AN ANALYSIS OF THE RELATIONSHIP BETWEEN STRATEGIC LINKAGE CONSISTENCY AND FIRM PERFORMANCE IN HOSPITAL SYSTEMS.

ABSTRACT

The study offers a preliminary perspective on strategic linkage consistency. The study examines two separate operationalizations of consistency by exploring: (1) the level at which linkages occur (centralized system-level vs. decentralized hospital-level); and (2) the type of linkage integration which is employed (vertical ownership vs. virtual contracting). The study examines three separate measures of hospital performance by exploring: (1) market share; (2) profitability; and (3) productivity. The study proposes a positive relationship between strategic linkage consistency and firm performance. The study proposes, also, that environmental conditions moderate this relationship so that in less competitive managed care environments strategic linkage consistency has less of an impact on hospital performance, while in highly competitive managed care environments strategic linkage consistency has more of an impact on hospital performance. The study tests these propositions in the health care industry. The study results support the first proposition and marginally support the second.

The impact of strategic fit on firm performance has received considerable attention in the organization behavior and management strategy fields. The literature addresses this issue from a number of different perspectives. One stream of research considers strategic fit in terms of the match between strategy and environment and claims that organizations with strategy-environment alignment tend to be more profitable and higher performing than those lacking such alignment (Shortell & Zajac 1990; Lamont, Marlin & Hoffman 1993). A second line of research discusses strategic consistency in terms of a fit between strategy and structure and argues that successful organizations are those which employ organizational designs consistent with their chosen strategies (Chandler 1962; Rumelt 1974; Horovitz & Thietart 1982; Harrison, Hall & Nargundkar 1993; Ginn, Young & Beekun 1995; Sriram & Anikeeff 1995). A third tract of research contemplates strategic fit from numerous multivariate perspectives and maintains that congruence among various components (i.e., competencies, resources, environmental threats, environmental opportunities, technology, systems, structure, process, management, staff, behaviors, values) determines organizational prosperity (Andrews 1971; Miles & Snow 1978; Jauch & Osborn 1981; Peters & Waterman 1982; Randolph & Dess 1984; Venkatraman & Camillus 1984; Kimberly & Zajac 1985; Harrison, Hall & Nargundkar 1993). Finally, a fourth group of studies addresses strategic fit by looking at the implications of matching genetic strategies over changing times (Zajac & Shortell 1989; Parnell 1994).

While there has been a great deal written about the ramifications of matching strategy to these sorts of elements, there has been a scarcity of research on the implication of matching strategy to strategy within a complex organization. Researchers, however, have acknowledged the importance of conducting intrastrategy research (Lenz 1980; Fredrickson & Mitchell 1984; Shortell, Morrison & Robbins 1985; Gulati & Lawrence 1996; Van de Ven 1997) and the present study intends to begin to fill this void by proffering the concept of strategic linkage consistency.

THEORETICAL GROUNDING

The present study explores the notion of linkage consistency explicitly and builds on the logic that it is not enough for a complex organization simply to have all of the right components, it is also necessary that there be a proper fit among them. The present study argues that strategic linkage consistency can positively impact firm performance through several mechanisms. For one, when a complex firm has linkage consistency, it becomes able to realize economies of scope and synergy among organizational sub-units which, in turn, allow it enhanced efficiencies in the production of products and delivery of services (Panzar & Willig 1981; Nayyar & Kazanjian 1993). Furthermore, having linkage consistency enables a firm to maintain a smooth flow of resources, achieve an easy coordination of activities, establish a sense of "collaborative know-how" and preserve a consistency to organizational routines (Pennings, Barkema & Douma 1994; Simonin 1997). Finally, having linkage consistency facilitates a coordinated information flow (Dodgson 1993), cultivating cross-functional integration (Lei, Hitt & Goldhar 1996), creating a context of understanding (Dove 1996), enhancing organizational learning and administrative processes and, consequently, leading to improved performance (Nass 1994). Taken together, these explanations lead to the following proposition:

Proposition 1: Strategic linkage consistency has a positive impact on firm performance.

Strategic linkage consistency might, however, have different implications for organizational success in dissimilar environments. Structural contingency theory (Lawrence & Lorsch 1967; Drazin & Van de Ven 1985) argues that environmental factors provoke organizations to pursue different organizational designs. Fredrickson and Mitchell (1984) suggest that "a firm's performance will be affected if the level of comprehensiveness in its strategic decision process is not consistent with the stability of its environment" (p. 405). The present study builds off of these works, proposing that different types of environments call for different degrees of linkage consistency. More specifically, the present study asserts that:

Proposition 2: Strategic linkage consistency has a more positive impact on firm performance for a firm in a turbulent environment than for a firm in a stable environment.

EMPIRICAL SETTING

The present study examines the relationship between strategic linkage consistency and firm performance by studying the health care industry. The health care industry is chosen for two reasons. First, it is an industry which affords a variety of health care systems/networks which make ideal case studies in which to examine issues of linkage consistency. Second, the industry exhibits variance in the magnitude and rapidity of change across geographically dispersed regions of the United States (American Practice Management 1993; Shortell et al. 1996; Bums et al. 1997) and so it allows for a consideration of both more and less turbulent environments.

The strategic linkages targeted by this study are threefold--hospital-physician linkages, hospital-insurance linkages, and hospital-service-mix linkages. This study examines whether hospitals within a health care system/network that pursue consistency across these linkage activities outperform those that pursue less consistent linkage strategies. The present study argues that if a hospital is consistent in the way in which it is linking up with its physicians, insurance and service-mix components then the hospital will reap positive performance results. While recognizing that these linkages do not necessarily represent the complete spectrum of ties which a hospital may posses, they are strategically important linkages likely to influence hospital performance. Coherence of these linkages can be investigated specifically in terms of: (1) the level at which each link occurs (centralized system-level vs. decentralized hospital-level); and (2) the type of linkage integration which is employed (vertical ownership vs. virtual contracting).

In health services research and the organizational literature at large, the centralization/decentralization issue has been widely examined (Starkweather 1981; Miller & Friesen 1984; Miller 1987; Shortell, Morrison & Friedman 1990) and has gained acceptance as a crucial area of study. The bulk of the attention on this topic has focused on comparing centralization to decentralization and outlining the pros, cons, and ramifications of each approach. An equally significant area of interest involves determining the different repercussions of vertical vs. virtual integration for health care systems/networks (Tennyson & Fottler 1997). Vertical integration refers to common ownership of stages of the production process. Virtual integration, by contrast, involves related stages of the health care value chain that are combined via long-term exclusive contracts, operating agreements, franchise deals, formal affiliations, partial equity investments, joint ventures, etc., without implying ownership by a sole body. After observing several decades of attachment activities, researchers are now contemplating the appropriateness of various integration strategies for the present day health care industry (Conrad & Shortell 1996; Robinson & Casalino 1996; Walston, Kimberly & Bums 1996) and, as with centralization/decentralization, the literature again focuses around a debate as to whether vertical ownership or virtual contracting is most effective for health care organizations.

While the present study acknowledges the importance of these debates, it ignores the specifics of these polarized arguments and examines the implications for each of them. In other words, the study looks at consistency of linkage strategies and asks, "as long as one linkage strategy (centralization, decentralization, vertical ownership, or virtual contracting) is being pursued uniformly, what does this imply for performance?" If the present research finds that linkage consistency does not have a positive impact on performance, then resolving the debate over which type/level of link is preferred becomes a moot point. If, on the other hand, the study finds that linkage consistency does have a positive impact on performance, then the ensuing step becomes to assess which type/level of link works best in various situations.

The purpose of the present research is to examine the extent to which linkage consistency is positively related to performance. As argued above, by having consistent linkage strategies, a hospital's resources flow smoothly, economies of scope result, organizational learning occurs and, therefore, it is expected that:

Hypothesis 1a: Hospitals that are pursuing consistent linkage strategies at the system-level (centralized or non-centralized) or consistent linkage strategies at the hospital-level (decentralized or non-decentralized) across their physician, insurance, and service-mix linkage activities outperform those that are pursuing a mixture of system-level and hospital-level links.

Hypothesis 1b: Hospitals that are pursuing consistent linkage strategies of ownership (or non-ownership) or consistent linkage strategies of contracting (or non-contracting) across their physician, insurance, and service-mix linkage activities outperform those that are pursuing a mixture of ownership and contracting links.

In graphic form, these hypotheses can be represented as drawn below. Hospitals in Figure 1a, with all solid links (indicating consistency with tight forms of strategic relationships, such as consistent centralization or consistent vertical ownership) between themselves and their product/service components, or hospitals in Figure 1b, with all dotted links (indicating consistency with loose forms of strategic relationships, such as consistent decentralization or consistent virtual contracting) between themselves and their product/service components will outperform hospitals in Figure 1c, with a combination of various textured links (indicating inconsistency with a mixture of tight and loose strategic relationships).

[Figure 1 ILLUSTRATION OMITTED]

Linkage consistency might, however, have different implications for hospitals' successes in dissimilar environments. More specifically, it is possible that the competitiveness of the market in which a hospital is situated influences the ability of strategic linkage consistency to impact hospital performance. In other words, in environments of instability (high managed care competition) linkage fit may have more of an impact on hospital performance than in environments of stability (low managed care competition) because in turbulent environments, uniform coherence can fundamentally impact organizational survival. As Shortell (1995) observes "the greater the prevalence of managed care activities in a given market [i.e., high managed care competition], the greater the likelihood that tighter, more-integrated alliance structures [i.e., those with linkage consistency] will outperform looser, less-integrated structures [i.e., those without linkage consistency]" (p. 191). The present study tests this argument more formally as:

Hypothesis 2: Managed care market competition moderates the proposed relationship between linkage consistency and hospital performance, so that in less competitive managed care environments linkage consistency has less of an impact on hospital performance while in highly competitive managed care environments linkage consistency has more of an impact on hospital performance.

DATA AND METHODS

DATA SAMPLE

The present study combines information from three data sets. The American Hospital Association (AHA) 1994 Annual Survey provides detailed data on hospitals' product/service component linkages as well as basic descriptive information. An historical data set created by the Health Care Investment Analysts (HCIA) provides 1994 hospital financial performance statistics on selected measures based on collected Medicare financial statements. The 1993 AHA/Northwestern University Quality Survey adds information on the competitiveness of the managed care markets in which the hospitals are located. Merger of these three data sets yields a sample of 1,291 hospitals available for analysis. These 1,291 hospitals tend to be larger in bed size and more likely to be not-for-profit than is true of the hospital universe at large (n = 5,264). In addition, these 1,291 hospitals represent only hospitals which are members of health care systems(1) or networks(2) (both terms are combined under the general label of system for the remainder of the study) since, by definition, freestanding facilities do not have relevance to the present study's discussion of linkage activities.

VARIABLES AND MEASURES (see Appendix for further detail)

The dependent variable in this study is hospital performance. The financial management literature on health care offers some guidance as to measures of hospital performance which have demonstrated face validity in prior research (Cleverley 1993). The present study employs three of these recommended measures, each of which targets a different dimension of hospital performance: productivity, profitability, and market share (Coyne 1985; Ginn, Young & Beekun 1995; Van Peursem, Pratt & Lawrence 1995; Zeller, Stanko & Cleverley 1996)(3). By examining multiple dependent variables as performance indicators, the study aims to shed light on which measures of hospital functioning are explicitly tied to strategic linkage consistency. For instance, it may be that a hospital which displays linkage consistency manifests positive performance primarily in terms of increased market share because linkage consistency enables the hospital to explicitly declare and target its niche. At the same time, linkage consistency may influence this hospital only negligibly in terms of increased productivity because linkage consistency may be related only to demandside/revenue and not to supply-side/cost performance indicators.

The key independent variables for the study are the operationalizations of centralized (system-level), decentralized (hospital-level), owned, and contracted strategies as they apply to the hospital-physician, hospital-insurance, and hospital-service-mix linkages. The strategies of centralization and decentralization are determined based on the level at which the link occurs: centralized means it transpires at the overall health system-level; non-centralized means it does not transpire at the overall health system-level; decentralized indicates that it happens at the hospital-level; and non-decentralized indicates that it does not happen at the hospital-level. Since it is possible, by definition, for a hospital to be simultaneously centralized and decentralized (i.e., to have strategic linkages at both the system-level and hospital-level), it is necessary to differentiate in this dual manner. In other words, while centralized and non-decentralized connote similar notions linguistically, they should not be considered equivalent terms for the purposes of this study. The strategies of ownership and contracting are operationalized in a similar fashion. Ownership is constructed as proprietorship over product/service components while non-ownership is the absence of such proprietorship. Similarly, but not reciprocally, contracting is defined as allying with another entity that is not owned by the hospital while non-contracting is defined as the absence of such an alliance. In other terms, ownership represents vertical integration while contracting represents virtual integration.

The definitions of these strategies are incorporated into measures of the hospitals' linkages. The hospital-physician link is operationalized based on hospitals' reported participation in eight different types of arrangements with physicians. The hospital-insurance link measure is taken from responses to questions about hospitals' involvement in each of two different forms of managed care arrangements. The hospital-service-mix link is operationalized into three separate measures: Basic, Outreach, and Long Term Care services.(4) In all, there are five linkages which are defined as individual independent variables: physician, insurance, basic services, outreach services, and long term care services. These five individual variable measures are transformed into measures of dispersion by standardizing the individual variables through a z-score transformation (in order to map all five linkage dimensions onto a comparable scale), grouping them into the four strategic categories (centralization (system-level), decentralization (hospital-level), ownership, and contracting), and creating a measure of dispersion for each of the four strategic categories (by comparing across the z-scores for the five linkage dimensions within each category and calculating the variance among them). A low score on these newly created dispersion measures is an indication of linkage consistency.

Additional independent variables are included in the analyses. A measure of managed care market competition based on the work of Westphal, Gulati and Shortell (1997)(5) is used to explore the moderating impact of the environment on the relationship between strategic linkage consistency and hospital performance. Control variables of interest in the present study are hospital bed size, ownership, location and teaching activity. Size is important because of its potential for creating economies of scope and scale in the delivery of services through the purchase of vertical and horizontal relations and for its potential for effectively reducing transaction expenses by decreasing the costs of writing and enforcing contracts (Williamson 1981). Ownership is relevant because of the implicitly different incentives, cultures, and strategies associated with different forms of hospital control. Location is meaningful because of its possible influence on the range and method of service delivery. Degree of teaching activity is considered because of its capacity to influence the degree of integration in the mix of services undertaken by a hospital (Bazzoli et al. 1999).

ANALYSES

The analyses are run in three stages. The first stage is a preliminary examination involving descriptive statistics and bivariate correlations between all of the proposed measures to examine issues of collinearity. The second step explores the main effects of linkage consistency on hospital performance. The third step tests an interaction model accounting for the interplay between managed care competition and linkage consistency.

RESULTS

Stage 1: Descriptive Information and Correlations

Means and ranges are provided in Table 1 in order to demonstrate how the 1,291 hospital sample breaks down descriptively. Table I also provides correlation matrices for all analytical variables. Potential collinearity among the independent variables became non-problematic once they were transformed into measures of dispersion, and the resulting correlations among the four categories of strategic linkage consistency are reasonable. Potential collinearity problems among the control variables were addressed with two minor adjustments. The not-for-profit and church affiliated categories were combined into one variable (non-profit ownership) because of their high correlation and their original derivation from the same scale item. Degree of teaching activity was dropped because it was correlated with hospital bed size and because a sense of hospitals' research/teaching capacities was already incorporated into the study via hospitals' responses to questions about whether their service mixes included the high technology functions (i.e., bum centers) that would be expected of teaching/research facilities.

                            TABLE 1
 DESCRIPTIVE INFORMATION AND PEARSON CORRELATION COEFFICIENTS
                    FOR ANALYTIC VARIABLES

          Variables                  Mean       S.D.

1.  System-level dispersion          0.61       0.74
2.  Hospital-level dispersion        0.80       0.54
3.  Ownership dispersion             0.72       1.65
4.  Contracting dispersion           0.86       1.50
5.  Size                           233.33     198.21
6.  Location                         0.68       0.47
7.  Non-profit owned                 0.72       0.45
8.  Government owned                 0.15       0.36
9.  Investor owned                   0.14       0.34
10. Competition                      9.70       7.00
11. FTE's-per 100 admits             5.43       3.02
12. Operating margin                 3.65       8.11
13. Market percentage               35.99      32.16

           Variables                  1             2

1.  System-level dispersion
2.  Hospital-level dispersion      0.12(***)
3.  Ownership dispersion           0.28(***)      0.15(***)
4.  Contracting dispersion         0.07(**)       0.01
5.  Size                           0.11(***)      0.26(***)
6.  Location                       0.17(***)      0.08(***)
7.  Non-profit owned               0.13(***)      0.04
8.  Government owned              -0.05           0.03
9.  Investor owned                -0.11(***)     -0.08(***)
10. Competition                    0.14(***)      0.01
11. FTE's-per 100 admits           0.04           0.13(***)
12. Operating margin               0.01          -0.02
13. Market percentage             -0.12(***)      0.02

           Variables                  3             4

1.  System-level dispersion
2.  Hospital-level dispersion
3.  Ownership dispersion
4.  Contracting dispersion         0.01
5.  Size                           0.07(***)      0.02
6.  Location                       0.08(***)      0.01
7.  Non-profit owned               0.04           0.06(**)
8.  Government owned              -0.01          -0.02
9.  Investor owned                -0.03          -0.07(**)
10. Competition                    0.05          -0.03
11. FTE's-per 100 admits           0.01           0.01
12. Operating margin               0.04           0.03
13. Market percentage             -0.03          -0.02

           Variables                  5             6

1.  System-level dispersion
2.  Hospital-level dispersion
3.  Ownership dispersion
4.  Contracting dispersion
5.  Size
6.  Location                       0.46(***)
7.  Non-profit owned               0.21(***)      0.14(***)
8.  Government owned              -0.12(***)     -0.24(***)
9.  Investor owned                -0.16(***)      0.07(***)
10. Competition                    0.17(***)      0.38(***)
11. FTE's-per 100 admits           0.16(***)      0.04
12. Operating margin               0.08(***)      0.04
13. Market percentage             -0.16(***)     -0.44(***)

           Variables                  7             8

1.  System-level dispersion
2.  Hospital-level dispersion
3.  Ownership dispersion
4.  Contracting dispersion
5.  Size
6.  Location
7.  Non-profit owned
8.  Government owned              -0.66(***)
9.  Investor owned                -0.63(***)     -0.17(***)
10. Competition                    0.05          -0.20(***)
11. FTE's-per 100 admits          -0.06(**)       0.17(***)
12. Operating margin              -0.10(***)     -0.01
13. Market percentage             -0.01           0.16(***)

           Variables                  9             10

1.  System-level dispersion
2.  Hospital-level dispersion
3.  Ownership dispersion
4.  Contracting dispersion
5.  Size
6.  Location
7.  Non-profit owned
8.  Government owned
9.  Investor owned
10. Competition                    0.14(***)
11. FTE's-per 100 admits          -0.11(***)     -0.03
12. Operating margin               0.14(***)     -0.02
13. Market percentage             -0.16(***)     -0.24(***)

           Variables                  11            12

1.  System-level dispersion
2.  Hospital-level dispersion
3.  Ownership dispersion
4.  Contracting dispersion
5.  Size
6.  Location
7.  Non-profit owned
8.  Government owned
9.  Investor owned
10. Competition
11. FTE's-per 100 admits
12. Operating margin              -0.16(***)
13. Market percentage             -0.07(**)       0.03

(**) p [is less than or equal to] 0.05

(***) p [is less than or equal to] 0.01

Stage 2: Main Model (Hypotheses [1.sub.a] and [1.sub.b]) Hypothesis [1.sub.a]

As presented in Table 2, at the system-level, linkage consistency (consistently centralized or consistently non-centralized) has a direct impact on market share in the predicted magnitude and direction. Neither productivity nor profitability are impacted by system-level linkage consistency. At the hospital-level, linkage consistency (consistently decentralized or consistently non-decentralized) has a positive impact on productivity. Neither market share nor profitability are impacted by hospital-level linkage consistency.

                              TABLE 2
                  REGRESSION RESULTS FOR MAIN MODEL

                         Market Share Model

Linkage consistency
 System-level      -2.27(**)
   dispersion     (-1.11)
 Hospital-level                  2.16
   dispersion                   (1.52)
 Ownership                                   0.10
   dispersion                               (0.48)
 Contracting                                             -0.59
   dispersion                                            (0.53)

Environment
 Managed care      -0.31(**)   -0.34(**)    -0.34(***)   -0.34(***)
   competition     (0.13)      (0.13)       (0.13)       (0.13)

Control variables(a)
 Size               0.01        0.01         0.01         0.01
                   (0.01)      (0.01)       (0.01)       (0.01)
 Location         -28.11(***) -28.44(***)  -28.56(***)  -28.52(***)
                   (2.08)      (2.07)       (2.08)       (2.07)
 Non-profit owned  -3.00       -2.93        -3.10        -3.01
                   (2.31)      (2.31)       (2.31)       (2.31)
 Investor owned   -14.49(***) -13.75(***)  -14.02(***)  -14.12(***)
                   (3.13)      (3.13)       (3.13)       (3.13)

Constant           61.70(***)  59.34(***)   60.85(***)   61.38(***)
                   (2.29)      (2.51)       (2.27)       (2.30)

Overall F          60.49(***)  60.03(***)   59.60(***)   59.86(***)

[R.sup.2]           0.22        0.22         0.22         0.22
                  (28.48)     (28.50)      (28.52)      (28.51)

                         Profitability Model
                         (Operating Margin)
Linkage consistency
 System-level       0.32
   dispersion      (0.31)
 Hospital-level                 -0.55
   dispersion                   (0.42)
 Ownership                                   0.20
   dispersion                               (0.13)
 Contracting                                              0.22
   dispersion                                            (0.15)

Environment
 Managed care      -0.07(*)     -0.06(*)    -0.06(*)     -0.06(*)
   competition     (0.03)       (0.03)      (0.03)       (0.03)

Control variables(a)
 Size               0.01(***)    0.01(***)   0.01(***)    0.01(***)
                   (0.00)       (0.00)      (0.00)       (0.00)
 Location          -0.17        -0.12       -0.14        -0.11
                   (0.58)       (0.58)      (0.58)       (0.58)
 Non-profit owned  -0.46        -0.48       -0.44        -0.47
                   (0.65)       (0.65)      (0.65)       (0.65)
 Investor owned     3.65(***)    3.51(***)   3.61(***)    3.61(***)
                   (0.86)       (0.82)      (0.86)       (0.86)

Constant            2.91(***)    3.42(***)   2.93(***)    2.84(***)
                   (0.64)       (0.70)      (0.64)       (0.64)

Overall F           7.70***      7.81(***)   7.89(***)    7.89(***)

[R.sup.2]           0.03         0.03        0.04         0.04
                   (7.99)       (7.99)      (7.99)       (7.99)

                        Productivity Model
                       (FTEs per 100 admits)

Linkage consistency
 System-level       0.09
   dispersion      -0.12
 Hospital-level                 0.43(***)
   dispersion                  (0.16)
 Ownership                                  -0.01
   dispersion                               (0.05)
 Contracting                                              0.02
   dispersion                                            (0.05)

Environment
 Managed care      -0.01       -0.01        -0.01        -0.01
   competition     (0.01)      (0.01)       (0.01)       (0.01)

Control variables(a)
 Size               0.00(***)   0.00(***)    0.00(***)    0.00(***)
                   (0.00)      (0.00)       (0.00)       (0.00)
 Location           0.05        0.09         0.07         0.07
                   (0.21)      (0.21)       (0.21)       (0.21)
 Non-profit owned  -1.54(***)  -1.50(***)   -1.53(***)   -1.54(***)
                   (0.24)      (0.24)       (0.24)       (0.24)
 Investor owned    -1.96(***)  -1.93(***)   -1.98(***)   -1.98(***)
                   (0.32)      (0.32)       (0.32)       (0.32)

Constan             6.19(***)   5.91(***)    6.22***      6.20***
                   (0.24)      (0.26)       (0.32)       (0.24)

Overall F          14.83(***)  16.07(***)   14.72(***)   14.75(***)

[R.sup.2]           0.07        0.07         0.06         0.06
                   (2.92)      (2.92)       (2.92)       (2.92)

(a) Government owned is the excluded category.

(*) p [is less than or equal to] 0.10

(**) p [is less than or equal to] 0.05

(***) p [is less than or equal to] 0.01

Hypothesis [1.sub.b]

Hospitals consistently pursuing strategies of ownership (or non-ownership) or contracting (or non-contracting) in their linkage activities do not excel over those pursuing non-consistent linkage strategies on any of the three performance indicators.

While not hypothesized explicitly, a significant direct effect is demonstrated for managed care competition in the market share and profitability models implying that hospital performance generally suffers in highly competitive managed care environments.

Stage 3: Interaction Model (Hypothesis 2)

Hypothesis 2

The findings presented in Table 3 demonstrate marginal support for the hospital-level productivity finding and no support for the system-level market share finding. In both cases the results are positive in sign but weak in magnitude. By and large, then, Hypothesis 2 must be rejected. Although not included explicitly in the second hypothesis, a significant interaction effect is found for ownership dispersion in the operating margin model indicating that, by itself, ownership linkage consistency is not related to hospital performance, but, in the presence of managed care competition, it is related to hospital performance.

                            TABLE 3
              REGRESSION RESULTS FOR INTERACTION MODEL

                         Market Share Model

Interaction term
  System-level dispersion x     0.09
   Managed care competition    (0.13)

  Hospital-level dispersion x          -0.14
   Managed care competition            (0.23)

  Ownership dispersion x                       -0.09
   Managed care competition                    (0.06)

  Contracting dispersion x                              0.07
   Managed care competition                            (0.10)

Linkage consistency
  System-level dispersion      -3.42(*)
                               (1.94)

  Hospital-level dispersion             3.50
                                       (2.60)

  Ownership dispersion                          1.08
                                               (0.83)

  Contracting dispersion                                -1.14
                                                        (0.92)

                         Profitability Model
                         (Operating Margin)

Interaction term
  System-level dispersion x    -0.02
   Managed care competition    (0.04)

  Hospital-level dispersion x           0.00
   Managed care competition            (0.06)

  Ownership dispersion x                        0.04(**)
   Managed care competition                    (0.02)

  Contracting dispersion x                              0.03
   Managed care competition                            (0.03)

Linkage consistency
  System-level dispersion       0.51
                               (0.54)

  Hospital-level dispersion            -0.58
                                       (0.72)

  Ownership dispersion                         -0.20
                                               (0.23)

  Contracting dispersion                                0.00
                                                       (0.26)

                        Productivity Model
                      (FTE's per 100 admits)

Interaction term
  System-level dispersion x    -0.01
   Managed care competition    (0.01)

  Hospital-level dispersion x           0.04(*)
   Managed care competition            (0.02)

  Ownership dispersion x                        0.00
   Managed care competition                    (0.01)

  Contracting dispersion x                             -0.01
   Managed care competition                            (0.01)

Linkage consistency
  System-level dispersion       0.19
                               (0.20)

  Hospital-level dispersion             0.06
                                       (0.27)

  Ownership dispersion                          0.01
                                               (0.09)

  Contracting dispersion                                0.08
                                                       (0.09)

DISCUSSION

Overall, linkage consistency with regard to strategies of ownership (or non-ownership) and contracting (or non-contracting) do not appear tied to any of the study's three measures of performance while linkage consistency with regard to system-level strategy (centralization or non-centralization) is tied to market share and linkage consistency with regard to hospital-level strategy (decentralization or non-decentralization) is tied to productivity. The competitiveness of the managed care environment appears only marginally to impact the relationship observed between hospital-level strategic linkage consistency and productivity and not at all for system-level linkage strategic consistency and market share.

With regard to market share, it appears that consistently centralizing or consistently not centralizing linkages at the system-level helps a hospital to define the market niche which it is targeting. Linkage consistency at the system-level helps system member hospitals focus their attention on system-wide entrepreneurial opportunities, facilitates hospitals' process of learning about markets, makes resource use easier to deploy, and enables system member hospitals to rally behind a uniform goal. In contrast, when a mixture of centralized and non-centralized linkage strategies is employed, market share gains are limited and hospitals suffer, possibly because of disincentives that are generated by such a "mixed" system.

The positive relationship with regard to productivity is substantiated by the social cognition perspective which argues that shared understandings yield shared mental models which result in better work processes (Stasser & Titus 1985; Cannon-Bowers, Salas & Converse 1993). A similar result was found by Meyer, Goes and Brooks (1993) who note that:

   ...members of high-performing hospitals...share sets of explicit values
   about what their organizations are doing and why... [and these] shared
   values inspire commitment and elicit cooperation... (p. 98).

It is important to point out that these two findings, however, are level-specific. It, also, is useful to discuss why they do not transfer from one level of analysis to the other. For instance, the finding that linkage consistency increases market share does not hold at the hospital-level. This makes sense, intuitively, because while market power of individual hospitals can vary, it is the aggregate market share of the system as a whole, not of its component hospitals individually, that makes for a high or low-performing system. Also, the finding that hospital-level linkage consistency increases productivity does not hold at the system-level. Again, this is understandable because day-to-day work routine issues make their impact more at the individual hospital-level than at the aggregate system-level. In summary, the differential findings for distinct levels of analysis may be explained by the unique dynamics occurring at each level.

Linkage consistency does not appear to have a significant impact on operating margin at either the system-level or the hospital-level. This may, in part, be due to hospitals' status as "loosely coupled" (Weick 1979) organizations represented by the relative separation of physicians as independent professionals from hospital employees. Given that physicians' decisions influence about 70-80 percent of hospital expenditures, this separation makes it more difficult to influence physician behavior in ways that might positively benefit the financial health of the hospital. One might expect to find a relationship between strategic linkage consistency and profitability in those hospitals with higher percentages of employed physicians. This is an area for further research.

The null finding that linkage consistency for strategies of ownership (or non-ownership) and contracting (or non-contracting) does not impact hospital performance can be interpreted in several ways. One explanation is that the dispersion measures of linkage strategies used in the study are specified too generally. To this end, it may be that because the study did not differentiate between ownership consistency and non-ownership consistency the results of the two types of linkage consistencies are canceling each other out. A second explanation for the null finding may be that it is truly better to have a mixture of ownership and contracting linkages. For instance, benefits might accrue to a hospital which enables medical professionals to maintain their historical autonomy through contracting mechanisms while, at the same time, reaps economy of scale benefits by buying service-mix components. These types of arrangements are referred to as "hybrids" (Conrad & Shortell 1996) and may, in the long run, prove to be the most viable arrangements. Finally, a third explanation for why linkage consistency for strategies of ownership and contracting is not found to impact hospital performance might be that these mechanisms impact different factors of performance than the three that were used as dependent variables in this study. To this end, return on equity and/or cash flow measures might be explored as additional hospital performance indicators.

The study yields less powerful support for the first set of hypotheses than anticipated. The weak results may be due to the fact that the analyses were unable to take into account hospitals' capabilities to execute strategy. To this end, while executing consistent strategy may have positive benefits for a health care system, not all hospitals may be capable of accomplishing and/or coordinating such an achievement. The fact that hospitals tend to be loosely coupled organizations (Weick 1979) may make it difficult for them to execute global strategies across entire health care systems. Alternatively, the weak results may be on account of a loss of interpretability due to the use of statistical factor analysis.

With regard to the second hypothesis, the competitiveness of the managed care environment generally does not exert moderating effects. While it is possible that hospitals in highly competitive managed care environments do, indeed, suffer when there is a lack of strategic linkage consistency, this effect is not particularly noticeable on the financial performance measures explored in the present study.

It is worth noting that strategic linkage consistency appears to explain only a small portion of the variance in financial performance differences across hospitals, however, the low [R.sup.2] values do not compromise the study's findings. In keeping with other models of "fit" as described by Golden (1989), the study's strategic linkage consistency theory is:

   ...less concerned with the overall explanatory power of the equation...and
   more... [concerned with] the significance of the term that specifically
   reflects the fit between or among a set of variables (Venkatraman,
   1989:3)... (p. 79).

Furthermore, although the [R.sup.2] values are small, the power of the effects is not. If, for example, values (using the sample means for all variables) are plugged into the first numeric column of Table 2, a comparison between a hospital with a poor system-level strategic linkage consistency score (one standard deviation above the mean) and a hospital with a good system-level strategic linkage consistency score (one standard deviation below the mean) amounts to about a 3.5 percent increase in market share. Similarly, if values (using the sample means for all variables) are plugged into the 10th numeric column of Table 2, a comparison between a hospital with a poor hospital-level strategic linkage consistency score (one standard deviation above the mean) and a hospital with a good hospital-level strategic linkage consistency score (one standard deviation below the mean) amounts to about a 0.5 percent increase in productivity. Given that the intention of this study was not to fully predict all of the determinants of financial performance, but, rather, to provide preliminary investigation of the direct and indirect relationships between strategic linkage consistency and firm performance, the modest amount of variance explained in the models is not surprising. To fully predict financial performance would require considerably more elaborate models, in order to include factors like the economic market conditions and state regulatory practices in which health care organizations operate. While including these types of additional control variables in the regression equations would likely serve to explain more variance in financial performance, such additional variables are not related to the study's theory of strategic linkage consistency and their inclusion in the models would compromise their parsimony.

CONCLUSION

The present study offers several insights. The findings speak to the issue of the appropriate organizational architecture for health care systems by suggesting that a holistic approach to centralization/decentralization strategy seems important while a more mixed hybrid-type approach to ownership/contracting strategy may be viable. The study also offers initial evidence of the relationship between strategic linkage consistency and firm performance. The findings suggest that, on selected measures, organizations which have coherent and well integrated strategic linkages tend to outperform organizations that do not. Furthermore, the degree to which this relationship matters appears to be marginally influenced by variation in environmental factors relative to market competition. If supported by further research, these findings might have considerable consequence for the management of organizations because they suggest a prescriptive formula for improving business operations.

The present study's findings prompt a number of avenues for future research. First, since there is some suggestion that linkage consistency does matter, future research could examine the specific types of different linkage consistency that have the greatest impact on performance. Second, primary data collection might be undertaken to examine additional links i.e., pharmaceuticals, nursing homes) which would be useful for strengthening the linkage consistency measures and expanding the study's arguments. An additional step might be to incorporate qualitative performance measures in order to investigate any intangible benefits which strategic linkage consistency might have for firms. Fourth, future studies should include additional outcomes measures in order to address the relationship between strategic linkage consistency and the quality of health care delivery by considering not just the economy and efficiency of operations, but, also, speaking to the effectiveness of the services provided. Fifth, field work/case study methodologies might be employed to gather more detail on the mechanisms through which strategic linkage consistency impacts firm performance. Finally, the relationship between and among strategic linkage consistency, internal firm capabilities and degree of alignment with external environmental domains is an area for further work.

The present study's findings should be considered within the context of several limitations. While the strategic integration activities of hospitals with regard to the linkages measured in the study are not likely to change markedly from year to year (Alexander 1990; Alexander, Fennell & Halpern 1993), the potential flaws of using cross-sectional data must be noted. There is a need for lagged performance data so that longitudinal analyses can be used to specify causality arguments more precisely and, also, to shed light on the long-term impact of strategic linkage consistency on firm performance. Also, additional measures of strategic linkage consistency should be developed in order to enhance reliability and validity of measurement. Finally, because the study is limited to one industry, generalization beyond health care should be done with caution. Additional industries should be examined in future research endeavors to see if similar results are obtained.

APPENDIX

DEFINITIONS AND VARIABLE DESCRIPTIONS

Panel A: Definitions

Closed Physician-Hospital Organization (CPHO): A physician-hospital organization that restricts physician membership to those practitioners who meet criteria for cost effectiveness and/or high quality (American Hospital Association, 1994).

Equity Model (EQUITY): Allows established practitioners to become shareholders in a professional corporation in exchange for tangible and intangible assets of their existing practices (American Hospital Association, 1994).

Foundation Model (FOUNDATION): A corporation, organized either as a hospital affiliate or subsidiary, which purchases both the tangible and intangible assets of one or more medical group practices. Physicians remain in a separate corporate entity but sign a professional services agreement with the foundation (American Hospital Association, 1994).

Group Practice Without Walls (GPWW): A hospital sponsors the formation of, or provides capital to physicians to establish, a "quasi" group to share administrative expenses while remaining independent practitioners (American Hospital Association, 1994).

Health Maintenance Organization (HMO): An entity that provides, offers or arranges for coverage of designated health services needed by plan members for a fixed, prepaid premium (United HealthCare Corporation, 1994:38).

Independent Practice Association (IPA): An IPA is a legal entity that holds managed care contracts. The IPA then contracts with physicians, usually in solo practice, to provide care either on a fee-for-services or capitated basis. The purpose of an IPA is to assist solo physicians in obtaining managed care contracts (American Hospital Association, 1994).

Integrated Salary Model (ISM): Physicians are salaried by the hospital or another entity of a health system to provide medical services for primary care and specialty care (American Hospital Association, 1994).

Management Services Organization (MSO): A corporation, owned by the hospital or a physician/hospital joint venture, that provides management services to one or more medical group practices. The MSO purchases the tangible assets of the practices and leases them back as part of a full-service management agreement, under which the MSO employs all non-physician staff and provides all supplies/administrative systems for a fee (American Hospital Association, 1994).

OBGYN: Obstetrics and gynecological services.

Open Physician-Hospital Organization (OPHO): A joint venture between the hospital and all members of the medical staff who wish to participate. The physician-hospital organization can act as a unified agent in managed care contracting, own a managed care plan, own and operate ambulatory care centers or ancillary services projects, or provide administrative services to physician members (American Hospital Association, 1994).

Preferred Provider Organization (PPO): A program in which contracts are established with providers of medical care ... [whereby there are] better benefits ... for services received from [these] preferred providers, thus encouraging covered persons to use these providers ... [but still allowing] benefits for non-participating providers' services (United HealthCare Corporation, 1994:60).

Panel B: Variable Descriptions

Dependent Variables

                                  Performance

Full Time Equivalents per 100     Productivity measure (lower
Admissions, Case Mix Adjusted     means better performing)

Operating Margin (without other   Profitability measure--operating
operating revenue)                income divided by revenues
                                  (higher means better performing)

Market Percent (based on          Market share measure (higher
hospital's market area            means better performing)
admissions)

Individual Independent Variables

                                  System-Level

Insurance                         System-level PPO and/or HMO
                                  -- dummy variable 0 or 1

MDs                               System-level IPA and/or GPWW
                                  and/or OPHO and/or CPHO and/or
                                  MSO and/or ISM and/or EQUITY
                                  and/or FOUNDATION--dummy
                                  variable 0 or 1

Basic Services                    System-level factor analyzed
                                  specialty, cardiology, surgery,
                                  diagnostic, emergency,
                                  [hitech.sup.a], pediatric and
                                  OBGYN/women's health
                                  services--percentage from
                                  0 to 100

Outreach Services                 System-level factor analyzed
                                  combination of averaged variables
                                  for community, geriatric,
                                  psychiatric and other
                                  services--percentage from
                                  0 to 100

LTC Services                      System-level factor analyzed
                                  combination of averaged variables
                                  for home health and long term
                                  care services--percentage from
                                  0 to 100

                                  Hospital-Level

Insurance                         Hospital-level PPO and/or
                                  HMO--dummy variable 0 or 1

MDs                               Hospital-level IPA and/or GPWW
                                  and/or OPHO and/or CPHO and/or
                                  MSO and/or ISM and/or EQUITY
                                  and/or FOUNDATION--dummy variable
                                  0 or 1

Basic Services                    Hospital-level factor analyzed
                                  combination of averaged variables
                                  for general acute, specialty,
                                  cardiology, surgery, diagnostic,
                                  emergency, [hitech.sup.a],
                                  pediatric and OBGYN/women's
                                  health services--percentage
                                  from 0 to 100

Outreach Services                 Hospital-level factor analyzed
                                  combination of averaged variables
                                  for community, geriatric,
                                  psychiatric and other
                                  services--percentage from
                                  0 to 100

LTC Services                      Hospital-level factor analyzed
                                  combination of averaged variables
                                  for home health and long term
                                  care services--percentage from
                                  0 to 100 for home health and
                                  long term care
                                  services--percentage from 0 to 100

                                  Ownership

Insurance                         System and/or hospital-level
                                  HMO--dummy variable 0 or 1

MDs                               System and/or hospital-level
                                  ISM and/or EQUITY and/or
                                  FOUNDATION model--dummy variable
                                  0 or 1

Basic Services                    Average of hospital-level and
                                  system-level factor analyzed
                                  combination of averaged variables
                                  for general acute, specialty,
                                  cardiology, surgery, diagnostic,
                                  emergency, [hitech.sup.a],
                                  pediatric and OBGYN/women's
                                  health services--percentage
                                  from 0 to 100

Outreach Services                 Average of hospital-level and
                                  system-level factor analyzed
                                  combination of averaged variables
                                  for community, geriatric,
                                  psychiatric and other
                                  services--percentage from 0
                                  to 100

LTC Services                      Average of hospital-level and
                                  system-level factor analyzed
                                  combination of averaged
                                  variables for home health and
                                  long term care
                                  services--percentage from 0 to 100

Insurance                         System and/or hospital-level
                                  PPO--dummy variable 0 or 1

MDs                               System and/or hospital-level
                                  IPA and/or GPWW and/or OPHO
                                  and/or CPHO and/or MSO
                                  model--dummy variable 0 or 1

Basic Services                    Contracted/joint venture-level
                                  factor analyzed combination of
                                  averaged variables for general
                                  acute, specialty, cardiology,
                                  surgery, diagnostic, emergency,
                                  [hitech.sup.a], pediatric and
                                  OBGYN/women's health
                                  services--percentage from
                                  0 to 100

Outreach Services                 Contracted/joint venture-level
                                  factor analyzed combination of
                                  averaged variables for community,
                                  geriatric, psychiatric and other
                                  services--percentage from

                                  0 to 100

LTC Services                      Contracted/joint venture-level
                                  factor analyzed combination of
                                  averaged variables for home
                                  health and long term care
                                  services--percentase from
                                  0 to 100

Transformed Independent Variables

                                  Linkage Consistency

System-Level Dispersion           Dispersion measure expressing the
                                  variance of values for z-scores
                                  of the 5 measures of system-level
                                  products/services (lower means
                                  more consistent)

Hospital-Level Dispersion         Dispersion measure expressing the
                                  variance of values for z-scores
                                  of the 5 measures of
                                  hospital-level products/services
                                  (lower means more consistent)

Ownership Dispersion              Dispersion measure expressing the
                                  variance of values for z-scores
                                  of the 5 measures of owned
                                  products/services (lower means
                                  more consistent)

Contracting Dispersion            Dispersion measure expressing
                                  the variance of values for
                                  z-scores of the 5 measures of
                                  contracted products/services
                                  (lower means more consistent)

Control Variables

                                  Controls

Size                              Total number of beds

Ownership                         3 separate indicators are used
                                  to represent government ownership,
                                  investor ownership, and non-profit
                                  status--dummy variables 0 or 1

Location                          Urban or rural--dummy variable
                                  0 or 1

Moderating Variable

                                  Moderator

Competitiveness of managed        Factor analyzed scale compiled
care environment                  from the mean of the following
                                  (when at least 2 of the 4 items
                                  is complete): (1) the number of
                                  HMOs in the same market area;
                                  (2) the percent of all patients
                                  paid on a capitated basis;
                                  (3) the number of hospitals who
                                  are direct competitors Of the
                                  focal hospital; and (4) the
                                  perceived intensity of competition

Including hitech as part of basic services was an empirically-derived decision. Theoretically, however, it might be more appropriate to consider the services included in the hitech category (i.e., trauma center, transplant services, extracorporeal shock-wave lithotripter (ESWL) and open heart surgery) as non-standard procedures even for reasonably technology-competent hospitals.

(1) A health care system is defined as "a corporate body that may own and/or manage health provider facilities or health-related subsidiaries as well as non-health-related facilities including freestanding facilities and/or subsidiary corporations" (American Hospital Association 1994).

(2) A health care network is defined as "a group of hospitals, physicians, other providers, insurers and/or community agencies that work together to coordinate and deliver a broad spectrum of services to their community" (American Hospital Association 1994).

(3) Financial management measures can be used to reflect the economy (indicating cost per number of inputs applied to manage health care activities), efficiency (indicating input/output ratios), or effectiveness (indicating outcome oriented measures such as enhanced health status of patients) of a health care organization (Van Perusem, Pratt & Lawrence 1995). The measures we employ in our study should be interpreted only as indicators of economy (full time equivalents per 100 admissions, case mix adjusted) and efficiency (operating margin) and not as indicators of effectiveness since we are unable to include variables to proxy for quality of care in our regressions. Nevertheless, measures of economy and efficiency play important roles for labor cost control and accountability reporting in health care organizations, both of which are critical to ensuring the maintenance of financially viable operations (Van Peursem, Pratt & Lawrence 1995; Zeller, Stanko & Cleverley 1996).

(4) Measures of the hospital-service mix links are consolidated from an initial 78 services into a 15 dimension service-mix classification scheme (general acute, pediatric, OBGYN/women's health, surgery, specialty, psychiatric, long term care, home health, geriatric, cardiology, diagnostic, emergency, community care, hi-technology, and other services) based on research by Dranove, Shanley and Simon (1992) and then factor analyzed to create three general service-mix constructs: Basic, Outreach, and Long Term Care (LTC) services. Principal Components factor analysis with OBLIMIN rotation of 15 service mix classification averages load at an absolute value of .40 or greater on three factors with Eigenvalues greater than 1.0: Basic (Eigenvalue = 6.53), Outreach (Eigenvalue = 1.28), and LTC (Eigenvalue = 1.13).

(5) Westphal, Gulati and Shortell (1997) used principal components factor analysis with OBLIMIN rotation of four items from the 1993 AHA/Northwestern University Quality Survey: (1) the number of HMOs in the same market area; (2) the percent of all patients paid on a capitated basis; (3) the number of hospitals who are direct competitors of the focal hospital; and (4) the perceived intensity of competition. These items load on a single factor (Eigenvalue = 1.64) which is then used as the measure of managed care competition in each hospital's market environment.

REFERENCES

Alexander, J.A. (1990) The Changing Character of Hospital Governance, Chicago: Hospital Research and Education Trust.

Alexander, J.A., M.L. Fennell & M.T. Halpern (1993) "Leadership Instability in Hospitals: The Influence of Board-CEO Relations and Organizational Growth and Decline," Administrative Science Quarterly, 38(1):74-99.

American Hospital Association (1994) Annual Survey of Hospitals: Data Base Documentation Manual, Chicago: Author.

American Practice Management (1993) Health System Market Profile Characteristics, New York: Author.

Andrews, K.R. (1971) The Concept of Corporate Strategy, Homewood, IL: Richard D. Irwin, Inc.

Bazzoli, G.J., S.M. Shortell, N. Dubbs, C. Chan & P. Kralovec (1999) "A Taxonomy of Health Care Networks and Systems: Bringing Order Out of Chaos," Health Services Research, 33(6):1683-1717.

Burns, L.R., G.J. Bazzoli, L. Dynan & D.R. Wholey (1997) "Managed Care, Market Stages, and Integrated Delivery Systems: Is There a Relationship?" Health Affairs, 16(6):204-217.

Cannon-Bowers, J.A., E. Salas & S. Converse (1993) "Shared Mental Models in Expert Team Decision Making" in N.J. Castellan, Jr. (ed.) Individual and Group Decision Making: Current Issues, Hillsdale, NJ: Lawrence Erlbaum Associates, Inc.

Chandler, A.D., Jr. (1962) Strategy and Structure, Garden City, NJ: Doubleday.

Cleverley, W.O. (1993) The 1993 Almanac of Hospital Financial and Operating Indicators, Columbus: Center for Healthcare Industry Performance Studies.

Conrad, D.A., & S.M. Shortell (1996) "Integrated Health Systems: Promise and Performance," Frontiers of Health Services Management, 13(1):3-40.

Coyne, J.S. (1985) "Measuring Hospital Performance in Multiinstitutional Organizations Using Financial Ratios," Health Care Management Review, 10(4):35-42.

Dodgson, M. (1993) "Organizational Learning: A Review of Some Literatures," Organizational Studies, 14(3) :375-394.

Dove, R. (1996) "Agile Knowledge Transfer: Reusable, Reconfigurable, Scalable," Production, 108(12):16-17.

Dranove, D., M. Shanley & C. Simon (1992) "Is Hospital Competition Wasteful?" Rand Journal of Economics, 23(2):247-262.

Drazin, R. & A. Van de Ven (1985) "Alternative Forms of Fit in Contingency Theory," Administrative Science Quarterly, 30(4):51 4-539.

Fredrickson, J.W. & T.R. Mitchell (1984) "Strategic Decision Processes: Comprehensiveness and Performance in an Industry With an Unstable Environment," Academy of Management Journal, 27(2):399-423.

Ginn, G., G.J. Young & R.I. Beekun (1995) "Business Strategy and Financial Structure: An Empirical Analysis of Acute Care Hospitals," Hospital & Health Services Administration, 40(2): 191-209.

Golden, B.R. (1989) Business Strategy as a Critical Contingency for the Corporate-Division Relationship, Unpublished doctoral dissertation, Northwestern University, Evanston, IL.

Gulati, R., & P. Lawrence (1996) Organizing Vertical Networks: A Design Perspective, Manuscript submitted for publication.

Harrison, J.S., E.H. Hall, Jr. & R. Nargundkar (1993) "Resource Allocation as an Outcropping of Strategic Consistency: Performance Implications," Academy of Management Journal, 36(5): 1026-1051.

Horovitz, J.H. & R.A. Thietart (1982) "Strategy, Management Design and Firm Performance," Strategic Management Journal, 3(1):67-76.

Jauch, L.R. & R.N. Osborn (1981) "Toward an Integrated Theory of Strategy," Academy of Management Review, 6(3):491-198.

Kimberly, J.R. & E.J. Zajac (1985) "Strategic Adaptation in Health Care Organizations: Implications for Theory and Research," Medical Care Review, 42(2):267-302.

Lamont, B.T., D. Marlin & J.J. Hoffman (1993) "Porter's Generic Strategies, Discontinuous Environments, and Performance: A Longitudinal Study of Changing Strategies in the Hospital Industry," Health Services Research, 28(5):623-640.

Lawrence, P.R. & J.W. Lorsch (1967) "Differentiation and Integration in Complex Organizations," Administrative Science Quarterly, 12(1):1-47.

Lei, D., M.A. Hitt & J.D. Goldhar (1996) "Advanced Manufacturing Technology: Organizational Design and Strategic Flexibility," Organization Studies, 17(3):501-523.

Lenz, R.T. (1980) "Environment, Strategy, Organization Structure and Performance: Patterns in One Industry," Strategic Management Journal, 1 (3):209-226.

Meyer, A.D., J.B. Goes & G.R. Brooks (1993) "Organizations Reacting to Hyperturbulence," in G.P. Huber & W.H. Glick (eds.) Organizational Change and Redesign: Ideas and Insights, New York: Oxford University Press.

Miles, R.E. & C.C. Snow (1978) Organizational Strategy, Structure, and Process, New York: McGraw-Hill.

Miller, D. (1987) "Strategy Making and Structure: Analysis and Implications for Performance," Academy of Management Journal, 30(1):7-32.

Miller, D. & P.H. Friesen (1984) Organizations: A Quantum View, Englewood Cliffs, NJ: Prentice-Hall, Inc.

Nass, C. (1994) "Knowledge or Skills: Which do Administrators Learn from Experience?" Organization Science, 5 (1) :38-50.

Nayyar, P. & R.K. Kazanjian (1993) "Organizing to Attain Potential Benefits from Information Asymmetries and Economies of Scope in Related Diversified Firms," Academy of Management Review, 18(4):735-759.

Panzar, J.C. & R.D. Willig (1981) "Economies of Scope," American Economic Review, 71 (2):268-272.

Parnell, J.A. (1994) "Strategic Consistency versus Flexibility: Does Strategic Change Really Enhance Performance?" American Business Review, 12(2):22-29.

Pennings, J.M., H. Barkema & S. Douma (1994) "Organizational Learning and Diversification," Academy of Management Journal, 37(3):608-640.

Peters, T.J. & R.H. Waterman, Jr. (1982) In Search of Excellence: Lessons

from America's Best Run Companies, New York: Harper & Row.

Randolph, A.W. & G.G. Dess (1984) "The Congruence Perspective of Organization Design: A Conceptual Model and Multivariate Research Approach," Academy of Management Review, 9(1): 114-127.

Robinson, J.C. & L.P. Casalino (1996) "Vertical Integration and Organizational Networks in Health Care," Health Affairs, 15(1):7-22.

Rumelt, R.P. (1974)Strategy, Structure, and Economic Performance, Boston: Harvard Graduate School of Business Administration.

Shortell, S.M. (1995) "Commentary" in A.D. Kaluzny, H.S. Zuckerman & T.C. Ricketts (eds.) Partners for the Dance: Forming Strategic Alliances in Health Care, Ann Arbor: Health Administration Press.

Shortell, S.M., R.R. Gillies, D.A. Anderson, K.M. Erickson & J.B. Mitchell (1996) Remaking Health Care in America, San Francisco: Jossey-Bass.

Shortell, S.M., E.M. Morrison & B. Friedman (1990) Strategic Choices for America's Hospitals: Managing Change in Turbulent Times, San Francisco: Jossey-Bass.

Shortell, S.M., E.M. Morrison & S. Robbins (1985) "Strategy Making in Health Care Organizations: A Framework and Agenda for Research," Medical Care Review, 42(2):219-266.

Shortell, S.M. & E.J. Zajac (1990) "Health Care Organizations and the Development of the Strategic Management Perspective" in S. Mick (ed.) Innovations in Health Care Delivery: New Insights into Organizational Theory, San Francisco: Jossey-Bass.

Simonin, B.L. (1997) "The Importance of Collaborative Know-How: An Empirical Test of the Learning Organization," Academy of Management Journal, 40(5): 1150-1174.

Sriram, V. & M.A. Anikeeff (1995) "Strategic Consistency and Performance: An Analysis of Real Estate Developers," Journal of Managerial Issues, 7(4):435-448.

Starkweather, D.B. (1981) Hospital Mergers in the Making, Ann Arbor: Health Administration Press.

Stasser, G. & W. Titus (1985) "Pooling of Unshared Information in Group Decision Making: Biased Information Sampling During Discussion," Journal of Personality and Social Psychology, 48(6):1467-1478.

Tennyson, D.H. & M.D. Fottler (1997) "Does System Membership Enhance Financial Performance in Hospitals?" in L.N. Dosier & J.B. Keys (eds.) Best Papers Proceedings, Statesboro, GA: Academy of Management.

United HealthCare Corporation. (1994) The Managed Care Resource: The Language of Managed Health Care and Organized Health Care Systems, Minnetonka, MN: Author.

Van de Ven, A. (1997) Managing Pluralistic Organizations in Health Care, Paper presented to the Department of Organization Behavior, Northwestern University, Evanston, IL.

Van Peursem, K.A., M.J. Pratt & S.R. Lawrence (1995) "Health Management Performance: A Review of Measures and Indicators," Accounting, Auditing & Accountability Journal, 8(5):34-70.

Venkatraman, N. (1989) "The Concept of Fit in Strategy Research: Toward Verbal and Statistical Correspondence," Academy of Management Review, 14(3):423-444.

Venkatraman, N. & J.C. Camillus (1984) "Exploring the Concept of "Fit" in Strategic Management," Academy of Management Review, 9(3):513-525.

Walston, S.L., J.R. Kimberly & L.R. Bums (1996) "Owned Vertical Integration and Health Care: Promise and Performance," Health Care Management Review, 21 (1):83-92.

Weick, K.E. (1979) The Social Psychology of Organizing, New York: Random House.

Westphal, J.D., R. Gulati & S.M. Shortell (1997) "Customization or Conformity? An Institutional and Network Perspective on the Content and Consequences of TQM Adoption," Administrative Science Quarterly, 42(2):366-394.

Williamson, O. (1981) "The Economics of Organization: The Transaction Cost Approach," American Journal of Sociology, 87(3):548-577.

Zajac, E.J. & S.M. Shortell (1989) "Changing Generic Strategies: Likelihood, Direction, and Performance Implications," Strategic Management Journal, 10(5):413-430.

Zeller, T.L., B.B. Stanko & W.O. Cleverley (1996) "A Revised Classification Pattern of Hospital Financial Ratios," Journal of Accounting and Public Policy, 15(2): 161-182.

The authors wish to thank Gloria Bazzoli, Robin Gillies, Ranjay Gulati, Paul Hirsch, Myron Roomkin, and Ed Zajac for their help in shaping this paper. The following groups provided data used in the study: the American Hospital Association Hospital Research and Educational Trust, the American Hospital Association Data Survey Group, the Joint Commission on Accreditation of Health Care Organizations, and the Health Care Investment Analysts.

Address for correspondence: Nicole L. Dubbs, Division of Health Policy & Management, Joseph L. Mailman School of Public Health, Columbia University, 600 West 168th Street, Room 606, New York, NY 10032 USA.