The leadership game: Matching the man to the situation

Fred E. Fiedler

ost people in management leadership Whether

training

would

accomplishes

agree that

group of petty officers, for example,

something.

no overall

it always does what it is intended

do is another

question.

one whose

behavior

formance

improved

leadership

training

changed

after he went about

who have gone through

leaders

leadership

training

per-

through

a

Unfortunately, as many

one training

and still perform

ever. Even more intriguing standing

or whose

program.

most of us also know after another

to

Most of us know some-

program as

are the many out-

who have had little at all-Joan

or no

of Arc being a

stellar example.

results.

On

performed

well as captains

studies of leadership

reveal

train-

the same disappointing

the average,

people

with

much

leadership

and majors

ated from military research. Nealey

perfor-

study in Canada,

their

basic

tasks as

who had gradu-

college. These experimenby results from field

and I found

no relationship

between amount of training and performance of post office managers as rated by their immediate

superiors.

correlations

between

performance

In addition, amount

of police

patrol

cent studies show similar

Empirical ing generally

mance. In a follow-up trainees

we found

in leadership

tal studies are supported

people

as poorly

differences

and noncommissioned infantry

I found

zero

of training

and

sergeants.

Re-

findings

for officers

officers of an American

division. This

does not necessarily

mean

that

training perform about as well as people with little or no training, and reviews by Stogdill;

leadership training need be ineffective. Quite the contrary. Our data suggest that leadership

Campbell, Dunnette, Lawler, and Weick; and others present no evidence that any particular leadership training method consistently improves organizational performance.

training, under certain conditions, systematically improves the performance of some leaders while it decreases the performance of

Kesearch by my associates and me has revealed the same disappointing results. When we compared a group of Belgian navy recruits and a well-trained and experienced

others. Obviously, we have to understand the conditions under which leadership training is effective if we are to make much this area. While

recognizing

progress

the legitimacy

in of

In leadership, --merely

the situation is the thing. There

people with the potential to be successful

tions or resounding

leadership

training

satisfaction I want

are, says Fiedler,

designed

to improve personal

my remarks

job

growth,

here to train-

First, let me briefly comment approaches.

a preliminary

program

support

as this

Last, I will describe the training

that we have developed

control,

on the basis

of this theory and that we are currently

vali-

dating.

behavior

is under

that a few weeks of telling

a leader how to behave or convincing kind

on pres-

training

some data that

is that leadership

of behavior

in the appropriate

I will propose

theory of leadership

well as present formulation.

Then

programs voluntary a certain

ing that aims to improve task performance as it is defined by an organization. ent training

leaders under certain condi-

failures under other conditions.

and to enhance

to confine

no born leaders

situations

emotion-charged,

relationships

to the subordinate

changes.

the fact that leadership

are highly

personal

is best will result

behavior

This ignores

him that

that mean

inter-

a great deal

as well as to his boss. We

probably

expect more change in interpersonal

behavior

than a routine

training

hope to deliver. The manner late to authority

figures

program

can

in which we re-

and subordinates

for most of us a very important

is

interaction

that we learn over many years. And it is very CURRENTPRACTICESBASEDON

d&cult

QUESTIONABLEASSUMPTIONS

emotional

indeed

to change

relationships.

such

It is essential,

fore, that we ask just how much Let us first look at present practices and, in particular, their underlying assumptions. One assumption

that

guides

many

training

typical

leader

actually

of leadership

that is re-

only in situations

all condi-

of control.

behavior

or attitude

performance

tions

and that

every trainee

to adopt. For example, thorities

contend

permissive,

under

therefore

needs

several prominent

au-

that a good leader has to be

participative,

or human-relations-

under

If we take a close look at the empiri-

change

his

leadership

behavior

in which he has a great deal

In situations

in which

a leader

is

pressure, in which there is considerable

uncertainty

and insecurity,

ior seems to depend ual’s personality situation.

oriented.

the

has over his own be-

Our studies suggest that a leader can voluntarily

to good

there-

control

havior.

pro-

grams is the notion that there is one ideal kind lated

significant

A

leadership

behav-

on the way the individ-

interacts didactic

with his leadership approach-telling

leader to be more considerate,

permissive,

a or

cal results, however, it is obvious that neither the permissive, considerate leaders nor the

decisive-is about as effective as telling someone that he should be more lovable or less

autocratic,

anxious.

directive

leaders

obtain

optimum

performance under all conditions. Yet any training program that seeks to develop the same kind implicitly

of leadership assumes

behavior

or attitude

that there is one best lead-

ership style. A second major

assumption

in many

A third assumption is that the more powerful and influential leader will be more effective because he will be able to make his group work harder on the organization’s tasks. On the basis of this assumption, many training

programs

try to increase

a leader’s

control

and influence

in various

give him human-relations can

make

himself

subordinates. These They

power,

Ph.D.

clc~gr.rrr II~ the Unil*crsit)

of Chicago

ate ulork. .4jtrr

onr ~‘car as an instructor

he moved

Illinois,

to the University

ofpsychology

and director

etfrctiveness

of psychology

and organization

of Washington,

where

organizational

the Unioersity

professor

Fellow

among tiveness,

Leadership

ment. Award

American

Personnel

ups also awarded

He received

Association

He is currently

of Administrative

Science

zational

and Human

of the American and the Journal

training,

Research

questionable

dinates

functions

will therefore

ness of participative

advice of unintelligent pect brilliant

of Applied

of

As

depends

in

and ability of the

people can hardly

ex-

answers.

Let me stress again, in participative

however,

management

that or in

any other kind of leadership approach is not necessarily bad practice, nor will it be ineffective for all trainees.

Academy

effective.

the effective-

The leader who listens to the

more

Performance,

shown,

large part on the intelligence group members.

with his subor-

management

Psychological Organi-

management

be more

Jon Blades has recently

consulting

editor

assumption

which holds that a leader who shares

his decision-making

training

in 1953. He

Quarterly,

the needs of the

is at the basis of participative

the

of the

of Consulting

of the American

Social Psychology.

M.

Manage-

Research

Association.

Management,

Effec-

well over 120 articles

Psychology

The Journal

at

In 1969 study.

the Outstanding

Arlrnrd by the Division

Iiehavior

An equally

professor

with Martin

satisfying

organization.

and one year

and Effective

for Outstanding

to take

with

of Leadership

He has published

are incom-

are apt

to interfere

of several books-

and papers on leadership.

needs

needs

a lead-

and of man-

Belgium.

and more recently

Chemers,

the leader’s

Where

At the very least, they are likely

an associate of the Univer-

A Theory

the fact that

is an arena in which

er’s and an organization’s patible,

and guest professor

is the author

them,

situation

ignores

precedence.

srty of INinois’s center for advanced Dr. Fledler

lies

In 1969 he

research

of Louvarn,

he was appornted

the leadership

Dr. Fiedler

of .4msterdnm

as Ford Faculty

approach

he also directs the

spent one year as Ftrlbright

power

the organiza-

the needs of his organization.

at the University

research group.

at the University

at

oj the group

research laboratory.

became projessor agement

where

as well as knowing

a leader must satisfy his own needs as well as

of

ulhcre he rtJentuat’ly &came

of an organi-

and punishments

This

in

that

him to skip most of his undergradfi-

Chicago,

technical

his expertise.

tion has to offer.

irntl

1947 and 1949 alter tafting examintrt~ons enuh(ed

knowing

the organization

what rewards wcrir~cci /?;.c wnstc’?s

may give a leader

so that he can make full use of his le-

within

E. Fietller

to work harder.

teach him the intricacies

gitimate

Fred

to his

will enable him

so that he can increase

zation

so that he

acceptable

his subordinates

programs

training

more

This supposedly

to motivate

ways. They

training

discriminating

Rather, about

we need to be

whom

we train

and the situation for which we attempt to train a particular leader. Most leadership training programs fail to do this because they give all trainees the same training despite the fact that practically all of the empirical evidence

tells us that the performance

of a group

de-

only as a co-worker

pends in part on the kind of task and the situ-

might

ation in which the leader has to operate.

admirable,

Where do we go from here?

otherwise

but also as a person who

have some acceptable,

if not

traits. The “high LPC” leader sees

close interpersonal

relations

as a requirement

for task accomplishment. Let me, however,

strongly

emphasize

THE CONTINGENCYAPPROACHTO

that we are here talking

LEADERSHIPTRAINING

ties of goals. We are not speaking

about leader

behaviors.

of the task

My position

on training,

not surprisingly,

based on the contingency effectiveness.

of a group or an organization

on the interaction

er’s personality

between

and the situation.

we have to match structure

model of leadership

In essence, this theory holds that

the effectiveness depends

is

the leader’s

the lead-

Specifically, motivational

(that is, the goals to which he gives

the highest priority) the situation

with the degree to which

gives the leader control

and in-

We measure the

(LPC). think

Least

Co-worker

This scale asks the individual of everyone

worked, with

the leader’s motivation

Preferred

with

whom

and then to describe

whom

he could

work

Scale

first to

he works at the

years

1 for the scale of oppos-

used to describe

the least pre-

ferred co-worker.) her least preferred and

rejecting

who describes

co-worker

terms

his or

in very negative

(a low LPC)

emotional

in effect

reaction

with whom he or she cannot

to people

work-in

effect,

“If I can’t work with you, you are no damn good!”

This is the typical pattern

relations

of a person

who, when forced to make the choice, opts first for getting on with the task and worries about his interpersonal relations later. Someone who describes even his least preferred co--worker in relatively more positive terms in effect looks at the individual not

might

the group

to suc-

cess. In this latter case the relationship-motivated, high LPC leader might be quite singleminded

about

general arousing LPC

accomplishing

the

we find that uncertain conditions

leaders

lations

tend

concentrate

is the case in situations

task.

and

In

anxiety-

to make

the

low

on the task, while on their re-

with their subordinates.

The opposite

in which the leader is

secure and in control. The other major factor in this theory is defined

by the “situational

leader

therefore,

indicates has

control

favorableness”

the degree and

influence

comes of the group interaction. measure

situational

has more

and,

the out-

We generally

favorableness

subscales:

task structure, members

to which

feels that he can determine

of three

An individual

while

of close interpersonal

the

he worked

with whom

shows a strong

tenance

behaviors,

be possible only by driving

that basically

with whom

(See Figure

the main-

Zeast well. This

ago or someone ing attributes

and pleas-

ant interpersonal

he has ever

the one person

can be someone moment.

accomplishment

priori-

might well call for very considerate

the high LPC leaders concentrate

fluence over the outcomes of his decisions. by

The

about different

on the basis

leader-member

relations,

and position power. The leader

control support

and him,

influence

if (1)

(2) he knows

his

exactly

what to do and how to do it, and (3) the organization

gives him the means to reward and

punish his subordinates. The

crucial

question

then

is to de-

termine the specific situations under which various types of leaders perform best. The contingency model has consistently shown that the task-motivated (low LPC) leaders tend to perform most effectively in situations in which their control and influence are very

Figure I LEAST PREFERREDCO-WORKER SCALE

Think someone person

of the person

with whom

you can work

least well. He may be someone

you knew in the past. He does not have to be the person with whom

you had the most difficulty

in getting

you work

with now, or

you like least well, but should

a job done. Describe

this person

to you. Pleasant

Friendly

Rejecting

Helpful

Unenthusiastic

Tense

Distant

Cold

Supportive

Boring

Quarrelsome

Self-assured

Gloomy

Open

Unpleasant

7

6

5

4

3

2

1

:-:-:-:8

7

6

5

:_:_:_:_: 4

3

2

1

:_:-:_ 1

2

3

:_:_:-:_ 4

5

6

7

8

:_:-:-:8

7

6

5

:_:4

3

:---:-: 2

1

:-:-:-:_ 1

2

3

4

:_:----5

6

:_-_-:_: 7

8

:-:-:-:1

2

3

4

:_:_:-:-: 5

6

7

8

:-:-:1

2

3

:-:-:-:-:-: 4

5

6

7

8

2

3

4

5

6

7

8

:-:_:_:8

7

6

5

4

_:_:_: 3

2

1

:-:-:-:8

7

6

5

4

3

:-:_: 2

1

:_:_:_:-:-:-:-:-: 1 2

3

4

5

6

7

8

:-:-:-:1

3

4

5

6

:-:-: 7

8

6

5

4

3

2

1

5

4

3

2

1

Unfriendly

:-:

Accepting

Frustrating

Enthusiastic

Relaxed

Close

:_:_:_:-:_:----:-:-:

Warm

:-:

:-:-

Hostile

Interesting

:-:2

Uncooperative

Harmonious

:-:-:-:-:-:-:-:-: 8

Efficient

.-:

8

1 Cooperative

:_:-:-

:_:-:-:-

7

Hesitant

:-:-:_:-:_:-:-:_:

Inefficient

8

7

6

:-:-:1

2

3

4

5

-:-:-: 6

7

8

:-:-:-:-:-:-:-:_: 8 7

6

5

4

3

2

1

:-:-:

Cheerful

Guarded

be the

as he appears

Figure 2 SCHEMATIC REPRESENTATIONOF THE PERFORMANCEOF RELATIONSHIPAND TASK-MOTIVATED LEADERS IN DIFFERENT SITUATIONAL FAVORABLENESSCONDITIONS

Task

Good

Relationship

I

Favorable

-

-

3

4

-

5

6

7

8

Good

Good

Good

Good

Poor

Poor

Poor

Poor

Task

High

High

Low

Low

High

High

Low

Low

Strong

Weak

Strong

Weak

Strong

Weak

Strong

Weak

Leader position

power

high and in situations low. By contrast, LPC)

leaders

tions in which

in which it is relatively

relationship-motivated

tend to perform their control

(high

best in situa-

and influence

is

Validating

described

the model

fully controlled

experiment

The

model

by the schematic

is most easily

drawing

axis shows the group’s or the

organization’s

performance.

the leader

by Chemers

and

showed that the contingency model for 28 percent of the variance in

in Figure

2. The vertical

“situational

The

with

control

horizontal

favorableness”-that

is, the degree to which the situation

This relationship has now been found in well over 50 dil’ferent studies; in fact, a careSkrzypek accounted

task performance.

axis indicates

moderate.

-

Unfavorable

Leader-member relations structure

-

motivated

Moderate

2

1

motivated

and

provides

influence.

The

solid line shows the performance of high LPC leaders, and the broken line the performance of low LPC leaders. As can be seen, the high LPC, or relationship-motivated, leaders generally perform

best in situations

in which their

11

relations

with subordinates

structure

and position

also perform subordinates position

well when

are low. They

their

relations

with

are poor but task structure

power

moderate

are good but task

power

are high

favorableness

2). Task-motivated

(both

and

situations

of

as defined

in Figure

leaders perform

best when

and Butler at the University

of Utah.

researchers

These

were low LPC persons. Half were assigned at random

to receive training,

fluence are either high or low.

task.

changing -that

group performance

the leader’s motivational

else by modifying While

his leadership

and the motivational

a part of personality, and uncertain

structure

per-

that is

this is clearly a difficult

process. It is, however, relatively

easy to modify

the leadership

can select a person not others,

situation.

for certain

kinds

we can assign

him

We

training most

A three-letter would English

pened in that particular

favorable

situation

will gradually

a zone of moderate

situational

move

favorableness.

Such a change in control and influence also change leadership

into

performance:

We would

motivated

leader who performs

favorable

zone will perform less well with while the relationship-motivated

well in the un-

training, leader should improve with training as he moves from the unfavorable to the moderately favorable Training mance others.

zone, toward the left of the graph. should, therefore, decrease perforof some

leaders

This was recently laboratory

but

increase

demonstrated

expei iment conducted

it for by a

by Chemers,

relations,

low positask if the

an unfavorable

therefore

leaders

would

expect that the perform

than would relationship-motivated training,

the task would The

LPC

leaders

better

than

would

become become

structured moderately

relationship-motivated,

should would

then

high

perform

task-motivated,

better

ones. With

relatively low LPC

leaders.

would

The task-

in

study, the groups had

were untrained-thus

and the situation

start off in an un-

words

situation.

favorable.

leaders who, for example,

an e at the end

and an unstructured

task-motivated

that

with

be an e.

only one-letter

very poor leader-member

in order

It follows

con-

are a and I, and so on. As it hap-

er’s control

influence.

letter would

word

training seeks to increase the favorableness of a situation-that is, it increases the leadand

Training

letters and then assuming

be the. The

leaders

leadership

to the

leaders such rules as count-

that the most frequent

certain

to increase his power and influence. As we said before,

sisted of teaching

tion power,

or

the others

unrelated

a series of cryptograms.

of jobs

tasks, give him more or less responsibility, we can give him leadership

ing

ing all the alphabet

situation.

it is possible, of course, to change

sonality

and

structure

is, the basic goals he pursues in life-or

while

an assignment

The group task consisted of decipher-

2 that

either by

four-

and psychol-

ogy students, with an ROTC cadet as the leader. Half the leaders were high and half

were given

we can improve

assembled

man groups composed of ROTC

all three factors that define their control and inIt should be clear from Figure

12

Rice, Sundstrom,

Figure

3 shows

the results

of this

study. As expected, the low LPC leaders performed

better than did high LPC

the unfavorable

situation,

while

leaders in high

LPC

leaders performed better in the moderately favorable situation. However, as the theory predicts but we would not normally expect, the low LPC leaders with training also performed less well than did the low LPC leaders who had not received training. Similar findings have been reported in real-life situations. For example, we conducted a study of 32 consumer cooperative

Figure 3

INTERACTION OF TRAININGANDLPC ON GROUP PRODUCTIVITY

I

I

Trained

companies measures

in which

we obtained

of performance

Untrained

objective

on all companies

in

tivated

leaders.

motivated

the federation,

When we then divided the gen-

less experience

eral managers

into those with high and those

than

with

low LPC

relatively those

little

with

concomitant Evaluations

scores, as well as those with

trained

experience

relatively

and trained

and

ship-motivated.

and

ilar results.

general

manager

had a favorable leadership situation. A relatively inexperienced and untrained manager would have correspondingly influence, hence a situation only moderately favorable. As Figure 4 shows, vated

leaders

performed

with

better

experience than

did

general

less control and that would be

The whether

and training highly

managers

and

task-motitraining

relationship-mo-

relationship-

with

relatively

performed

better

experienced

and

who were relation-

Several other studies give simquestion

usually

the leader could change

arises

as to

his motiva-

tional structure or his behavior to suit the situation. I would not want to preclude this possibility, but I also really do not think that this is done very easily. As we said before, leadership is a very ego-involving relationship, and in such relationships

the

the

managers

the more

experience

training, we obtained Figure 4. from several judges indicated that

the experienced

did

and training

high

However,

general

it is very d&cult

to

control our behavior. It is certainly much more emotionally charged than, say, the interaction between

a salesman

and a customer

or

13

Figure 4 THE PmsChfED CHANCE IN PERFORMANCE OF I~ZLATIONSHID-AND TASK-MOTIVATED COMPANY MANAGERS AS A FUNCTION OF INCREASED EXPERIENCE

102.00

101 .oo

Relationship motivated

(high

Task motivated

(low

LPC)

LPC)

100.00

99.00

,” .-ii

98.00

”

c

LI m c 2

z

z 0”

Mean

-

97.00

96.00

-

92.00 Experience Leadership Situation

High

Experience

Intermediate

a lawyer and his client. We are talking about patterns of interaction that are fairly central to our personality. The degree to which a person is affected by his relations with others, or the degree to which he is driven to get a

14

Low

Experience Favorable

job done, is not very easily changed

from one

day to the next. I don’t really think

that you

can make someone who is cold and businesslike into a warm, cuddly leader in the course of a few hours or even days. Chris Argyris’s

account of these point all too well.

A

NEW

difficulties

documents

this

APPROACH TO TR.~ININc

Let me now get to the point of this symposium. What kind of training would the contingency model call for? We have tended to look at people as infinitely malleable, as infinitely

capable

of changing

being changed even though

their

behavior

the behavior

of

we are trying

change may have been acquired lifetime.

and

by just a few hours of training

In contrast,

we have tended as relatively

rigid.

see themselves

very little control

over their

the problem

to

to look

inflexible

and

and emotional

situation.

degree

our accessibility to which

the

closeness of our rela-

tions with others. However, modify

This is clearly not true. We have to teach peo-

what

We may not be able to change warmth

as having

work

they may just explain

is and then let him run with the

ball.

over a whole

at the organization Most people

it; for another,

we can frequently

to subordinates,

the

we share information,

and

the extent to which contacts with subordinates

ple that they have much more control over the

are formal and businesslike

relevant

and relaxed. We can give detailed, step-by-step

aspects of their own leadership

than they generally them, therefore, ation

jobs

realize. We have to teach

that they can change the situ-

so that it will better

match

their

per-

sonality.

task instructions

or informal,

or general

policies and guide-

lines. We can use our position some conditions

social,

power

and share decision

under making

under others. The

research

on

the

contingency

model shows that effective leadership

Martin

depends

on maintaining

the right match of personality

and of situation.

We can certainly

I have grammed

teach people

called Leader

in

managers

which they are likely to succeed and those in

situation,

the particular

for

Match

how

Linda

Mahar,

a self-administered

manual

situations

how to recognize

Chemers,

developed

leadership

that

how to determine

training

attempts

to diagnose

their

to teach leadership

the kind of situa-

which they are likely to be less effective. And

tion that best matches their personality

we can tell them, “If you avoid jobs in which

tivational

you are likely to fail, you are bound

situation

to be a

pattern,

style. One validation

We now have reason to believe that we can teach people with reasonable accuracy

cessfully

nates and their

to which

superiors

degree to which

their

subordi-

are supportive,

a task is structured,

the

and the

degree to which they have position power.

or mo-

to modify

the

study has now been suc-

completed,

using

second-level

ers of a volunteer

public

health

that

Latin

America.

operates

one involves

in

middle

managers

lead-

organization Another

of a govern-

ment agency. In these studies the leaders who were trained

The next step is to give trainees guid-

how

so that it does match their leadership

success.”

to assess the degree

and

and pro-

significantly

with

Leader

Match

performed

better than did leaders in a com-

ance in seeking or developing leadership situations in which they are most likely to be

parable, randomly selected control the time of this writing, a third

group. At validation

successful,

study also seems to be producing

significant

or to modify

their

situations

to

match their personalities. We can also train them in ways to provide their subordinate

aging, and we hope to obtain further

leaders with conditions

during

tivational They

patterns. Successful

may

that match

their mo-

results. These early results are highly encourthis coming year. In summary,

leaders do this intuitively.

say about

a person

that

“you

have to give him a lot of backing if you want him to be effective.” For one person, they may spell out in detail what to do and how to do

evidence

my own position

is that

we must train people differentially-not everyone should be trained to behave in the same way or to adopt the same attitudes. In fact, we will be better served by training our leaders in how to change their leadership situations

15

than

in

how

Leadership match

to

of person

change

and

personality

ing this balance. success

that

recent

studies

show

requires

promise

years,

more that

and

not

this

weeks.

approach

for the future

Our

training and they

holds

to

co-authored md

and Company, summary work

SELECTED

The

most

gency Theory

16

extensive

model of

description

can be found

read

Ralph

Leadership

Effectiveness

contin-

Fiedler’s (McGraw-

A

(Scott,

1975), provides presents

of the current

a less

some

of the

as well as on

1974). Another Dunnette,

(h4c(;raw-Hill,

popular

description

refer-

and Weick’s and

Effrc-

who

would

like

a more

of recent work on leadership read

Psycho/ogy

Today

somewhat

less popular,

You Make Leaders Organizutronal

excellent

Lawler,

1970).

reader

should

should

of Leadership

Performance,

Behavior,

tiveness

in a com-

literature

Handbook

Stogdill’s

Free Press,

training

of the

in Fred

Chemers,

The leader who is interested

The

HIBLIOCRADHP

M.

Management

on validation

plete bibliography

hlanageri~l

CD

Martin

and

in-

selection, and experience.

ence is Campbell.

training.

by

technical

scales. A subsequent

Eflective

technical

recent

some of the valida-

and provides

Foresman

(The

considerable

of leadership

volume

in detail

about the various

training,

so that reason

formation

more

successful, have

tion studies

Leadership

the situations

situations We

to

uncertain

model

to be most

effectively.

Hill, 1967). In it, he reviews

proper

trying

with

can recognize their

a

way of achiev-

of contingency

they tend

perform

personalities.

situation,

It is an effort

can modify

believe

their requires

is the hard

that leaders

in which they

change

effectiveness

Fred

(February

Fiedler’s

description,

More Effective?”

Dynamics

article

in

1973). A clear, but

(Autumn

“How appears 1972).

Do in