They report that they like to work. This flnding is particularlyinteresting, because our research on achievement motivationhas led many commentators to argue that achievement motivationpromotes the "Protestant work ethic." Almost the preciseopposite is true. People who have a high need to achieve hke to

THE McKTNSEY QUARTERLYOO

get out of work by becoming more efficient. Tbey would like tosee the same result obtained in less time or with less effort. Butmanagers who have a need for institutional power actually seemto like the discipline of work. It satisfles their need for gettingthings done in an orderly way.

^ They are more altruistic. They seem quite willing to sacriflcesome of their own self-interest for the welfare of the organizationthey serve. For example, they are more willing to make con-tributions to charities.

• They have a keen sense of justice. It is almost as if they feel thatif a person works hard and sacrifices for the good of the organ-ization, he or she sbould and will get a just rewaixl for the effort.

Tt is easy to see how each of these traits helps a person become agood manager, concerned about what the institution caa achieve.

And there is one more attribute we discovered in studying thebetter managers at George Prentice's company. They were moremature. Mature people can be most simply' described as lessegotistic. Somehow their positive self-image is not at stake in whatthey are doing. They are less defensive, more willing to seek advicefrom experts, and have a longer-range view. They accumulatefewer personal possessions and seem older and wiser. It is as if theyhave awakened to the fact that tbey are not going to live foreverand have lost some of the feeling that their own personal future isall that important.

Many US businessmen fear this kind of maturity. They suspect thatIt will make tbem less hard driving, less expansion-minded, andless committed to organizational effectiveness. Our data do notsupport tbeir fears. These fears are exactly the ones George Prenticehad hefore attending the workshop. Afterward he was a moreeffective manager, not despite his loss of some of the sense of hisown importance, but because of it. The reason is simple- his sub-ordinates believed afterward that he genuinely was more concernedabout the company than about himself. Where once tbey respectedhis confidence but feared him, they now trust him, Once he sup-ported their image of him as a "big man" by talking about the newPorsche and the new Honda he bad bought; when we saw himrecently he said, almost as an aside: "I don't huy things anymore."

George Prentice was able to change his managerial style afterlearning more about himself m a workshop. But does self-know-

SUMMER 1977

Fxhibit Average scores before and after training

Percentile ranking of average scores (national norms)0 10 20 30I \

40 50 60

Sense of responsibilityI

Rewards receiv ed

Organizational clarity

Team spirit

1 -,—..•... , 1 1. - 1 ' i .•<

: 1. **•.-•

1 1

Scores from over 50 salesmen ;[* ̂ 1 Before managertraining

R T I After managertraining

ledge generally improve managerial behavior? Some people mightask: "What good does it do to know, if I am a manager, that Ishould have a strong power motive, not too great a concern aboutbeing liked, a sense of discipline, a high level of maturity, and acoaching managerial style? What can I do about i t?" The answer isthat giving managers analytical information about themselves m asupportive setting can enable them to change.

Changing managerial styleGonsider the results shown in Exhibit III, which compares "before"and "after" scores. Once again we use the responses of sub-ordinates to give some measure of the effectiveness of managers.To judge by their subordinates' responses, the managers wereclearly more effective afterward. The subordinates felt that theywere given more responsibility, that they received more rewards,that the organizational procedures were clearer, and that moralewas higher. These differences are all statistically significant.

40THE McKINSEY QUARTERLY

But what do these differences mean in human terms? How did themanagers change? Sometimes they decided they should get intoanother line of work. This happened to Ken Briggs, for example,who found that the reason he was doing so poorly as a managerwas because he had almost no interest in influencing others. Heunderstood how he would have to change if he were to do well inhis present job, but in the end decided, witb the help of manage-ment, that he would prefer to work back into his first love, sales.

Ken Briggs moved into "remaindering," to help retail outlets forhis company's products get rid of last year's stock so that theycould take on each year's new styles. He is very successful in thisnew role; he has cut costs, increased dollar volume, and in timebas worked himself into an independent role selling some of the oldstock on his own in a way that is quite satisfactory to the business.And he no longer has to manage anybody.

in George Prentice's case, less change was needed. He was ob-viously a very competent person with the right motive profile for atop managerial position. When he was promoted., he performedeven more successfully than before because he realized the need tobecome more positive in his approach and less coercive in hismanagerial style.

But what about a person wbo does not want to change his job anddiscovers that he does not have the right motive i)rofile to be amanager?

The case of Gharlie Blake is instructive. Charlie was as low inpower motivation as Ken Briggs, his need to achieve was aboutaverage, and his affiliation motivation was above average. Thus hehad the affiliative manager profile, and, as expected, the moraleamong his subordinates was very low. When Charlie learned thathis subordinates' sense of responsibility and perception of a re-ward system were in the 10th percentile and that team spirit was intho 30th, be was shocked. When shown a film depicting threemanagerial climates, Charlie said he preferred what turned out tobo tbe authoritarian climate. He became angry when tbe workshopleader and other members in the group pointed out the limitationsof this managerial style. He became obstructive in the group pro-cess and objected strenuously to what was being taught.

In an interview conducted much later, Charlie said: "I blew mycool. When I started yelling at you for being all wrong, I got evenmadder when you pointed out that, according to my style question-

SUMMKR1977

naire, you bet that that was just what I did to my salesmen. Downunderneath I knew something must be wrong. The sales perform-ance for my division wasn't so good. Most of it was due to meanyway and not to my salesmen. Obviously their reports that theyfelt very little responsibility was delegated to them and that Ididn't reward them at all had to mean something. So I finallydecided to sit down and try to figure what I could do about it. Iknew I had to start being a manager instead of trying to do every-thing myself and blowing my cool at others because they didn'tdo what I thought they should. In the end, after I calmed down, Irealized that it is not so bad to make a mistake; it's bad not tolearn from it."

After the course, Charlie put his plans into effect. Six monthslater, his subordinates were asked to rate him again. He attended asecond workshop to study these results and reported: "On the wayhome I was very nervous. I knew I had been working with thoseguys and not selling so much myself, but I was very much afraid ofwhat they were going to say about how things were going in theoffice. When I found out that the team spirit and some of those otherlow scores had jumped from around 30th to the 55th percentile, Iwas so delighted and relieved that I couldn't say anything all daylong."

When he was asked how he acted differently from hefore, he said:"In previous years when the corporate headquarters said we had tomake 110 percent of our original goal, I had called tbe salesmen inand said, in effect: This is ridiculous; we are not going to make it,but you know perfectly well what will happen if we don't. So get outthere and work your tail off.' The result was that I worked 20 hoursa day and they did nothing.

"This time I approached it differently, I told tbem three things.First, they were going to have to do some sacriflcing for the com-pany! Second, working harder is not going to do much good be-cause we are already working about as hard as we can. What willbe required are special deals and promotions. You are going tohave to figure out some new angles if we are to make it. Third, I'mgoing to back you up. I'm going to set a realistic goal with each ofyou. If you make that goal but don't make tbe company goal, I'llsee to it that you are not penalized. But if you do make the companygoal, I'll see to it that you will get some kind of special rewards."

When the salesmen challenged Charlie saying he did not haveenough influence to give them rewards, rather than becoming

THE McKINSKY QUARTERLY

angry, Charlie promised rewards that were in bis power to give -such as longer vacations.

Note that Charlie has now begun to behave in a number of ways thatwe found to be characteristic of the good institutional manager.He is, above all, higher in power motivation - the desire to in-fluence his salesmen - and lower in his tendency to try to do every-thing himself. He asks the men to sacrifice for the company. Hedoes not defensively chew them out when they challenge him buttries to figure out what their needs are so that he can influencethem. He realizes that his job is more one of strengthening them.And he is keenly interested in giving them just rewards for theirefforts.

The changes in his approach to his job have certainly paid off. Thesales flgures for his office in 1973 were up more than 16 percent over1972 and up still further in 1974 over 1973. In 1973 his gain over theprevious year ranked seventh in the nation; in 1974 it ranked third.And he wasn't the only one in his company to change managerialstyles. Overall sales at his company were up substantially in 1973as compared with 1972, an increase which played a large part inturning the overall company performance around from, a S15million loss in 1972 to a S3 million profit in 1973. The companycontinued to improve its performance in 1974 with an 11 percentfurther gain in sales and a 38 percent increase in profits.

Hidden shortcomings

Of course not everyone can be reached by a workshop. HenryCarter managed a sales office for a company which had very lowmorale (around the 20th percentile) before be went for training.When morale was checked some six months later, it had not im-proved. Overall sales gain subsequently reflected this fact since itwas only 2 percent above the previous year's figures.

Oddly enough, Henry's problem was that ho was so well liked hyeverybody that he felt little pressure to change. Always the life ofthe party, he is particularly popular because he supplies otbermanagers with special hard-to-get brands of cigars and wines at adiscount. He uses his close ties with everyone to bolster his positionin the company, even though it is known that his office does notperform well compared with others.

His great interpersonal skills became evident at the workshop whenhe did very poorly at one of the business games. When the dis-

.SUMMER1977 43

cussion turned to why he had done so badly and whether he actedthat way on the job, two prestigious participants immediatelysprang to his defense, explaining away Henry's failure by arguingthat the way he did things was often a real help to others and thecompany. As a result, Henry did not have to cope with suchquestions at all. He had so successfully developed his role as alikeable, helpful friend to everyone in management that, eventhough his salesmen performed badly, he did not feel under anypressure to change.

Checks and balances

What have we learned from Ken Briggs, George Prentice, CharlieBlake and Henry Carter? Principally, we have discovered whatmotive combination makes an effective manager. We have alsoseen that change is possible if a person has the right combinationof qualities.

Oddly enough, the good manager in a large company does not havea high need for achievement, as we define and measure that motive,although there must be plenty of that motive somewhere in hisorganization. Tbe top managers shown here have a high need forpower and an interest in influencing others, botb greater than theirinterest in being liked by people. The manager's concern for powershould be socialized - controlled so that the institution as a whole,not only the individual, beneflts. Men and nations witb this motiveprofile are empire builders; they tend to create high morale and toexpand tbe organizations they head.

But there is also a danger in this motive profile; empire buildingcan lead to imperialism and authoritarianism in companies just asit does in nations. The same motive pattern which produces goodpower management can also lead a company or a nation to try todominate others, ostensibly in tbe interests of organizationalexpansion. Tbus it is not surprising that big business has had to beregulated from time to time hy Federal agencies. And it is mostlikely that international agencies will perform the same regulativefunction for empire building nations.

For an individual, the regulative function is performed by twocharacteristics that are part of the profile of the very best managers

a greater emotional maturity, where there is little egotism, and ademocratic, coaching managerial style. If an institutional powermotivation is checked by maturity, it does not lead to an aggressive,egotistic expansiveness.

THE McKINSEY' QUARTEliLY

For nations, this checking means that they can control theirdestinies beyond their borders without being aggressive and hostile.For individuals, it means they can control their subordinates andinfluence others around them without resorting to coercion or to anauthoritarian management style. Real disinterested statesmanshiphas a vital role to play at the top of both nations and companies.

Summarized in this way, what we have found out through empiricaland statistical investigations may just sound like good commonsense. But the improvement over common sense is that now thecharacteristics of the good manager are objectively known.Managers of corporations can select those who are likely to be goodmanagers and train those already in managerial positions to bemore effective with more confidence.

David MeClclland isprofessor of psychology at Harvard Univcrsitv;David Burnham is president of McHcr and C'ompany, a l)ehavi()ralscicncf consulting firm in Boston. This article, winner of aMcKinsey Foundation Award for 197H. is roi)rint('(l bv pormissiotifrom the March April 1976 issue of the Harvard liusiness Review.Co])yright y 1976 by thf President and Follows of llai'vaffi Colloj^c.All rights I'csorvcd.

SUMMER 1977