THE DYNAMICS OF CHARISMATIC LEADERSHIP IN ORGANIZATIONS
Chanoch Jacobsen
Robert J. House
Faculty of Industrial Engineering and Management
Technion, Israel Institute of Technology
Haifa, Israel 32000
ABSTRACT
An integrated theory of charismatic leadership was developed, and
a System Dynamics model built of that theory. Applying the model
to J.F. Kennedy, four data sets relating to the Peace Corps were
successfuly reproduced by the model with the same set of initial
values, attesting to the sensitivity of the general theory and
model to the manifestations of JFK’s charismatic leadership.
THE PROBLEM
The frequent need of organizations for radical, yet legitimate
changes has led to renewed interest in charismatic or, as it is
sometimes called, transformational leadership. There are two
traditions of theory and research in this field. Social scientists
have looked at the situational conditions for the emergence of
charismatic leaders, and the contingencies in which personal
charisma will or will not be "routinized" into a stable social
order (e.g., Weber, 1922; Eisenstadt, 1968; Bendix, 1985; Willner,
1984; Glassman and Swatos, 1985). Psychologists have concentrated
on the personal characteristics of charismatic leaders and the
dynamics whereby followers are drawn to support the leader’s vision
(e.g., House, 1977; Burns, 1978; Tichy and Devannah, 1986; Conger
and Kanungo, 1988).
Since neither of the two approaches covers the phenomenon in its
entirety, we have integrated them into one comprehensive theory of
charismatic leadership. Then, in order to permit the empirical
testing of this theory, we have applied the strategy advocated by
Jacobsen, Bronson, and Vekstein (1990), and constructed a System
Dynamics model of the integrated theory. All the model variables
and coefficients have been defined to match the concepts and
behaviors postulated by the theory. This paper presents the
results of our first attempts to reproduce empirical data with the
model. ‘
THEORY. AND MODEL
The following summarizes the theory briefly. A social situation is
conducive to charismatic leadership if one or more of the following
conditions apply: (1) Important moral values are perceived to be
involved; (2) the relationship between performance and goals is
unclear and ambiguous; (3) the situation is unstable, and (4)
requires exceptional efforts (Shamir, House, and Arthur, 1989).
When, in such a situation, there is a person with the appropriate
motivation (McClelland, 1985; Avolio and Bass, 1988; Sashkin,
1988:142), s/he will articulate a vision of change, causing people
to recognize that person as leader and identify themselves with the
vision.
The leader will then arouse followers by communicating his/her
confidence in their ability, and the expectation that they will
realize the vision through their own activities (Bass, 1985; Eden,
1990: 128). An elite of active followers will be enthused by the
leader’s role-model of commitment to commit themselves also, and
autonomously reinforce themselves to exceptional performance. If
the vision is consonant with the existing power structure, the
improved performance will be routinized into an institutionalized
order. If it is not, routinization will depend on the support. the
leader has from the collective identity of the three types of
followers: passive, active, and committed. In any case, the
collective identity will ameliorate the conditions which had made
the situation conducive to charismatic leadership initially.
But some of the elite, being now mere participants in routine
activities, will no longer be autonomously reinforced and become
disenchanted. Others will defect, having committed themselves only
to solve a personal problem (Smelser, 1963: 256-7). As a result,
a bureaucracy develops to maintain performance. Bureaucratization
causes depersonalization of those who are still active, and an (at
least apparent) displacement of the original visionary goals, so
that even passive followers become alienated. Thus the situation,
though changed, reverts to what had made it conducive to
charismatic leadership in the first place.
We have constructed a System Dynamics model of this theory (Fig.1)-
The present study set out to test whether the model, built as it is
on the basis of a theory, can reproduce the empirical data relating
to an actual charismatic leader, President John F. Kennedy (for the
code, see Appendix).
AN EMPIRICAL CASE
The Peace Corps, a volunteer service for giving grass-roots aid to
developing countries, was announced by John F. Kennedy shortly
before being elected, in his vision to "get the country moving
again". It was "the part of the aid effort which best expressed
the spirit of the New Frontier" (Schlesinger, 1965:557). When,
eight months later, the first Peace Corps Volunteers met with JFK
prior to departing overseas, they regarded him not merely as their
formally elected President, but identified with him personally
(Wofford, 1980:249ff). Kennedy, aware of their commitment to his
vision, told them that they were "the most immediate response that
the country has seen to the whole spirit which I tried to suggest
in my inaugural" (Rice, 1985:303).
JFK was a charismatic leader. In terms of our theory, he had
articulated a vision of change, causing people to identify
themselves as his followers; he had aroused many of them to want to
participate actively. in realizing the vision; and he had enthused
the select elite among the active participants to commit themselves
- 268 -
-
-
&
3
Flow Diagram
Figure 1.
by joining. As for institutionalization, Rice (1985:ix) calls the
Peace Corps JFK’s "most affirmative and eduring legacy", and the
evidence supports such an assessment. Delegating extensive powers
to Sargent Shriver, the first P.c. director, he authorized the
formal independence of the new bureaucracy from the Administration
Establishment. At the time of its 25th anniversary in 1986, over
6000 Volunteers were active world-wide in 62 countries (General
Accounting Office, 1989).
There also is evidence that disenchantment, depersonalization, goal
displacement and alienation was felt at least by some of the
Volunteers (Cowan, 1970). It seems reasonable, therefore, to take
the Peace Corps as an empirical manifestation of JFK’s charismatic
leadership, affecting individuals as well as the organizational
structure of his administration. Thus our model of charismatic
leadership should be able to reproduce these effects with the
corresponding model variables.
DATA AND RESULTS -
We were able to locate four different time-series relating to the
growth and subsequent routinization of the Peace Corps. The Annual
Reports of the Peace Corps provide the yearly number of
applications received. Based of U.S. Census figures, these were
converted into yearly percentages of the relevant population,
namely college graduates of that year. This data set (JFK1)
corresponds to the model variable AROUSE. The same source also
gives the annual numbers of actual trainees and volunteers who
entered service. These figures, similarly converted into yearly
percentages (JFK2), correspond to the rate of commitment to the
cause (COMMIT).
The U.S. Budget for the years 1963-1988 provides the annual number
of permanent paid staff employed by the Peace Corps. This, when
related to the yearly number of Volunteers (JFK3), gives a
reasonable indicator of bureaucratization (BUREAU). Finally, the
Peace Corps has also published the number of volunteers annually
who terminated their service overseas before their regular two-year
tour of duty was up. These, computed as a percentage of the ELITE
(JFK4), are an indirect indicator for the rate of disenchantment
(DISENCH) for this period.
A measure of the of fit between model and data is given by TI
(Trend Index), which is the percentage of data variance reproduced
by the corresponding model variable. The TI values obtained were
are follows:
JFK1 (AROUSE) = 90.0%,
SFK2 (COMMIT) = 84.4%,
JFK3 (BUREAU) = 36.4%,
JFK4 (DISENCH) = 80.3%
The plots (Figure 2) all show a fairly good fit between the trends,
even for JFK3 and BUREAU where the TI was rather low.
Encouraging as these results are, we must emphasize that these
simulations do not represent a test of the theory, since the model
c
ae
SS ——-m am —--—-m
a
2
x
1 i
x \
=e b2s- +L
# 1961, 19d. ASBL. 1927, a 1561. AML SEL. 1980,
. “nM ‘ . “KE . "
—— soe —-—-™
1 ir
— [’ t
n / \ \
“Y | S- wy =
ha 197, 13 1380 arr 190. 1981 1M
. ee " . : “mkt . :
Figure 2. Plotted Output
was designed to be congruent with the theory, both in its structure
and its dynamics. But they do show that the model, while
functioning as the theory would predict, can reproduce the complex
developments of an actual case of charismatic leadership.
Moreover, the empirical adequacy of the theory can now be
systematically tested by running the model against data on other
charismatic leaders. If the model can reproduce further data sets
also, solely from the internal dynamics of its fixed coefficients
acting on the initialized parameters, then we may conclude that the
theory provides an adequate explanation of the data. Conversely,
if the model fails to reproduce the data, then either the theory
(and model) need to be revised or should be rejected.
REFERENCES
Avolio, B., and B.M. Bass. 1988. Transformational leadership:
Charisma and beyond. In J.G.Hunt, B.R. Beliga, H.T.Dachler,
and C.A. Schriesheim (eds.), , Emerging Leadership Vistas.
Lexington, MA: Heath & Co.
Bass, Bernard M. 1985. Leadership and Performance Beyond
Expectations. New York: Free Press.
Bendix, Reinhard 1985. Reflections on charismatic
leadership. Pp. 17-25 in R.M. Glassman and W.H. Swatos,
op. cit.
Burns, J. MacGregor. 1978. Leadership.
New York: Harper & Row.
Conger, Jay A., and R.N.Kanungo, eds. 1988. Charismatic
Leadership: The Elusive Factor in Organizational
Effectiveness. San Francisco: Jossey-Bass.
Cowan, Paul. 1970. The Making of an Un-American: A Dialogue
with Experience. -New York: Viking.
Eden, Dov. 1990. Pygmalion in Management.
Lexington, MA: Heath & Co.
Eisenstadt, Shmuel N. 1968. Max Weber on Charisma and
Institution Building. Chicago: U. of Chicago Press.
General Accounting Office. 1989. Peace Corps: A Statistical
Profile. Washington, D.C.: Govt. Printing Office.
Glassman, Ronald M., and W.H. Swatos (eds.). 1985. Charisma,
History, and Social Structure. New York: Greenwood Press.
House, Robert J. 1977. A 1976 theory of charismatic
leadership. In J.G. Hunt and L.L. Larson (eds.),
Leadership: The Cutting Edge. Carbondale, Ill:
Southern Illinois University Press.
Jacobsen, C., Bronson, R, and D. Vekstein. 1990. A strategy
for testing the empirical adequacy of macro-sociological
theories. Journal of Mathematical Sociology, 15(2): 137-48.
McClelland, D. 1985. Human Motivation.
Chicago: Scott, Foresman.
Rice, Gerard T. 1985. The Bold Experiment.
Notre Dame, IN.: University of Notre Dame Press.
Sashkin, Marshall 1988. The visionary leader. Pp.122-160 in
J.A. Conger and R.N. Kanungo, op.cit.
Schlesinger, Arthur M. 1965. A Thousand Days: John F. Kennedy
in the White House. New York: Houghton Mifflin.
Shamir, Boas, Robert J. House, and Michael B. Arthur. 1989.
The Transformational Effects of Charismatic Leadership:
A Motivational Theory. Working Paper 89-16, The R. H.
Jones Center, Wharton School, University of Pennsylvania.
Smelser, Neil J. 1963. A Theory of Collective Behavior.
New York: The Free Press.
Tichy, N.M., and M.A. Devanah. 1986.
The Transformational Leader. New. York: Wiley.
Weber, Max. 1922. Gesammelte Aufsaetze zur Wissenschaftslehre.
Tuebingen: Mohr.
Willner, A.R. 1984. The Spellbinders: Charismatic Political
Leadership. New Haven, CO: Yale University Press.
Wofford, Harris L. 1980. Of Kennedys and Kings: Making Sense
of the Sixties. New York: Farrar, Straus, Giroux.
APPENDIX
* CHARISMA Revised: 19-1-92
A Model of Charismatic Leadership in Organizations
MACRO STAT(X,MEAN, VAR)
INTRN CUMX, CUMXSQ, ITIME, ELPST
CUMX . K=CUMX. J+DT#X.T
ITIME=TIME
CUMXSQ. K=CUMXSQ.J+DT* (X.J*X.J)
CUMXSQ=0
VAR. K= (CUMXSQ.K/ELPST. K) - (MEAN. K*MEAN. K) +1E-30
[END
Brsarapepran
NAzZaANARAHAHAAAAAANA
OP BP PZ PP AP PAZ MP APA BP
SECTION I. INITIALIZATIONS
(For JFK1, JFK2, JFK3, JFK4)
Initial Identifiers (pcet.)
Initial Active (pet.)
Initial Elite (pet.)
Leaders’s Motives (pet.)
Leader’s Vision (prob.)
Leader’s Communication Skill (prob.)
Leader’s Personalized Charisma (pct.)
Structural Consonance (pct.)
Intl. Situational Conduciveness (pct.)
Ousting of Current Powerholders (Flag)
IPERF=1.0 Initial Performance (Output/Input)
STAFF=17.69 Salaried Staff (pct.) i
Delay from IDENTIF to ACTIVE
Delay from ACTIVE to ELITE
DELATI=1 Delay from ACTIVE to IDENTIF
DELPTR=2 Delay from PERFORM to ROUTINE ;
TIME=BEGTIME
BEGTIME=1961
ENDTIME=1987
SECTION II. VARIABLES AND RELATIONSHIPS
FADE. K=TABHL (TFADE, TIME. K, BEGTIME, ENDTIME, *
(ENDTIME-BEGTIME) /10)
TFADE=1/.97/.92/.85/.75/.65/.57/.50/.45/.42/.40
Fading of charisma curve
ACCEPT. K=ACCEPT.J+DT* (ALIEN.JK-RECOG.JK)
ACCEPT=100-INIDEN-INACT-IELITE Pet. Accepters
CONTAG.K=1+TABLE (TCONTAG, COLLECT.K,0,100,10)
TCONTAG=0/.03/.08/.17/.3/.5/.7/.83/.92/.97/1
Contagion Multiplier
ARTIC.K=((MOTIVES/100) *VISION)*FADE.K Articulation (prob.)
IDENTIF.K=IDENTIF.J+DT* (RECOG.JK-AROUSE.JK+*
DEPERS .JK-ALIEN.JK)
IDENTIF=INIDEN Pet. Identifiers
EXPCON. K=100* ( (ARTIC. K*COMMUN) *ACTMUL. K*FADE. K)
Expressions of Confidence
ACTMUL. K=TABHL (TACTMUL, ACTIVE. K,0,100,10)
TACTMUL=1.5/1.47/1.41/1.33/1.24/1.16/1.1/1.06/1.03/1.01/1
Multiplier from ACTIVE
EEXPCON.K=CLIP(100,EXPCON.K,EXPCON.K,100) Safety Clip
ACTIVE. K=ACTIVE.J+DT* (AROUSE.JK-COMMIT.JK*
+DISENCH.JK-DEPERS.JK)
ACTIVE=INACT Pet. Active Participants
ROLEMOD. K=( (EEXPCON. K/100) *MODRF .K) * ( (100-PERSON) *
/100) *FADE.K Role Modeling (probability)
MODRF . K=TABHL (TMODRF, PERCHA.K,0,2,.2)
TMODRF=1/.97/.92/.85/.7/.5/.3/.15/.08/.03/.001
Metamorphic Reduction Factor
RROLEMOD. K=CLIP(1,ROLEMOD.K,ROLEMOD.K, 1) Safety Clip
ELITE. K=ELITE.J+DT* (COMMIT. JK-DISENCH.JK)
- 274 -
i
seams
Suir
BP Pp Pp PP BPP APP Pr APP Pp Ae ee
i i eo
ELITE=IELITE Pet. Elite
COLLECT. K=100-ACCEPT.K Collective Identity (pet.
COLLRF . K=TABLE (TCOLLRF , COLLECT.K,0,100,10)
TCOLLRF=1.0/.97/.92/.83/.70/.50/.30/.17/.08/.03/.001
Collective Identity Reduction
AUTREIN.K=(ACTIVE.K/2) +ELITE.K
Automonously Reinforced (pct.)
PERFORM. K=IPERF* (CONDEF.K*AUTEFF.K) Prfomnce (ratio)
CONDEF . K=TABLE (TCONDEF , CONDUCE.K, 0, 100,10)
TCONDEF=1/.97/.92/.85/.7/.5/.3/.15/.08/.03/.001
Conduciveness Multiplier
CONDUCE. K=INCOND*EXTFOR.K*COLLRF.K Sitnl. Conduciveness
EXTFOR. K=TABHL (TEXTFOR , TIME.K, 0,36, 6)
PEXTFOR=1/1/1/1/1/1/1 Exogenous Forces
AUTEFF . K=TABLE (TAUTEFF , AUTREIN.K,0,100,10
TAUTEFF=1/1.1/1.25/1.4/1.7/2/2.3/2.6/2.75/2.9/3
Autonomous Reinforcement Multiplier
PERCHA.K=PERFORM.K/IPERF Performance change (d.less)
PEREFF .K=TABHL (TPEREFF , PERCHA.K,1,3,.2)
TPEREFF=1/3/8/15/30/50/70/85/92/97/100
Performance Multiplier
SUPPORT. K=CLIP (COLLECT. K, PEREFF.K,OUST, .5)
Support for OUST
ROUTINE. K=DLINF3 (SUPPORT.K, DELPTR) *CONSON/100
Routinization (pcet.)
BUREAU. K=(STAFF+( (ACTIVE. K+ELITE.K) *ROUTINE.K) ) /*
( (ACTIVE. K+ELITE. K) * IMMUNE) Bureaucratization (pct.)
DISPLAC. K=COLLECT. K* (BUREF.K/100)
Goal Displacement (pct.)
BUREF . K=TABLE(TBUREF , BUREAU.K,0,100,10)
TBUREF=5/8/13/20/30/45/60/70/77/82/85
Bureaucratization Multplr
SECTION III. RATES OF CHANGE
RECOG. KL=ACCEPT. K*ARTIC.K* (CONDUCE. K/100) *CONTAG.K
Pet. Recognition/year
AROUSE. KL=DELAY1 ( (EEXPCON.K/100) , DELITA) *IDENTIF.K
Pet. Arousal/year
coMMrT. KL=DELAY1 (RROLEMOD.K, DELATE) *ACTIVE.K
Pct.Commitment/year
HAVEN=10 Yrs. people.can shelter in ELITE
FRINGE. K=TABHL (TFRINGE, TIME.K, BEGTIME, BEGTIME+*
HAVEN, HAVEN/10)
TFRINGE=15/14.8/14.5/14/13/11/7/3/1/.5/.2
Lunatic fringe (pct.)
DISENCH. KL= (ROUTINE. K/ (100-PERSON) ) *ELITE.K*FRINGE.K
Pet. Disenchantment/year
DEPERS . KL=ACTIVE. K*DELAY1 (DEPERF.K, DELATI)
Pct. Depersonalization/year
DEPERF. K=TABLE (TDEPERF , BUREAU.K,0,100,10)
TDEPERF=0/.03/.08/.17/.3/.5/.7/.83/.92/.97/1
Depersonalization Multplr
ALIEN. KL=(DISPLAC.K/100) *IDENTIF.K Pct. Alienation/year
275 om
ap
PPP Pp Pree Pp Pepe PED D
TERMINATION
TERM. K=CLIP(0,1,0, IDENTIF.K) *CLIP(0,1,0,ACTIVE.K) *
*CLIP(0,1,0, ELITE. K) Run termination
LENGTH.K=CLIP(TIME.K,ENDTIME,0,TERM.K) Run length
DATA SETS
JFK1.K=TABLE (TIFK1, TIME.K,1961,1987,1) Applicants (AROUSE)
TIFK1=2.575/3.878/6.083/7.387/6.306/5-909/4.557/3.492/*
2.446/1.773/2.307/2.451/2.628/2.283/2.185/1.352/*
1,035/1.018/1.360/1.218/1.162/1.077/1.233/1.019/*
1.017/0.883/0.714
JFK2 .K=TABLE (TIFK2, TIME.K, 1961, 1987,1) Joiners (COMMIT)
TIFK2=.108/.545/1.197/1.631/1.983/2.176/1.936/1.585/1.225/*
+887/.616/.563/.574/.609/.533/.443/.428/.527/.474/*
+451/.408/.398/.402/.417/.456/.415/.376
JFK3 .K=TABLE (TIFK3,TIME.K,1961,1987,1) Paid staff (BUREAU)
TIFK3=17.69/13.73/12.91/10.14/7.63/7.11/7.65/8.81/9.33/*
10.68/13.26/14.14/13.10/12.30/15.27/17.44/17.83/*
14.90/16.20/16.11/16.73/14.08/14.06/13.38/12.49/*
13.30/16.74
JFK4.K=TABHL (TIFK4, TIME.K,1961,1986,1) | Leavers (DISENCH)
TIFK4=.01/.141/.399/.574/.793/.975/.960/.861/.642/.448/*
+286/.239/.258/.290/.226/.164/.169/.221/.185/.163/*
-140/.129/.125/.125/.139/.129
TREND INDEX CALCULATIONS (with MACRO)
DEVSQU1.K=(AROUSE.KL-JFK1.K) * (AROUSE. KL-JFK1.K) i
NUMER1.K=STAT (DEVSQU1.K,MN1.K,VN1.K) numerator for TI1 |
DENOM1. STAT (JFK1.K,MD1.K,VD1.K) denominator for TI1
T11.K=(1~(MN1.K/VD1.K) ) #100 Trend Index for JFK1
DEVSQU2.K=(COMMIT. KL-JFK2.K) * (COMMIT. KL-JFK2.K)
NUMER2 . K=STAT (DEVSQU2.K,MN2.K,VN2.K) numerator for TI2
DENOM2.K=STAT(JFK2.K,MD2.K,VD2.K) denominator for TI2
TI2.K=(1-(MN2.K/VD2.K) ) #100 Trend Index for JFK2
DEVSQU3 . K= (BUREAU . K-JFK3 . K) * (BUREAU. K-JFK3 .K)
NUMER3 . K=STAT (DEVSQU3.K,MN3.K,VN3.K) numerator for TI3
DENOM3 . K=STAT (JFK3.K,MD3.K,VD3.K) denominator for TI3
TI3.K=(1-(MN3.K/VD3.K)) *100 Trend Index for JFK3
DEVSQU4 . K= (DISENCH . KL-JFK4 . K) * (DISENCH. KL-JFK4 .K)
NUMER4 . K=STAT (DEVSQU4.K,MN4.K,VN4.K) numerator for TI4
DENOM4 .K=STAT(JFK4.K,MD4.K,VD4.K) denominator for TI4
TI4.K=(1-(MN4.K/VD4.K) ) #100 Trend Index for JFK4
SAVE ACCEPT, IDENTIF, ACTIVE, ELITE, COLLECT, CONDUCE, CONTAG, “
EXTFOR, ARTIC, FADE, EEXPCON , RROLEMOD , AUTREIN , PERFORM, *
PERCHA, ROUTINE, BUREAU, DISPLAC, RECOG , AROUSE, COMMIT, *
DISENCH, DEPERS , ALIEN, FRINGE, JFK1, JFK2, JFK3 , JFK4, *
TI1,TI2,TI3,TI4
SPEC DT=.125,PRTPER=1, PLTPER=1, SAVPER=1
