M-Form Reorganization in Diversified Firms
James J. Hoffman
TEXASTECHUNIVERSITYRobert J. Williams
UNIVERSITY OFNORTHALABAMABruce T. Lamont
FLORIDASTATEUNIVERSITYScott W. Geiger
UNIVERSITY OFSOUTHFLORIDAWe identify two factors that affect how quickly highly interdependent In the only study to explicitly examine this issue (Lamont,
diversified firms recover from the performance disruptions that typically Williams, and Hoffman, 1994), diversification strategy was
accompany M-form reorganizations: (1) organizational tenure and (2) found to be an important contingency affecting the length of
the tenure of top management teams. We find that top management teams the performance recovery period following M-form adoption.
with a rare combination of long organizational tenure and relatively short Unrelated diversifiers, with limited linkages among their
busi-team tenure are associated with faster performance recovery in diversified nesses and, thus, relatively easier reorganization tasks, were
firms with extensive business unit interdependencies. The effects of mana- found to recover more quickly than vertical or related
diversifi-gerial tenure on recovery time were found to be inconsequential in unre- ers, firms with more complex interdependencies among their
lated diversified firms where the reorganization task is less complex. different businesses. It was also observed that the variance in
J BUSN RES2000. 50.287–296. 2000 Elsevier Science Inc. All rights recovery times for the vertical and related diversifiers was
reserved. much greater than that for the unrelated diversifiers,
sug-gesting that other factors besides strategy may be important in explaining the performance recovery times of the more highly interdependent firms.
A
lthough improved performance is a common motivation One factor (besides strategy) that may be important in for most major structural reorganizations, such as the explaining the performance recovery times of diversified firms adoption of the multi-divisional structure (or M-form), is managerial tenure. Specifically, organization tenure and it usually deteriorates during implementation and may take top management team (TMT) tenure are probably the two years to reach or exceed pre-reorganization levels. Indeed, dimensions of managerial tenure that would have the largest some firms may never recover from M-form reorganizations. impact on a firm’s ability to recover from a structural reorgani-Given that the M-form is the dominant structure for managing zation. This is because organizational tenure is directly related large, complex, particularly diversified corporations in the to one’s knowledge and power bases within the firm while United States and throughout the world (Galunic and Eisenh- the amount of time managers have spent on the top manage-ardt, 1996; Hoskisson, Hill, and Kim, 1993; Markides and ment team is directly related to the manner in which top Williamson, 1996; Rumelt, Schendel, and Teece, 1994), iden- management team members approach their duties and interact tification of the factors that help explain whether and how with other team members. Although several studies have in-quickly firms recover from M-form reorganizations is criticalvestigated the effect of managerial tenure on organizational to both organizational scholars and the managers of growing
outcomes, most have only focused on one dimension of mana-and diversifying firms contemplating M-form adoption.
gerial tenure, and none have examined the effects of manage-rial tenure on recovery time after M-form reorganization. In Address correspondence to James J. Hoffman, Ph.D., Department of Manage- order to extend earlier research efforts, the objective of the ment, College of Business, Texas Tech University, Lubbock, TX 79409; Tel.:
(806) 742-4004; E-mail address: [email protected] current study is to empirically examine whether managerial
Journal of Business Research 50, 287–296 (2000)
2000 Elsevier Science Inc. All rights reserved. ISSN 0148-2963/00/$–see front matter
tenure (organizational tenure and TMT tenure) interacts with suggested that firms pursuing certain strategies, specifically unrelated diversification and, to a lesser extent, vertical inte-firm strategy (vertical integration, related diversification, and
unrelated diversification) to affect firm performance recovery gration, benefit more from adopting the M-form structure than firms that have pursued related diversification (Hoskis-time after M-form reorganization.
son, 1987; Lamont et al., 1994). In addition, it can be theo-rized that certain characteristics of the firm’s top management
Background
team may play a pivotal role in how well a firm is able torealign systems, culture, personnel, and procedures with the Examination of factors that affect firm recovery time after
new structure and to ultimately recover from the M-form M-form reorganization is important because firms increasing
reorganization. their levels of diversification generally change their structures
from the functional (U-form) to the multi-divisional (M-form)
structure (Amburgey and Dacin, 1994; Chandler, 1962;
Theoretical Development
Hoskisson et al., 1993; Williamson, 1975). In theory, the
M-form is supposed to lead to increased firm efficiency as The idea that top management team characteristics affect orga-nizational outcomes is not new. Hambrick and Mason (1984) operating decisions are delegated to the newly created
semi-laid the conceptual foundation for the relationship between autonomous divisions (Bettis and Chen, 1990; Chandler,
top management team characteristics as managerial age, ten-1962; Hill, Hitt, and Hoskisson, 1992; Hoskisson, Harrison,
ure, education, and functional background and firm perfor-and Dubofsky, 1991; Hoskisson et al., 1993; Roberts perfor-and
mance. The work of Hambrick and Mason (1984) focused on Greenwood, 1997; Russo, 1991; Williamson, 1975). While
the possible impact of basic demographic factors on manage-the firm’s head office retains strategic decision-making
respon-rial strategic choices related to firm strategy and related risk, sibilities, routine operational decisions are assigned to the
innovation, firm structural dimensions, and firm growth and various divisions that are directly involved in the firm’s
pro-performance. duction and distribution activities. An M-form reorganization
Several studies have confirmed empirically that such a generally involves dramatic changes in structure, systems,
con-relationship does, in fact, exist between TMT characteristics trols, and power and decision-making centers within the firm
and firm performance. For example, Norburn and Birley (Miller and Friesen, 1984; Roberts and Greenwood, 1997;
(1988) found that teams (the firm’s dominant coalition) pos-Tushman and Romanelli, 1985).
sessing greater functional experience and wider educational The underlying theory behind firms’ adoption of the
training experienced better performance outcomes. Wiersema M-form structure is that in large, complex corporations the
and Bantel (1992, p. 91) found that “cognitive perspectives, M-form enhances performance resulting from improved
op-as reflected in a team’s demographic characteristics, are linked erating and strategic controls. Research suggests that the
to the team’s propensity for change.” Thomas, Litschert, and M-form enhances firm performance by shifting operational
Ramaswamy (1991) found that the greater the alignment be-decision making and responsibility to decentralized product
tween firm strategy and TMT characteristics, the better the or geographic divisions. Such decentralization serves to
im-firm’s performance. More recently, Boeker (1997) found that prove the firm’s efficiency as decisions are made by managers
firm performance moderated the effects of TMT characteristics closer to the firm’s production and distribution activities
(Arm-on corporate strategic change. our and Teece, 1978; Bettis and Chen, 1990; Galbraith and
One TMT characteristic that has been studied extensively Kazanjian, 1986; Williamson, 1975). Further, utilization of
over the years is managerial tenure. Results from several stud-the M-form allows stud-the firm’s separate divisions to exercise
ies have suggested that there is a link between managerial greater operational autonomy, thereby reducing the routine
tenure and organizational outcomes (e.g., Barker and Pat-decision-making activities of top managers. Top managers are
terson, 1996; Boeker, 1997; Finkelstein and Hambrick, 1990; better able to concentrate on broad strategic issues facing
Greening and Johnson, 1996; Pfeffer, 1983). To date however, the firm, while the burden of operational decision making
no study has examined the possibility of a link between mana-is reduced as a result of decentralization (Chandler, 1962;
gerial tenure and recovery time from M-form adoption, nor Galbraith and Kazanjian, 1986; Williamson, 1975).
have any studies examined the effect of firm strategy on this In addition to the above benefits associated with M-form
relationship. adoption there are also drawbacks. The major drawback is
that it may take years for the firm to realign systems, culture,
Managerial Tenure and Recovery Time
personnel, and procedures with the new structure (Chandler,
1962; Galbraith and Kazanjian, 1986; Lamont et al., 1994; Long managerial tenure has been linked with commitment to the status quo (Boeker, 1997; Michel and Hambrick, 1992; Roberts and Greenwood, 1997). It can be theorized, however,
that the extent of this drawback may be somewhat contingent Wiersema and Bantel, 1992), reduced learning and increased inertial responses (Hambrick, 1994, 1995; Virany, Tushman, on the type of diversification strategies which the firm has
infor-mation processing (Finkelstein and Hambrick, 1990, 1996; progress through the disruptive reorganization process. These managerial attributes should also provide the firm with needed Hambrick, 1994, 1995), and a tendency for managers to
main-stability during the process. tain past business patterns (Barker and Patterson, 1996; Boeker,
1997; Staw and Ross, 1980). Arguments have been made
Firm Strategy Recovery Time
that TMTs possessing long organization tenure and long TMT
In addition to suggesting that managerial tenure can affect a tenure tend to be more conservative, that they do not engage
firm’s performance recovery time following M-form adoption, in as much strategic experimentation (relative to TMTs with
research also suggests that a firm’s strategy prior to M-form shorter organizational and TMT tenure), and tend to pursue
implementation affects the magnitude and nature of the reor-imitative strategies directly in line with industry trends, and
ganization problem (Hill and Hoskisson, 1987; Hoskisson, thus their organizations perform close to the industry average.
1987; Hoskisson, Harrison, and Dubofsky, 1991, Lamont et Long tenure has also been related to increased first-order
al., 1994). Specifically, the level of interconnectedness or inter-learning among management team members and lower levels
dependence among the various businesses and/or functional of second-order learning (Hambrick, 1995; Virany et al.,
units affects M-form implementation (Hill and Hoskisson, 1992). TMTs that exhibit first-order learning continually and
1987; Hoskisson, 1987; Hoskisson, Harrison, and Dubofsky, incrementally improve at those things that offer them a
com-1991; Hoskisson, Hitt, and Hill, com-1991; Jones and Hill, 1988; petitive advantage (Tushman and Keck, 1989; Virany et al.,
Markides and Williamson, 1996). The level of interconnected-1992). This is because first-order learning is based on
experi-ness reflects the extent to which each of the firm’s parts, ence and the competence that comes from working with the
or divisions, is dependent on and supports the firm’s other same actors on a limited number of fixed tasks over a long
divisions (Thompson, 1967). The level of interconnectedness period of time. In contrast, second-order learning is
character-across units affects the types of controls and systems required ized by a shift in core assumptions and decision-making
prem-to achieve smooth, corporate-wide coordination (Hill and ises. It involves unlearning prior premises and standard
op-Hoskisson, 1987; Jones and Hill, 1988), with higher levels of erating procedures (SOPs). It also requires that the firm develop
interconnectedness requiring more complicated and costly new decision-making frames, new SOPs, and new interpretive
coordinating mechanisms (Thompson, 1967). schemes (Bartunek, 1984; Virany et al., 1992).
Different strategies tend to produce different patterns of Prior research suggests that the results of long managerial
interconnectedness between the various businesses and func-tenure should be a reluctance of managers to undertake major
tional units of the corporation (Hill and Hoskisson, 1987; organizational changes, such as an M-form reorganization.
Hoskisson, Harrison, and Dubofsky, 1991; Jones and Hill, Nevertheless, the accumulation of managerial tenure may
pro-1988). Thompson (1967) identified three patterns of work vide certain benefits. New managers generally enter their
posi-flow interdependence between organizational units: pooled, tions lacking extensive knowledge about contacts and
proce-serial, and reciprocal. Each pattern of interdependence repre-dures needed to successfully perform their duties, and such
sents a greater level of interconnectedness and a more complex knowledge must be gradually obtained over time (Boeker,
and costly coordination problem. With pooled interdepen-1997; Gabarro, 1987; Hambrick and Fukutomi, 1991). In
dence, typical in unrelated diversifiers (Jones and Hill, 1988), addition to knowledge of the job, team members enhance and
each unit operates largely independently of other units except solidify their power over time (Pfeffer, 1983). The CEO, in
for a common input source or ownership structure. With serial particular, accumulates power through the cultivation of
inter-interdependence, the dominant pattern in vertically integrated nal and external relationships with other executives and
orga-firms (Jones and Hill, 1988), the output of one unit is the input nizations. The CEO, as well as other TMT members, also
to another unit. Reciprocal interdependence is exemplified by accrues power by accumulating expert knowledge of the firm complex exchanges between units, where the actions of one and its processes, and by granting rewards to organizational unit affects the actions and outcomes of another. Related members (Boeker, 1997; Hambrick and Fukutomi, 1991). In diversifiers generally possess this type of interdependence most cases, this accrual of power can occur only through among their units, as they typically serve overlapping markets, managerial tenure. Thus, while the dysfunctional aspects of drawing on overlapping sources of supply, frequently using lengthy managerial tenure have been addressed in the empiri- similar and interdependent technologies (Jones and Hill, cal literature, the potentially positive influence of lengthy ten- 1988). Considering only the level of interconnectedness asso-ure seems to have been largely overlooked. ciated with the various strategies, related diversifiers should Although long managerial tenure may contribute to TMTs face the most complicated and costly reorganization problem being adverse to major organizational change, the greater ex- in implementing satisfactory M-form controls and systems pertise of long-tenured managers may prove invaluable to since they have the highest level of interconnectedness. Verti-firms who have formally committed to undertaking a struc- cally integrated firms face the next most complicated and tural reorganization. Once the reorganization is underway it costly reorganization problem while the reorganization task seems reasonable to assume that a manager’s knowledge and confronting unrelated diversifiers is the least problematic
As previously stated, performance recovery requires effec- longer organization tenure will perform no differently (i.e., will have similar performance recovery times) tive corporate-wide coordination and efficient operations
within the newly created M-form. Given that long organization than those unrelated diversified firms whose TMTs have shorter organization tenure.
and team tenure may lead to shared ideologies, standardized
ways of communicating, a more homogeneous perspective, H3b: Following M-form implementation, unrelated diver-and an overall greater understdiver-anding of the day-to-day work- sified firms that have top management teams with ings of the organization (Tushman and Keck, 1989; Virany longer TMT tenure will perform no differently (i.e., et al., 1992), TMTs with longer tenure are likely to be better will have similar performance recovery times) than able to facilitate corporate-wide coordination, efficient opera- those unrelated diversified firms whose TMTs have
tions, informal communication flows, and other integration shorter team tenure.
devices. Such long-tenured managers have established the necessary knowledge and power bases to bring about more
Methodology
effective reorganizations. Since a high level of firm unit
inter-dependence is most strongly associated with firms pursuing
Sample
related diversification and vertical integration strategies, theThe sample consisted of the 76 firms identified as adopters positive aspects associated with long tenure should manifest
of the M-form. The 76 firms make up a large percentage of themselves through more rapid recovery times. Therefore, the
the firms identified in the Armour and Teece (1978) and following hypotheses are offered.
Hoskisson (1987) studies and are the same firms used by La-H1a: Following M-form implementation, related diversi- mont et al. (1994). These studies identified firm strategy and fied firms that have top management teams with the year in which the firms initiated and completed the transi-longer organization tenure will perform better (i.e., tion to the M-form, thus permitting an unambiguous identifi-have shorter performance recovery times) than those cation of prior strategy and the timing of structural change. related diversified firms whose TMTs have shorter Firm strategy classifications were based on data from
an-organization tenure. nual reports and Moody’s Industrial Manuals. As Hoskisson
(1987) noted, the firms’ strategies primarily corresponded to H1b: Following M-form implementation, related
diversi-the dominant-vertical, related-constrained, and acquisitive con-fied firms that have top management teams with
glomerate strategies in the familiar Rumelt (1974) typology. longer TMT tenure will perform better (i.e., have
shorter performance recovery times) than those
re-Measures and Data Collection
lated diversified firms whose TMTs have shorter team
tenure. Organization tenure (TENORG) and tenure on the top
man-agement team (TENT) were the independent variables. Data H2a: Following M-form implementation, vertically
inte-on managerial tenure were based inte-on demographic informatiinte-on grated firms that have top management teams with
found in Dun and Bradstreet’s (1962–1968)Reference Book of longer organization tenure will perform better (i.e.,
Corporate Managementsfor selected years during 1962–1986. have shorter performance recovery times) than those
The top management team consisted of the CEO and senior vertically integrated firms whose TMTs have shorter
vice-presidential positions for each firm. This is the same organization tenure.
definition that Michel and Hambrick (1992) and Hambrick H2b: Following M-form implementation, vertically
inte-(1994) used in their work on top management teams. Using grated firms that have top management teams with
this definition, the largest team had nine members while the longer TMT tenure will perform better (i.e., have
smallest team had three members. On average, the teams had shorter performance recovery times) than those
verti-six members. cally integrated firms whose TMTs have shorter team
TMT and organization tenure were measured in chronolog-tenure.
ical years, with team tenure measuring the number of years that individual team members had spent on the TMT, and The reorganization task faced by unrelated diversifiers
organization tenure measuring the number of years that team should be less complex than with firms that utilize vertical
members had been employed with the organization. The team integration and related diversification strategies, as the need
tenure and organization tenure for each firm’s TMT were calcu-to retain high levels of unit coordination and mutual
adjust-lated as an average of the individual team members’ tenures. ment are much less for these types of firms. With unrelated
diversifiers, the lower levels of firm unit interdependence
PERFORMANCE RECOVERY TIME. Industry-adjusted return on as-should reduce the need for those positive attributes associated
sets (ROA) was used as the performance indicator in the with long tenure. Therefore,
construction of the performance recovery time (RECOV) vari-able. ROA was chosen for its comparability with previous studies H3a: Following M-form implementation, unrelated
La-mont et al., 1994). Following a procedure used in other studies implementation time (IMTIME) to influence performance re-covery it was included as a control variable. M-form imple-(e.g., Judge and Zeithaml, 1992; Lamont et al., 1994), each
mentation time was measured as the time required to fully firm’s annual ROA was adjusted for industry involvement to
implement the M-form. TMT tenure heterogeneity was also control for the performance effects of industry differences, as
included as a control variable. The need to control for TMT suggested by Dess, Ireland, and Hitt (1990). Following the
tenure heterogeneity was deemed necessary due to the nature same procedures as Lamont et al. (1994), industry-adjusted
of member interactions that normally occur among individuals ROA was calculated using firm ROA divided by industry ROA.
who have served on the TMT for different time periods Specifically, each firm’s annual ROA was divided by the
(Boeker, 1997; Hambrick, 1994, 1995). TMT tenure heteroge-corresponding annual ROA for the firm’s dominant industry.
neity (TENTHET) was measured as the coefficient of variation The dominant industry was operationalized using the IRS
of TMT tenure. industry classification system, equivalent to the three-digit
External to the firm, changes in gross national product SIC system. Due to the diverse revenue sources of unrelated
(GNP) reflecting general business cycles should also exert diversified firms, great care was taken to properly identify the
influence on the firm’s ability to recover to prior performance industry providing the greatest revenue source for the firm.
levels. Consistent with prior research on M-form reorganiza-As in prior studies (i.e., Hoskisson, 1987; Hoskisson,
Har-tion (Hoskisson, 1987; Lamont et al., 1994) GNP changes rison, and Dubofsky, 1991; Lamont et al., 1994) the dominant
were intended to control for the business cycle and time period industry was then identified for each of the unrelated
diversi-differences across the sample. Gross national product was fied firms.
measured as the annual percentage changes in GNP during Performance recovery time (RECOV) was measured as the
the recovery period. number of years, after the M-form reorganization was
com-pleted (as reported by prior studies, i.e., Hoskisson, 1987;
Analytical Procedures
Hoskisson, Harrison, and Dubofsky, 1991; Lamont et al.,
Because performance recovery time (RECOV) was a time-1994) for each firm’s industry-adjusted ROA to reach or
ex-event metric and some of the data were right-censored (i.e., ceed the firm’s average industry-adjusted ROA for the
five-not all firms recovered), a regression technique known as year period prior to reorganization. Out of 76 firms, 57 firms
survival analysis was used to test the hypotheses. When sur-experienced a performance decline during the transition
pe-vival analysis is utilized, censored observations are used in riod. For those firms that experienced no performance decline
estimating parameters, thereby avoiding biases that result from RECOV was coded as “0.” If a firm did not reach or exceed
eliminating observations or treating them as though an event its pre-change five-year performance average within 10 years,
(e.g., recovery) has occurred when it has not. Survival analysis RECOV was coded “10” in the analysis. Although 15 firms
is a more generally appropriate technique for analyzing data required longer than 10 years to recover, the recovery times
in which the dependent variable is elapsed time to some event of those firms were “right censored” at 10 years in the survival
(e.g., time to performance recovery) and the data are censored analysis.
(e.g., the recovery time of certain firms is unknown)
(Ander-CONTROL VARIABLES. A decision was made not to use a con- son, Auquier, Hauck, Oakes, Vanaele, and Weisberg, 1980; trol group in the current study. This decision was made be- Cox and Oakes, 1984). A preliminary analysis indicated that cause there is no meaningful way to measure the study’s recovery time was best modeled as following a Weibull distri-dependent variable, recovery time, for a control group of firms bution (Cox and Oakes, 1984), with the effects of the variables that did not experience a structural reorganization, and the on recovery time modeled as multiplicative. The SAS LIFEREG associated performance drop from which they would need to program was used with the logarithm of RECOV serving as recover, thus making useful comparisons between our sample the dependent variable (SAS, 1990).
and a control group rather difficult. Subgroup analysis was used to examine if managerial tenure
In place of a control group, five variables were identified (organization tenure and TMT tenure) interacted with firm that should exert influence on the tenure–recovery time rela- strategy type (vertical integration, related diversification, and tionship, and all five were included as control variables in the unrelated diversification) to affect firm performance recovery model. From an internal perspective, firm size, reflecting the time after M-form reorganization. Subgroup analysis was used magnitude of the reorganization task, and the firm’s prior instead of interaction terms because the use of interaction performance should affect the firm’s performance recovery terms is not possible when using the multiplicative modeling time. Firm size (SIZE) was measured as the logarithm of the approach in the SAS LIFEREG program. Subgroup analysis is firm’s employees in the year prior to reorganization, and prior an appropriate technique to use in this situation because it falls performance (ROAP) was measured as the firm’s average per- under the interaction perspective of measuring fit (Venkatra-formance during the five-year period prior to M-form imple- man, 1989). The use of subgroup analysis requires that a sample
mentation. be split into two or more groups. Subgroups are then tested
Table 1. Descriptive Statistics
Correlations
Variables Mean SD 1 2 3 4 5 6 7 8
RECOV 4.10 3.64
TENORG 23.27 8.85 20.14
TENT 9.07 3.36 0.35** 0.44**
SIZE 10.01 1.08 20.17 20.14 20.10
GNP 7.90 1.53 20.02 0.08 20.11 20.01
IMTIME 2.97 2.49 20.12 20.15 20.13 0.191 20.15
ROAP 1.04 0.50 0.45** 20.05 0.18 20.08 20.14 0.05
TENTHET 0.53 0.25 20.39* 0.09 0.11 20.46** 0.08 20.201 0.12
n576.
1p,0.10; *p,0.05; **p,0.01.
When using this procedure in the current study, separate tenure, however, is opposite to the effect hypothesized in H1b, suggesting that following M-form implementation re-regression analyses were conducted for each of the firm
strat-egy types. Each regression analysis consisted of one basic lated diversified firms that have top management teams with shorter team tenure recover more quickly than those related model which regressed performance recovery time (RECOV)
on the two independent variables (organizational tenure and diversified firms whose TMTs have longer team tenure. TMT tenure) and the five control variables. The equation For the vertical integration group, the parameter estimate representing the basic model is as follows: for organization tenure was negative and statistically signifi-cant (p,0.05), while the parameter estimate for team tenure RECOV5 B01B1(TENORG) 1B2(TENT)1 B3(SIZE)
was not significant. The results suggest that vertically inte-1B4(ROAP) 1B5(GNP)1B6(IMTIME) grated firms having TMTs with longer organization tenure
tend to recover more quickly following M-form reorganization 1B7(TENTHET)1Error
than firms having TMTs with shorter organization tenure. The results also suggest that top management team tenure is not
Results
a significant factor influencing recovery time for vertically integrated firms following a reorganization. Therefore, the The descriptive statistics for the variables are presented inresults offer support for H2a, but offer no support for H2b. Table 1 and the survival analysis results are presented in
For the unrelated diversification group, the parameter esti-Table 2. With the exception of (GNP) and (SIZE), there were
mates for both organization tenure and team tenure were not statistically significant correlations between the control
vari-statistically significant. These results support both H3a and ables (i.e., IMTIME, ROAP, and TENTHET) and the
indepen-H3b, that following M-form implementation unrelated diversi-dent and dependiversi-dent variables, thus supporting their inclusion
fied firms that have top management teams with longer organi-in the analysis. As expected, the correlation between the firm’s
zation tenure and TMT tenure do not recover more quickly prior performance (ROAP) and recovery time was positive
than those unrelated diversified firms whose TMTs have and significant, as was implementation time with firm size.
shorter tenure. Surprisingly, however, correlations between tenure
heteroge-neity and each of recovery time and firm size were negative
and significant, suggesting faster recovery times may be linked
Discussion
to more heterogeneous teams. Larger firm size was linked toThe findings generally support the importance of prior strategy less heterogeneous teams, a surprise given the greater
opportu-as a contingency affecting the relationship between managerial nities for diversity in large firms.
tenure and firm performance recovery after M-form reorgani-Results from the survival analysis indicate that for the
re-zation. We saw this as a logical extension of previous research lated diversified firms the parameter estimates for both
organi-that has examined the effect of managerial tenure on organiza-zation tenure and team tenure were statistically significant
tional outcomes, and also as a logical extension of previous (p,0.001). The significant negative parameter estimate for
research linking strategy to the performance effects of adopting organization tenure indicates that following M-form
imple-the M-form in large corporations. mentation related diversified firms that have top management
What emerges from the findings is the profile of a TMT teams whose members possess lengthy organization tenure
that may be most effective in implementing reorganizations tend to recover more quickly than those related diversified
in highly interdependent firms: the rare combination of a TMT firms whose TMTs have shorter organization tenure. This
Table 2. Survival Analysis Results (Dependent Variable: RECOV)
Log Independent Parameter Likelihood Strategy Variable Estimate Chi-Square Ratio
Related Div: Intercept 1.282 0.196 232.322
(right-censored) Tenorg 20.464 22.773***
Tent 0.194 11.752***
Size 20.086 0.092
GNP 9.705 1.016
Imtime 20.081 0.734
Roap 0.347 0.867
Tenthet 20.944 6.102*
Vertical Int: Intercept 1.143 0.223 223.733
(right-censored) Tenorg 26.654 7.090*
Tent 0.010 0.029
Size 0.113 1.305
GNP 24.349 0.053
Imtime 0.001 0.000
Roap 2.742 20.118***
Tenthet 20.030 1.544
Unrelated Div:1 Intercept 26.886 3.5641 210.210
(right-censored) Tenorg 20.337 2.444
Tent 0.476 1.637
Size 0.197 0.366
GNP 31.004 8.711**
Imtime 0.094 0.486
Roap 0.165 0.137
Tenthet 1.118 1.648
n576.
1p,0.10; *p,0.05; **p,0.01; ***p,0.001.
1Two-tailed significance
Where:
RECOV5Performance Recovery Time. Tenorg5Organizational Tenure. Tent5Top Management Team Tenure. Size5Organizational Size.
GNP5Gross National Product. Imtime5M-form implementation Time. Roap5The Firm’s Prior Performance.
Tenthet5Top Management Team Tenure Heterogeneity.
and short team tenure. For firms that have formally committed The unexpected finding linking short TMT tenure to more rapid recovery times requires further analysis of the dynamics to a reorganization, long organization tenure among team
members appears to be preferred. As predicted, such managers of TMT membership and group processes. Some recent work (Hambrick, 1994, 1995; Finkelstein and Hambrick, 1996) would have developed a solid knowledge base and an
exten-sive set of relationships within their firms and industries. Such makes the important observation that top management groups do not always function well as “teams.” Relevant to this study, detailed knowledge of the firm and its business is probably
not transferable to other firms (Kotter, 1982). Extensive un- is a posited negative curvilinear relationship between team tenure and team effectiveness and its ability to implement derstanding of an organization’s culture, systems, processes
and people, that is likely to only come with experience is organization-wide strategic change. That is, it takes time for group members to develop shared views and establish interac-critical to the mounting of organization-wide redesign efforts
and the recombining of these design elements around a new tion patterns that are necessary for effective communication and decision making, but only up to a point. After that point, structure (Weick, 1993). Further, by definition, long-tenured
managers have been successful, and a successful track record increases in team tenure are associated with less internal com-munication and superficial exchanges among members. It is generally serves as the basis for the accumulation of power
(Hambrick and Fukutomi, 1991). Therefore, organizational further posited that shared awareness and corroboration of new strategic imperatives eventually diminish, making it diffi-knowledge and power are derived from long organization
tenure, and apparently provide greater benefits for reorganiza- cult for the “team” to effectively mount organization-wide changes in a timely fashion. Given that none of the TMTs in tions involving firms with higher levels of interdependence
a period of time beyond that necessary for effective team and when, a team reaches this point, it may cause its firm to follow the pattern of decline and possible extinction as de-functioning (1.5 years) found in small group research. Our
scribed by Hannan and Freeman (1984). If a management results are consistent with the curvilinear team tenure–team
team is at risk of losing its responsiveness, strategy and OD effectiveness hypothesis, and particularly supportive of the
researchers may wish to examine proper mechanisms for diminishing effectiveness of long-tenured teams in
implement-maintaining the quality and freshness of management teams. ing large-scale organizational changes such as M-form
reorga-Given the findings, periodic changes in TMT membership nizations, at least in complex interdependent firms.
may become necessary among firms with high levels of unit While Hambrick and Fukutomi (1991) examined changes
interdependence that are contemplating a major structural associated with increasing CEO tenure, their basic arguments
reorganization. In such cases, however, care must be taken may also be relevant to all top managers. For example, early
not to overly disrupt the functioning of the firm’s management in their TMT tenures team members may experience higher
team during the crucial reorganization period. Further, it levels of commitment to achieving the firm’s goals (Hambrick
should be remembered that the inferences drawn from this and Fukutomi, 1991). Task knowledge will rise, and
informa-study about the ideal combination of long organization ten-tion diversity and task interest should be highest early in
ure and short TMT tenure among managers in firms having the team member’s tenure (Hambrick and Fukutomi, 1991;
highly interdependent units are related to periods of structural Kotter, 1982; Miller, 1991). With the exception of firm
knowl-change. This same combination of managerial characteristics edge and power, the team member’s interest and information
may not be as valuable to an organization during more routine sources may diminish over time. This decline in enthusiasm
operating periods. and/or information diversity may reduce the team member’s
As with most studies, there are limitations associated with overall effectiveness. Also, if groupthink (Janis, 1972)
devel-the present study that provide opportunities for future research. ops within the team the overall loss in managerial effectiveness
Future studies may utilize more recent data, a larger sample, will be magnified.
and may also want to examine other aspects of top manage-Although somewhat speculative, short TMT tenure coupled
ment teams. Differences in organizational governance structures with long organization tenure may provide the optimal
combi-limit the ability of a management team to influence performance. nation for undertaking a structural reorganization. When a
An examination of these variables would be an appropriate firm has managers who have relatively recently joined the
extension of the regression model tested in this study. TMT, and who possess significant organization tenure, these
Overall, it is hoped that results from the current study managers not only want to make things happen, but also
have provided additional insight into the manner in which know what and how to get things done. Chandler (1962)
the type of strategies firms pursue affects the relationship provides anecdotal support for this assertion, particularly in
between managerial tenure and firm performance recovery his case study description of the M-form reorganizations of
after M-form reorganization. It is also hoped that the current GM and DuPont. For example, Alfred Sloan, the primary
study serves as a point of reference for future studies that executive responsible for the successful M-form reorganization
examine the effects of strategy type on the relationship be-at GM, was a long-tenured organizbe-ational member but recent
tween other management variables and other organizational newcomer to the TMT, as were his two key assistants
(Don-outcomes. aldson Brown and John Lee Pratt) in putting the plan into
action. Similarly, the set of executives who engineered the
References
rapid DuPont reorganization had spent nearly all of their
Amburgey, Terry, and Dacin, Tina: As the Left Foot Follows the
business careers with the company, were rapid climbers of
Right? The Dynamics of Strategic and Structural Change.Academy
the administrative ladder, and, with the exception of Fletcher
of Management Journal37 (1994): 1427–1452.
Brown, were relatively new to the ranks of top management.
Anderson, Sharon, Auquier, Antoine, Hauck, W., Oakes, David,
At least in the case of firms possessing high levels of unit Vandaele, W., and Weisberg, H.:Statistical Methods for Compara-interdependence, the findings of the present study offer sup- tive Studies, Wiley & Sons, New York. 1980.
port for the conclusions presented above. The results also Armour, Henry O., and Teece, David J.: Organizational Structure suggest that when firm unit interdependence is minimal (i.e., and Economic Performance: A Test of the Multidivisional
Hypoth-esis.Bell Journal of Economics9 (Spring 1978): 106–122.
the unrelated diversifiers), the link between managerial tenure
Barker, Vincent, and Patterson, Paul: Top Management Team Tenure
and rapid reorganizations is greatly reduced or non-existent.
and Top Manager Casual Attributions at Declining Firms
At-The implications of the results presented are significant for
tempting Turnarounds.Group and Organization Management21
strategy researchers. The interactive patterns found in this
(1996): 304–337.
study may provide some explanation for apparently
contradic-Bartunek, Joan J.: Changing Interpretive Schemes and Organizational
tory findings of the strategic choice proponents and organiza- Restructuring. Administrative Science Quarterly 29 (September tional ecologists. Although speculative, it is very possible that 1984): 355–372.
at some point a TMT loses some of its ability to adapt to Bettis, Richard A., and Chen, A.: Organizational Structure, Financial
Market Performance and Vertical Integration, inRisk, Strategy,
and Management, Richard A. Bettis and H. Thomas, eds., JAI, Versus Incremental M-form Reorganization on Performance: A Time-Series Exploration of Intervention Dynamics. Journal of Greenwich, CT. 1990, pp. 183–205.
Management11 (Fall 1985): 55–70. Boeker, Warren: Strategic Change: The Influence of Managerial
Char-acteristics and Organizational Growth.Academy of Management Hoskisson, Robert E., Harrison, Jeffrey S., and Dubofsky, David A.: Capital Market Evaluation of M-form Implementation and Journal40 (1997): 152–170.
Diversification Strategy. Strategic Management Journal 12 (May Chandler, Alfred D.:Strategy and Structure: Chapters in the History
1991): 271–279. of American Industrial Enterprise, MIT Press, Boston, MA. 1962.
Hoskisson, Robert E., Hill, Charles W., and Kim, Hicheon: The Cox, David R., and Oakes, D.:Analysis of Survival Data, Chapman &
Multidivisional Structure: Organizational Fossil or Source of Hall, London. 1984.
Value.Journal of Management19 (Summer 1993): 269–298. Dess, Gregory G., Ireland, R. Duane, and Hitt, Michael A.: Industry
Hoskisson, Robert E., Hitt, Michael A., and Hill, Charles W.: Manage-Effects and Strategic Management Research.Journal of Management
rial Risk Taking in Diversified Firms: An Evolutionary Perspective. 16 (March 1990): 7–27.
Organization Science2 (August 1991): 296–314. Dun & Bradstreet,Reference Book of Corporate Managements, Dun &
Janis, Irving L.:Victims of Group Think, Houghton Mifflin, Boston, Bradstreet, New York. Selected years.
MA. 1972. Finkelstein, Sydney, and Hambrick, Donald. C.:Strategic Leadership:
Jones, Gareth R., and Hill, Charles W.: Transaction Cost Analysis of Top Executives and Their Effects on Organizations, West, St. Paul,
Strategy-Structure Choice.Strategic Management Journal9 (March-MN. 1996.
April 1988): 159–172. Finkelstein, Sydney, and Hambrick, Donald C.: Top Management
Judge, William, and Zeithaml, Carl P.: Institutional and Strategic Team Tenure and Organizational Outcomes: The Moderating Role
Choice Perspectives on Board Involvement in the Strategic Deci-of Managerial Discretion.Administrative Science Quarterly35
(Sep-sion Process.Academy of Management Journal35 (October 1992): tember 1990): 484–503.
766–794. Gabarro, John J.:The Dynamics of Taking Charge, Harvard Business
Kotter, John P.:The General Managers, Free Press, New York. 1982. School Press, Boston, MA. 1987.
Lamont, Bruce T., Williams, Robert J., and Hoffman, James J.: Perfor-Galbraith, Jay R., and Kazanjian, Robert K.:Strategy Implementation:
mance During M-form Reorganization and Recovery Time: The Structure, Systems, and Process, West, St. Paul, MN. 1986.
Effects of Prior Strategy and Implementation Speed.Academy of Galunic, D. Charles, and Eisenhardt, Kathleen: The Evolution of Management Journal37 (February 1994): 153–166.
Intracorporate Domains: Divisional Charter Losses in
High-Tech-Markides, Constantinos, and Williamson, Peter: Corporate Diversifi-nology, Multidivisional Corporations. Organization Science 7
cation and Organizational Structure: A Resource-Based View. (1996): 255–282.
Academy of Management Journal39 (1996): 340–367. Greening, Daniel, and Johnson, Richard: Do Managers and Strategies
Michel, John G., and Hambrick, Donald C.: Diversification Posture Matter? A Study in Crisis.Journal of Management Studies33 (1996):
and Top Management Team Characteristics.Academy of Manage-25–52.
ment Journal35 (March 1992): 9–37. Hambrick, Donald: Fragmentation and the Other Problems CEOs
Miller, Danny: Stale in the Saddle: CEO Tenure and the Match Have With Their Top Management Teams.California Management
Between Organization and Environment.Management Science37 Review37 (1995): 110–128.
(January 1991): 34–52. Hambrick, Donald C.: Top Management Groups: A Conceptual
Inte-Miller, Danny, and Friesen, Peter H.:Organizations: A Quantum View, gration and Reconsideration of the Team Label, inResearch in
Prentice Hall, Englewood Cliffs, NJ. 1984. Organizational Behavior, L.L. Cummings and B.M. Staw, eds., JAI
Press, Greenwich, CT. 1994. Norburn, David, and Birley, Susan: The Top Management Team and Corporate Performance. Strategic Management Journal 9 (May-Hambrick, Donald C., and Fukutomi, Gregory D.S.: The Seasons of
June 1988): 225–237. a CEO’s Tenure. Academy of Management Review 16 (October
1991): 719–742. Pfeffer, Jeffrey:Research in Organizational Behavior, JAI Press, Green-wich, CT. 1983.
Hambrick, Donald C., and Mason, Phyllis: Upper Echelons: The
Organization as a Reflection of its Top Managers.Academy of Roberts, Peter, and Greenwood, Royston: Integrating Transaction Cost and Institutional Theories: Toward a Constrained-Efficiency Management Review9 (April 1984): 193–206.
Framework for Understanding Organizational Design Adoption. Hannan, Michael T., and Freeman, John: Structural Inertia and
Orga-Academy of Management Review22 (1997): 346–373. nizational Change.American Sociological Review49 (April 1984):
149–164. Rumelt, Richard, Schendel, Dan, and Teece, David: Fundamental
Issues in Strategy, Harvard University Press, Boston, MA. 1994. Hill, Charles W., Hitt, Michael A., and Hoskisson, Robert E.:
Cooper-ative Versus Competitive Structures in Related and Unrelated Rumelt, Richard:Strategic, Structure, and Economic Performance, Har-vard University Press, Cambridge, MA. 1974.
Diversified Firms.Organization Science3 (November 1992): 501–
521. Russo, Michael V.: The Multidivisional Structure as an Enabling
Device: A Longitudinal Study of Discretionary Cash as a Strategic Hill, Charles W., and Hoskisson, Robert E.: Strategy and Structure
in the Multiproduct Firm.Academy of Management Review 12 Resource.Academy of Management Journal34 (September 1991): 718–733.
(April 1987): 331–341.
Hoskisson, Robert E.: Multidivisional Structure and Performance: SAS Institute, Inc.,SAS-STAT User’s Guide, Version 6, SAS Institute, Inc., Cary, NC. 1990.
The Contingency of Diversification Strategy.Academy of
trative Scenarios.Journal of Applied Psychology65 (June 1980): Venkatraman, N.: The Concept of Fit in Strategy Research: Toward Verbal and Statistical Correspondence.Academy of Management 249–260.
Review14 (July 1989) 423–444. Thomas, Anisya S., Litschert, Robert J., and Ramasway, Kannan:
The Performance Impact of Strategy-Manager Coalignment: An Virany, Beverly, Tushman, Michael L., and Romanelli, Elaine: Execu-tive Succession and Organization Outcomes in Turbulent Envi-Empirical Examination.Strategic Management Journal12 (October
1991): 509–522. ronments: An Organizational Learning Approach.Organization
Science3 (February 1992): 72–91. Thompson, James:Organizations in Action, McGraw-Hill, New York.
1967. Weick, Karl E.: Organization Redesign as Improvisation, in
Organiza-tion Change and Redesign: Ideas and Insights for Improving Perfor-Tushman, Michael L., and Keck, Sara L.: Environmental and
Organi-mance, G. Huber and W. Glick, eds., Oxford University Press, zation Context and Executive Team Characteristics: An
Organiza-New York. 1993. tion Learning Approach. Working Paper, Columbia University
Graduate School of Business, New York, 1989. Wiersema, Margarethe F., and Bantel, Karen A.: Top Management Team Demography and Corporate Strategic Change.Academy of Tushman, Michael L., and Romanelli, Elaine: Organizational
Evolu-Management Journal35 (March 1992): 91–121. tion: A Metamorphosis Model of Convergence and Reorientation,
inResearch in Organizational Behavior, L. Cummings and B. Staw, Williamson, Oliver E.:Markets and Hierarchies: Analysis and Antitrust Implications, MacMillan, New York. 1975.
