Cultural determinants of customer- and learning-oriented value systems and their joint effects on firm performance

Journal of Business Research 58 (2005) 1340 – 1352

Cultural determinants of customer- and learning-oriented value systems and their joint effects on firm performance Cengiz Yilmaz*, Lutfihak Alpkan1, Ercan Ergun2 Gebze Yuksek Teknoloji Enstitusu, Isletme Fakultesi, Cayirova Fabrikalar Yolu No: 101 41400 Gebze, Kocaeli, Turkey Received 1 November 2003; received in revised form 1 March 2004; accepted 1 June 2004

Abstract The authors examine the effects of firm-level reflections of two societal–culture factors, collectivism and power distance, as well as organizational cultural strength, on the development of customer- and learning-oriented value systems in organizations. The joint effects of customer and learning orientations on components of firm performance are also investigated. The main theses of the study are that a customer- and learning-oriented organizational value system is more likely to develop and improve firm performance when (1) complemented by appropriate societal–culture factors and (2) supported by a bstrongQ organizational culture. The results obtained from a sample of manufacturing firms in Turkey generally support the study theses. The study uses multiple informants from each firm and measures cultural factors and performance assessments from the viewpoints of different informants, thus achieving a greater level of measure specificity while at the same time eliminating same-source bias. Implications of the study findings for managers and researchers are discussed. D 2004 Elsevier Inc. All rights reserved. Keywords: Firm performance; Collectivism; Power distance; Customer orientation; Learning orientation

1. Introduction The topic of organizational culture and its impact on firm performance has become a core area of interest for organizational researchers and practitioners (e.g., Denison and Mishra, 1995; Schein, 1985). Specifically, two cultural value systems, a customer orientation and a learning orientation, have been suggested as the defining characteristics of modern organizations and crucial drivers of superior firm performance (e.g., Argyris and Schon, 1978; Sinkula et al., 1997). A customer orientation (hereafter, CO) is manifested as the organization and its members focus

* Corresponding author. Tel.: +90 262 653 8497x1470; fax: +90 262 654 3224. E-mail addresses: [email protected] (C. Yilmaz)8 [email protected] (L. Alpkan)8 [email protected] (E. Ergun). 1 Tel.: +90 262 653 8497x1222; fax: +90 262 654 3224. 2 Tel.: +90 262 653 8497x1321; fax: +90 262 654 3224. 0148-2963/$ - see front matter D 2004 Elsevier Inc. All rights reserved. doi:10.1016/j.jbusres.2004.06.002

their efforts on understanding and satisfying customers (Deshpande et al., 1993). A learning orientation (hereafter, LO), on the other hand, reflects the extent to which the organization values the acquisition, improvement, transfer, and utilization of knowledge (Slater and Narver, 1995). While closely related, CO and LO exert distinct (as well as synergistic) effects on firm performance (Hult et al., 2001; Sinkula et al., 1997). Managers in a variety of industries therefore face the complex challenges of (1) building customer- and learning-oriented organizational value systems and (2) using such capabilities to achieve superior organizational performance. To accomplish both challenges, managers need to consider two distinct mechanisms that are influential in the shaping of an organization’s culture. The first mechanism concerns the impacts of the broader society on organizational culture. Societal–culture factors exert a strong influence on the cultural orientations of organizations and may even become the core elements of an organization’s culture (Schein, 1985). This is largely because

C. Yilmaz et al. / Journal of Business Research 58 (2005) 1340–1352

individuals bring into the organization’s culture the values they acquired earlier in their societies (Hofstede et al., 1990). To the extent that an organization is composed of people from a singular sociocultural identity (e.g., a national culture or subcultural elements within a nation), the cultural values of that particular societal group establish a basic ground, a cultural foundation to be acted upon, for managerial attempts to move the organization towards a more customer- and learning-oriented stance (Newman and Nollen, 1996). The second mechanism concerns the fact that organizational members acquire, identify with, or internalize the specific values of the organization largely through the socialization processes at the workplace (e.g., Pascale, 1985). Thus, through effective workplace socialization, managers can generate a congruence of customerand learning-oriented values across the organization. In fact, factors that facilitate the socialization process, such as management support, facilitative leadership, and internal communication, are seen as the strongest drivers of CO and LO (e.g., Conduit and Mavondo, 2001; Slater and Narver, 1995). Research attempting to provide guidance on how to build and utilize CO and LO should therefore consider both: (1) firm-level reflections of societal–culture factors and (2) factors related to the socialization processes within the firm. In view of this, the present study focuses on two societal–culture factors highly relevant to the emergence of customer- and learning-oriented organizational value systems, collectivism and power distance (Hofstede, 1980), as well as on the ultimate criterion for assessing the degree of within-firm socialization, that is, cultural strength (e.g., Denison, 2000). Collectivism refers to the subordination of personal interests to the interests of a larger work group, and power distance relates to the degree of asymmetry in power relationships recognized by organizational members (Hofstede, 1980). Cultural strength concerns (1) bwho and how many [of the organizational members] accept the dominant value set and (2) how strongly, deeply, or intensely the values are heldQ (Fisher, 1997, p. 47; Schein, 1985). Effective socialization processes lead to strong organizational cultures where all members share, coherently and intensely, a common set of beliefs and values (e.g., Schall, 1983). Our first research objective is to investigate how these three cultural factors facilitate or inhibit CO and LO in organizations. We then turn our attention into the effects of CO and LO on components of organizational performance. We examine the joint effects of CO and LO on financial and qualitative components of firm performance. In addition, regarding these effects on performance components, we adopt a contingency approach and posit that cultural strength not only facilitates or inhibits CO and LO but also (positively) moderates their effects on firm performances. Thus, overall, we investigate the theses that CO and LO are more likely to develop and achieve maximum effectiveness when they are (1) complemented by appropriate societal–

1341

culture factors and (2) supported by a bstrongQ organizational culture. To test these ideas, we use data collected from the managers and employees (N=1349) of 134 manufacturing firms in Turkey. The firms in this sampling context rely heavily on ethnic and/or geographical kinship in their recruitment policies. Therefore, each firm is a typical reflection of one of the various subcultures in Turkey, which presents a rare opportunity for examining the effects of societal-level cultural factors on organizational phenomena. We conduct our analyses at the firm level using multiple informants from each firm, which is considered a major plus for measurement quality. More important, we use top-level managerial assessments for measuring firm performance and examine its relationships with measures of cultural factors obtained from other informants in the firm, which eliminates concerns about same-source bias and adds substantially to the value of our research.

2. Proposed model and research hypotheses Fig. 1 provides the conceptual basis for the hypotheses we develop. The hypotheses sequentially consider (1) the effects of the societal–culture factors and cultural strength on CO and LO, (2) effects of CO and LO on components of firm performance, and (3) the moderating role of cultural strength within the relationships of CO and LO with the performance components. The framework also includes market dynamism, firm size, and industry as control variables affecting directly firm performances. Given the widespread of firms included in the sample, controlling for the effects of these factors enables to us separate out the impacts of culture on firm performances (cf. Slater and Narver, 1995; Liu et al., 2002). Each phase of the relationships in Fig. 1 is discussed in the following sections. 2.1. Societal–culture factors as bases of CO and LO Researchers have long recognized that the cultural values of the broader society in which an organization operates may have profound impacts on its culture (e.g., Gordon, 1985; Hofstede et al., 1990). Regarding the role of national culture, Hofstede et al. (1990) assert that individuals enter organizations as adults, with the bulk of their values acquired in early youth bfirmly in placeQ (p. 312). The societal–culture elements individuals bring into a company may therefore play a major role in the evolution of its specific culture, particularly in cases where the organization’s members share a common (sub)cultural background. What societal–culture factors are relevant to the development of CO and LO, then? Hofstede (1980) identifies four societal–culture dimensions—collectivism, power distance, masculinity, and uncertainty avoidance—to explain differences across nations. Each one of these dimensions characterizes how work is conducted in organizations

1342

C. Yilmaz et al. / Journal of Business Research 58 (2005) 1340–1352

Fig. 1. Hypothesized model.

(Chatman and Barsade, 1995) and therefore exerts influence on CO and LO. However, even an extended list of the societal–culture factors would fail to represent an exhaustive combination of the cultural factors affecting organizational values. In this research, we focus specifically on collectivism and power distance because these two factors bear more directly on internal integration issues within firms (Barkema and Vermeulen, 1997; Hofstede, 1980). Note that the emergence of CO and LO also relate primarily to internal organizational dynamics, such as quality of interpersonal interactions, open expression of views, information sharing, and cooperation. The conceptual domain captured jointly by collectivism and power distance is therefore crucial for understanding the emergence of CO and LO in organizations, as we discuss next. 2.1.1. Collectivism The cultural trait of collectivism captures the relative importance people give to the interests of a larger workgroup (e.g., coworkers) as opposed to personal interests (Wagner and Moch, 1986). Unlike individualistic cultures, which value independence and competition, collectivist cultures encourage the subordination of personal interests to the goals of a larger work group and put more emphasis on sharing and cooperation (Yilmaz and Hunt, 2001). Collectivist cultures give priority to supportive organizational practices, interpersonal connectedness, group solidarity, joint responsibility, and harmony (Doney et al., 1998; Newman and Nollen, 1996). Out of this cohesion grows a greater proclivity (than individualistic cultures) to (1)

exchange information and ideas, (2) support and assist each other, (3) discuss problems openly and constructively (Chen et al., 1998), and (4) develop commitment to the organization (Wasti, 2002). Therefore, the generation, dissemination, and utilization of knowledge and ideas, including those pertaining to the customers, are generally easier in collectivist organizations than in individualistic ones. Note that such learning and responsiveness to market information may also occur in individualistic cultures. However, in individualistic cultures, these actions are viewed as instrumental—i.e., people engage in such actions to reach their own personal goals. Therefore, the likelihood that such behaviors will be observed in individualistic cultures depends largely on the degree to which they are supported by controls and rewards (Chen et al., 1998). As a result, ceteris paribus, CO and LO are more likely to be observed in collectivist cultures than in individualistic cultures. H1. The greater the degree of collectivism, the greater is the level of (a) a customer orientation and (b) a learning orientation. 2.1.2. Power distance Power distance is the extent to which less powerful members of an organization accept and expect as much as their superiors that power is distributed unequally across hierarchical levels (Hofstede, 1980). A high degree of power distance leads to a less participative stance in decision making (Newman and Nollen, 1996), greater reliance on rules and procedures, and higher levels of subordinate

C. Yilmaz et al. / Journal of Business Research 58 (2005) 1340–1352

submissiveness (Van Oudenhoven, 2001). In such organizations, people will have little inclination to (1) take responsibilities outside the immediate scope of their jobs, (2) act on urgent marketplace information, or (3) provide individual input into strategy and planning. Furthermore, asymmetric power relationships are generally characterized as lacking informal communication patterns, which may (1) impede knowledge acquisition through exploration and generative learning (i.e., learning that occurs when the organization is willing to question its long-held assumptions) and (2) inhibit diffusion of knowledge within the organization (Slater and Narver, 1995). As a result, a high level of power distance have many negative implications for both CO and LO. H2. The greater the degree of power distance, the less is the level of (a) a customer orientation and (b) a learning orientation. 2.1.3. CO and LO: interrelationships and effects on firm performance While several authors have noted the strong association between CO and LO (e.g., Slater and Narver, 1995; Hurley and Hult, 1998), the question of whether CO is a causal driver of LO or vice versa; or, whether the relationship is truly bidirectional, is still under debate. Both value systems emphasize information processing and use, but they also differ from each other on several grounds. A customeroriented value system influences customer-related information processing activity and prioritizes its use in the strategic process. It is therefore reasonable to assert that CO facilitates LO. Alternatively, as Baker and Sinkula (1999) assert, LO is more inclusive than CO in terms of both the scope of the information processed (i.e., organization-wide processing all forms of information, including learning from mistakes) and the higher order examination of this activity (i.e., challenging long-held, bdeeperQ norms or assumptions and questioning even systems that bworkQ). Thus, LO could also contribute positively to CO. We therefore hypothesize a bidirectional relationship between the two value systems. H3. A positive (bidirectional) relationship exists between a customer orientation and a learning orientation. Next, regarding the effects of CO and LO on firm performance, we distinguish between the performance components that relate to external organizational outcomes, i.e., financial and market performance (profitability, sales growth, and market share), and internal organizational processes, i.e., qualitative firm performance (quality improvements, innovativeness, employee satisfaction, and employee commitment). Schein (1990) asserts that organizations need to cope with the dual problems of external adaptation versus integral integration. We hypothesize that CO and LO exert positive effects on both performance components, independently from each other, thus each enabling organizations to cope simultaneously with the two challenges identified by Schein (1990). Supporting

1343

these hypotheses, prior research suggests that firms with CO and LO have superior capability not only for providing customer value and responding effectively to opportunities but also for improving internal processes (Celuch et al., 2002; Senge, 1990; Slater and Narver, 1995, 2000). H4. The greater the level of a customer orientation, the greater is the level of (a) financial and market performance and (b) qualitative firm performance. H5. The greater the level of a learning orientation, the greater is the level of (a) financial and market performance and (b) qualitative firm performance. Notice, that our theoretical model and empirical procedures also enables exploration of the relative effects of the two value systems on each performance component. Specifically, we expect that (1) CO will have a stronger effect on financial and market performance than LO and (2) LO will have a stronger effect on qualitative performance than CO. These expectations are consistent with the view that CO and LO complement each other in terms of facilitating organizational success. Given the lack of supporting or confronting evidence in prior research, however, we do not develop formal hypotheses regarding these expectations. The aforementioned differences between the two value systems make these expectations worthy of exploring. A high level of CO, with its inherently external focus (Baker and Sinkula, 1999), should have immediate effects on the speed and the effectiveness of the firm’s responses to external contingencies, particularly to those relating to customer expectations, thereby facilitating sales and financial earnings. On the other hand, whereas a learning-driven organization should also have an external emphasis through its information acquisition processes, learning is primarily an internally driven process because it requires that information is disseminated across and utilized by the organization’s units. These properties of LO should make it a stronger predictor (than CO) of qualitative performance aspects, such as quality improvements, innovativeness, and favorable employee attitudes. 2.2. The role of cultural strength Another quality of organizational culture that is as important as the underlying traits is its degree of strength (e.g., Schein, 1985; Schall, 1983). Understanding the dominant values in the firm and their strength is essential not only for building CO and LO but also for effectively transforming these value systems into appropriate forms of collective action. In strong cultures, subculture members— individuals, groups, departments, and so on—share a common set of values (Harris, 1998). In addition, the dominant values are held so deeply and intensely in strong culture firms that individuals adhere to them with great commitment (Sathe, 1983). As such, a strong culture may or may not be consistent with the specific set of values that are

1344

C. Yilmaz et al. / Journal of Business Research 58 (2005) 1340–1352

suitable for the emergence of CO and LO, but it certainly connotes a solid, organization-wide agreement with regard to what is important or unimportant, appropriate or inappropriate, and right or wrong. The relationships between the construct of cultural strength and CO and LO is therefore contingent upon the dominant value set. A strong culture should (1) facilitate CO and LO if its dominant elements fit the aspects of CO and LO and (2) inhibit CO and LO if it fails to exhibit such a fit. For this reason, we do not develop formal hypotheses concerning the direct effects of cultural strength on CO and LO. We examine the nature of these relationships in our sampling context in an exploratory fashion. We do, however, develop hypotheses regarding the moderating effects of cultural strength on the relationships of CO and LO with performance indicators. Saffold (1988) asserts that, in order for a firm’s culture to influence performance outcomes, it needs to be strong and must possess distinctive traits. Strongly shared values improve the coordination of activities and goal alignment across organizational units (Deal and Kennedy, 1982). Accordingly, we suggest that the effects of CO and LO on firm performances will grow stronger as cultural strength increases. Effective transformation of customer- and learning-oriented values into appropriate strategies, processes, and collective actions requires agreement across organizational units (i.e., a strong culture). When such cohesiveness does not exist, the complex information-processing tasks required for the implementation of CO and LO, as well as the organization-wide utilization of such information in operational procedures, will be much harder to systematize. Indirect evidence for this view comes from research that connects group cohesiveness with greater employee commitment, prosocial and ethical behavior, and reduced job stress (e.g., Kidwell et al., 1997; Posner et al., 1985). Other relevant works reveal that firms with a long history (in comparison to younger firms where shared experiences are minimal) show stronger effects of CO and LO on effective information processing, innovativeness, and firm performance (Callantone et al., 2002; Sinkula, 1994). H6. The positive effect of a customer orientation on (a) qualitative firm performance and (b) financial and market performance is stronger when cultural strength is high than when cultural strength is low. H7. The positive effect of a learning orientation on (a) qualitative firm performance and (b) financial and market performance is stronger when cultural strength is high than when cultural strength is low.

3. Method The data were collected via face-to-face interviews (using standardized questionnaires) with multiple inform-

ants (top-level managers and at least five more employees) from 134 manufacturing firms located within the Kocaeli industrial district of Turkey. Firms from all major industrial areas in this district that have more than 50 employees (medium- and large-sized firms according to the formal classification system in Turkey) were included in the sampling frame. Highest level executives (e.g., general managers) of each of the 646 such firms in the sampling area were first contacted via phone to solicit their cooperation. After a series of contacts, 143 firms agreed to participate in the study. These firms were then visited by one of the researchers and structured interviews were conducted with their general managers. Managers of nine firms were not to be reached or decided not to participate in the study after seeing the questionnaire items, however. Thus, our firm-level sample size is 134, which corresponds to a response rate of 21%. The general managers responded to the questions pertaining to cultural factors and provided assessments of firm performance and market dynamism. Next, depending on firm size, a number of additional employees (ranging from 5 to 35) from each firm were interviewed. Specific attention was given to select these informants from different departments and disparate hierarchical levels, also to conduct the interviews in a private environment, to obtain a wide spectrum of views and sincere responses. Among these respondents, those holding a managerial position answered the same set of items with the general managers, and those at lower level positions responded to items measuring the cultural factors only. In total, 1349 usable questionnaires were obtained. 3.1. Sample characteristics The sample includes firms from 17 different industries (e.g., machinery and equipments, packaged food products, chemical products, household appliances, textiles, and automotive). Sixty-five of the participating firms operate predominantly in consumer markets, while 69 firms are in industrial markets. The average number of employees in the participating firms is 250 (S.D.=259). As to the individual respondents, 134 are top-level managers, 464 are senior or midlevel managers, 325 are white-collar employees, and 426 are blue-collar employees. Of the 1349 respondents, 1041 are male, 768 are university or higher level graduates, 412 are high-school graduates, and 169 are graduates of secondary or primary school. The average respondent age is 34 years (S.D.=12.4), and the average tenure in the organization is 8 years (S.D.=10). Nonresponse bias was assessed by comparing (1) the proportion of firms from each industry, (2) number of employees in each firm, and (3) respondent characteristics (i.e., age, gender, education, and organizational tenure) with the average industry figures of the sampling frame, and the appropriate t tests (pairwise or independent samples tests, depending on the nature of comparison) indicated no significant differences.

C. Yilmaz et al. / Journal of Business Research 58 (2005) 1340–1352

1345

maximum likelihood estimation procedure in Lisrel 8.3. Measurement analyses were conducted at the individual level; thus, the effective sample size is 1349. Prior to the estimation of the confirmatory measurement model, singlefactor exploratory factor analyses are conducted for each scale to assess unidimensionality. In each of these analyses, a single factor is extracted (using a cutoff point of eigenvalue=1), suggesting that our measurement scales are unidimensional. Next, consistent with our measurement theory, the 34 items measuring the cultural factors (collectivism, power distance, strength, CO, and LO) are hypothesized to load on five distinct factors in the measurement model. In addition, the five items measuring market and financial performance and the four items measuring qualitative performance are averaged to create composite indicants for each of these formative measures, which are then posited to load on two distinct performance factors in the measurement model. Measurement error terms for these performance indicants are set at 10% of their observed variances. This seven-factor model results in a significant chi-square statistic [v 2(575)=2593], as expected, given the large sample size. The resulting goodness-of-fit indices suggest that the model fits the observed covariances well, however [comparative fit index (CFI)=.90; goodness-of-fit index (GFI)=.90; root mean square residual (RMR)=.047; root mean square error of approximation (RMSEA)=.05]. In addition, all items load significantly on their respective constructs (with the lowest t value being 12.09), providing support for the convergent validity of measurement items. Finally, discriminant validity is obtained for all constructs because the variance extracted for each construct is greater than its squared correlations with other constructs (Fornell and Larcker, 1981). The resulting latent factor correlations are displayed in Table 1.

3.2. Measures All study measures use five-point Likert scales with anchors strongly disagree (=1) and strongly agree (=5), unless otherwise noted. The coefficient alpha estimates, Lisrel-based internal consistency estimates, and the average variance extracted in each measure are at or beyond the threshold levels suggested by Nunnaly (1978) and Fornell and Larcker (1981). Measurement items and the results of these reliability analyses are provided in Appendix A. Items measuring collectivism (7 items) and power distance (8 items) are selected from the scales used in Dorfman and Howell (1988), Erez and Earley (1987), Chen et al. (1998), Sigler and Pearson (2000), and Robbins and Mukerji (1994). Items in the original versions of these scales assess individual-level cultural orientations. Wordings of the items are modified slightly to emphasize that the evaluations reflect the respondents’ expectations regarding their respective organizations. Similarly, items measuring customer orientation (four items), learning orientation (eight items), and cultural strength (seven items) are selected from a larger set of items used by Denison (e.g., 2000) and Denison and Neale (1994) to measure aspects of organizational culture that are conceptually identical with the constructs of interest. As to firm performance, five items are used to measure the financial- and market-related performance, and four items are used to measure qualitative performance. All performance items assess the average level of firm performances within the preceding three years, using five-point scales anchored at much worse than competition (=1) and much better than competition (=5). Finally, the measure of market dynamism includes four items assessing the rate of changes in customer preferences, competitors’ strategies, product characteristics, and technology; firm size is measured by asking the number of employees in each firm; and industry is coded by asking the firms’ primary industrial area of operation.

3.4. Tests of hypotheses

3.3. Measure validation

Firm-level aggregates of the individual-level measures of the cultural constructs were used to test the hypotheses. This analysis was first done by creating individual-level composite scores for each construct and then by aggregating these

The measures were evaluated by conducting a confirmatory factor analysis of the item covariance matrix, using the Table 1 Descriptive statistics for the scales and construct correlations (1) Collectivism (2) Power distance (3) Cultural strength (4) Customer orientation (5) Learning orientation (6) Financial and market performance (7) Qualitative performance (8) Market dynamism (9) Number of employees (10) Industry

Mean

S.D.

1

2

3

3.66 2.76 3.56 3.76 3.76 3.44 3.81 3.50 250 –

.43 .55 .37 .47 .45 .75 .59 .42 259 –

1 .53 .63 .51 .66 .35 .45 .29 .02 .03

.53 1 .50 .35 .47 .23 .39 .13 .005 .025

.70 .54 1 .51 .78 .30 .53 .28 .03 .10

4 .52 .45 .61 1 .60 .36 .41 .47 .03 .03

5

6

7

8

9

.74 .57 .88 .69 1 .32 .51 .30 .01 .05

.32 .30 .41 .33 .42 1 .56 .02 .15 .06

.51 .51 .63 .52 .68 .65 1 .12 .14 .12

1 .24 .05

1

.24 .14

The figures above the diagonal are the latent factor correlations obtained from the measurement model. The figures below the diagonal are the correlations across the aggregated scales used as input in the path analyses.

1346

C. Yilmaz et al. / Journal of Business Research 58 (2005) 1340–1352

scores within each firm. In addition to the scales assessed in the measurement model, measures of market dynamism, firm size, and industry are also included in these analyses as control variables. As noted before, top-level managerial assessments of firm performances and market dynamism are used in our analyses, rather than the aggregated scores of all respondents, which diminishes concerns about same-source bias. Also provided in Table 1 are the correlations across these aggregated scales. For the direct effects in Fig. 1, we estimate the path coefficients linking the study constructs using the structural equations modeling methodology in Lisrel 8.3. In specifying this model, aggregated (firm-level) composite scores of the cultural constructs are used as single indicants, and the measurement error terms for these variables are fixed at one minus the estimated coefficient alpha of the (individuallevel) scale of each construct. Consistent with the procedure used in the measurement model, measurement error terms for the performance indicators, market dynamism, and firm size are fixed at 10% of their observed variances. In addition, consistent with the received theory, the two endogenous performance indicators are allowed to covary

in this analysis by letting the relevant model parameter (w 2,1) to be estimated freely. As expected, the chi-square statistic obtained from the estimation of this model is nonsignificant [v 2(12)=19.93], and all goodness-of-fit indices suggest a good model fit (CFI=.98; GFI=.97; SRMR= .038). The hypothesized nomological network of relationships explain 38%, 82%, 23%, and 39% of the observed variances in CO, LO, financial performance, and qualitative performance, respectively. The parameter estimates for the hypothesized paths are provided in Table 2. The results suggest that collectivism is positively related to CO (standardized path coefficient, c i =.34, Pb.01) and LO (c i =.24, Pb.01), while the estimated path coefficients linking power distance to CO and LO are both nonsignificant. Although not hypothesized formally, cultural strength is also found to exert strong positive effects on CO (c i=.32, Pb.01) and LO (c i=.71, Pb.01). In addition, consistent with that hypothesized, the bidirectional path linking CO and LO is positive and significant (w i=.13, Pb.01). As to effects on performance indicators, CO is found to exert a positive influence on financial performance (b i=.31, Pb.01) and LO is found

Table 2 Parameter estimates Hypothesized path

Nonstandardized parameter estimate

Direct effects analyses [v 2(12)=19.9; CFI=.98; GFI=.97; SRMR=.038] CollectivismYcustomer orientation .34 CollectivismYlearning orientation .24 Power distanceYcustomer orientation .01 Power distanceYlearning orientation .01 Cultural strengthYcustomer orientation .32 Cultural strengthYlearning orientation .71 Customer orientationYqualitative performance .10 Customer orientationYfinancial and market performance .31 Customer orientationX learning orientation .13 Learning orientationYqualitative performance .51 Learning orientationYfinancial and market performance .19 Qualitative performance X financial and market performance .36 Market dynamismYqualitative performance .04 Market dynamismYfinancial and market performance .13 Number of employeesYqualitative performance .13 Number of employeesYfinancial and market performance .15 IndustryYqualitative performance .01 IndustryYfinancial and market performance .01

Standardized parameter estimate .34 .24 .01 .01 .32 .71 .10 .31 .13 .51 .19 .36 .04 .13 .14 .17 .07 .05

Hypothesized path

Moderator

Standardized parameter estimate

Moderating effects analyses Customer orientationYqualitative performance

Cultural strength

Low .06 (t=0.30) High .17 (t=1.21) Low .00 (t= 0.01) High .45 (t=3.22)*** Low .59 (t=3.43)*** High .18 (t=1.17) Low .29 (t=1.56)* High .09 (t=0.54)

Customer orientationYfinancial and market performance

Cultural strength

Learning orientationYqualitative performance

Cultural strength

Learning orientationYfinancial and market performance

Cultural strength

* Pb.1 (one-tailed tests). ** Pb.05 (one-tailed tests). *** Pb.01 (one-tailed tests).

t Value 2.57*** 2.52*** 0.12 0.09 2.50*** 5.24*** 0.89 2.46*** 3.04*** 3.79*** 1.53* 6.62*** 0.49 1.35* 1.77** 1.97** 0.92 0.62 v 2(1) Diff.

0.20 4.26** 5.48** 0.63

C. Yilmaz et al. / Journal of Business Research 58 (2005) 1340–1352

to exert a positive influence on qualitative performance (b i=.51, Pb.01). Contrary to that hypothesized, but somewhat consistent with the expectation that CO and LO will have distinct effects on financial and qualitative performances, CO is found unrelated to qualitative performance and the path coefficient linking LO with financial and market performance is positive but only marginally significant (b i=.19, Pb.1). Finally, among the three control variables, only firm size is found to have significant effects on financial performance (b i=.17, Pb.05) and/or qualitative performance (b i=.14, Pb.05). Next, to assess the hypothesized moderating effects of cultural strength, we conduct two-group comparisons within the structural equation modeling methodology. The sample was split into two groups (high and low cultural strength) based on the median value of the strength variable (3.58). The paths linking CO and LO with performance indicators were then compared between the two groups in a step-bystep procedure. In the first step, the parameter from CO to financial performance was constrained to be equal between the low- and high-strength groups. This parameter was then allowed to be estimated freely. The difference in the chisquare statistics obtained from the two estimations is significant (Dm2=4.26, df=1, Pb.05), which suggests that high- and low-strength cultures differ in terms of the COYfinancial performance relationship. Tests following the same procedure suggest that the LOYqualitative performance link is also significantly different between the high- and low-culture-strength groups (Dm2=5.48, df=1, Pb.05). Contrary to our expectations, however, no significant differences exists between the two samples in terms of (1) the COYqualitative performance link and (2) the LOYfinancial performance link. Examining the coefficients for the significant moderating effects (see Table 2), we find the following: (1) the COYfinancial performance link is positive and significant in the high-strength group (b i=.45, Pb.01), and reduces to the point of nonsignificance in the low-strength group. That is, CO exerts a stronger positive effect on financial performance in strong culture firms than in weak culture firms. (2) The LOYqualitative performance link is positive and significant in the low-strength group (b i=.59, Pb.01), and reduces to the point of nonsignificance in the highstrength group. That is, in sharp contrast to that hypothesized, as cultural strength increases, the LOYqualitative performance link weakens.

4. Discussion Organizational scientists have long recognized that cultural elements are among the most stable predictors of firm performance (e.g., Senge, 1990). Cultural capabilities are generally hard to imitate (due to their socially complex and causally ambiguous nature), and thus are characterized as significant sources of sustained competitive advantage

1347

(Barney, 1986). In addition, whereas no one strategy is capable of leading to superior performance in the long run (Jacobson, 1992), cultural value systems, such as CO and LO, enable firms to foresee changes, make adjustments, and achieve superior long-term performance (Slater and Narver, 1995). Accordingly, in this study, we explore the viability of three important theses: (1) customer- and learning-oriented organizational value systems constitute critical competitive resources, acting jointly as drivers of superior organizational performance, and each having positive effects independently from one another on differing aspects of organizational performance; (2) CO and LO are more likely to grow if complemented by appropriate societal–culture factors; and (3) for a more effective firm, CO and LO should be supported by a strong organizational culture. Our analyses provide general support for the theses. In addition, several insights are revealed as to how CO and LO develop and affect firm performances, and three such implications deserve further discussion. First, the bivariate correlation coefficients in Table 1 suggest that CO and LO relate positively to both financial and qualitative components of firm performance, which is consistent with the extant literature. However, when the joint effects of these factors are investigated, CO appears to be related directly to financial and market performance only; and, while exerting a marginally significant positive effect on financial performance, LO relates more strongly to qualitative performance (i.e., employee satisfaction and commitment, quality improvements, and innovativeness). These findings suggest that CO and LO complement one another as they jointly improve firm performances. Thus, obviously, superior long-term performance is by far more likely when a firm puts emphasis on both value systems. Indeed, pursuing only CO may (mis)lead firms to focus excessively on the expressed needs of existing customers (Hamel and Prahalad, 1991) and/or to overemphasize the importance of past successes in decisions (Celuch et al., 2002). Such firms may fail to notice latent customer needs, emerging markets, and opportunities for change. By contrast, a pure LO could be problematic because of its so-called binside-outQ orientation (Day, 1994). Occupied so deeply with creating knowledge and finding new methods of doing things, purely learningoriented firms may have a self-centered view of the external world, giving insufficient importance to the voice of customers and other marketplace actors, thus failing to understand the long-term trends in the competitive arena. Second, we show that the firm-level reflections of national cultural factors do influence customer- and learning-oriented value systems in organizations. For instance, the interpersonal connectedness growing out of a collectivist culture is shown to exert positive effects on both CO and LO. This means that building CO and LO in individualistic firms entails greater challenges than in collectivist firms. Managers in individualistic firms need to put greater emphasis on the design of rewards and controls that take into account the instrumentalist motives of the individu-

1348

C. Yilmaz et al. / Journal of Business Research 58 (2005) 1340–1352

alistic personnel. In addition, the finding that power distance is unrelated to CO and LO should not be taken as conclusive. The observed negative correlation coefficients linking power distance to CO and LO (see Table 1) are in fact in line with our expectations that high power distance would inhibit the voicing and sharing of new insights and information and therefore negatively influence CO and LO. These negative effects are attenuated in the path analytic model, however, probably due to the bovershadowingQ effect of the cultural strength factor. Note that a strong organizational culture seems to be the prime driver of CO and LO in our sample. Note also that such a strong positive role for cultural strength is possible only when the shared cultural elements are largely consistent with aspects of CO and LO. It appears that, when such a strongly shared support exists in the organization’s culture, the otherwise deteriorating effects of power distance on CO and LO do not surface and operate as hypothesized. In such circumstances, cultural strength seems to act as a full mediator of the effects of power distance on CO and LO. Because high power distance firms generally lack informal communication patterns, these firms also fail to develop value congruence or a strong organizational culture, which in turn hinders CO and LO. In support of this view, when cultural strength is excluded from the model, power distance exerts negative effects on both value systems. Third, we show that the effects of (1) CO on financial and market performance and (2) LO on qualitative performance are moderated by cultural strength. The moderating role of cultural strength within the COYfinancial performance link is consistent with our expectations; that is, CO exerts a greater positive effect on financial firm performance in strong organizational cultures than in weaker cultures. Note that employees in strong cultures (1) bcan blindly coordinate activities and economize on communication costsQ, (2) bguess what their managers want them to do, rather than being explicitly told soQ (Fisher, 1997, p. 48), and (3) know the organization’s goals and work for them (e.g., Gordon, 1985). A strong culture, thus, can improve effectiveness because it increases horizontal and vertical efficiencies and facilitates goal alignment (Deal and Kennedy, 1982). Such aspects of strong cultures aid the development of systems and processes that enable rapid and decisive organizational responses to external contingencies. Acting on customer information is easier in such decisive contexts. And, unless such a unity exists, getting different organizational units to act consistently and coherently towards providing customer value is much harder. On the other hand, the observed moderating role of cultural strength within the LOYqualitative firm performance relationship is in the opposite direction of that expected. We find that LO exerts a positive effect on qualitative firm performance under the low cultural strength condition, but this effect size reduces gradually to the point of nonsignificance as the organizational culture becomes stronger. This finding relates to the

negative aspects of strong cultures. Regardless of their content, strongly shared norms and values may (1) hinder open expression of different views, (2) discourage the search for new methods and applications, (3) limit variety in opinions and approaches, and (4) decrease an organization’s long-term ability to adapt its systems, processes, beliefs, and behaviors to changing conditions (e.g., Kotter and Heskett, 1992). The lack of flexible thinking and open sharing of ideas in strong culture firms seems to be inhibiting the ability of the firms in our sampling context to transform learning-oriented values into appropriate strategies and actions. This, in turn, weakens the observed relationship between LO and qualitative performance. Strong culture firms aiming to create an effective LO should therefore give specific importance to flexibility and openness. One final issue concerns the finding that the aforementioned negative aspects of strong cultures do not appear to have a significant bearing in the mechanisms through which CO influences performance components. One plausible explanation for this finding is that the primary mechanisms through with CO influences firm performances do not have much to do with openness and/ or adaptability. That is, CO seems to enhance performance components primarily through rapid and decisive organizational responsiveness to external and internal contingencies (a capability that is fostered by cultural strength), rather than through open expression of views and/or adaptability (a capability that is hindered by cultural strength). 4.1. Limitations and future research directions Our study has a number of conceptual and methodological limitations, which nevertheless provide fruitful avenues for future research. Foremost among these is the organizational-level use of the societal–culture factors. As noted before, our sampling context provides a rare opportunity to explore the effects of societal–cultural factors on organizational phenomena in that the organizations in our sample are typical reflections of the various geographically and ethnically diverse subcultures in Turkey. Still, however, the observed variability in cultural assumptions across the organizations (or subcultures) of a single nation may not profile cross-national differences adequately. Therefore, future research should address the relationships explored in this study with samples obtained from multiple nations. Such research would also extend the generalizability of our findings. Another potential limitation concerns the multidimensional nature of firm performance. Dealing with the various components of firm performance always presents a challenging set of problems for researchers. As Denison and Mishra (1995, p. 205) note, investigation of firm performance requires that it be bdefined by a complex set of stakeholders, who may hold differing, incompatible, and

C. Yilmaz et al. / Journal of Business Research 58 (2005) 1340–1352

changing criteriaQ. This study distinguishes between components of firm performance that relate to the internal organizational processes and external organizational outcomes. Empirical analyses of the performance items suggest a two-factor solution that is in line with this conceptualization. We show that CO is a useful predictor of financial and market performance, and that LO is a more potent predictor of qualitative performance. Future research might explore the links between specific cultural orientations and related performance components in further detail. Finally, note that our study relies on measures of cultural traits obtained from a broad range of organizational members. This constitutes a major advantage over prior studies that used single informants only and relied mostly on top manager perspectives—thus failing to control for informant bias and to reflect different respondent perspec-

1349

tives. Nevertheless, the reliance on multiple informants bears the problem of aggregating the responses to the organizational level. To our knowledge, no processes exist to do this without error.

5. Conclusion In conclusion, this study reveals that customer- and learning-oriented organizational value systems are easier to develop when they are complemented by collectivist cultural assumptions and when they are supported by a strong culture. In addition, while both CO and LO facilitate albeit different aspects of firm performance, CO is shown to be more effective in strong cultures, and LO is found to play a more prominent role in contexts where the underlying values are not shared so strongly amongst

Appendix A. Measurement scales and analyses

Source

Dorfman and Howell (1988)

Erez and Earley (1988) Chen et al. (1998)

New item New item

Sigler and Pearson (2000)

Robbins and Mukerji (1994)

Scale

In this firm, our shared views and values are such that. . . Collectivism: (1) Group welfare is more important than individual rewards. (2) Group success is more important than individual success. (3) Individuals may be expected to give up their goals in order to benefit coworker success. (4) Working with a group is better than working alone. (5) People should pay absolutely no attention to coworker views when deciding what kind of works to do. (R) (6) Everyone should be held jointly responsible for each other’s performance. (7) Group decisions are more important than individual decisions. Power distance: (1) People at lower levels in organizations have a responsibility to make important decisions for people around them. (2) People at lower levels in the organization should not have power in the organization. (3) It is often necessary for a supervisor to emphasize authority and power when dealing with subordinates. (4) Managers should be careful not to ask the opinions of subordinates too frequently. (5) A manager should avoid socializing with his/her subordinates at the job. (6) Subordinates should not disagree with their manager’s decisions. (7) Managers should not delegate difficult and important tasks to subordinates.

Standardized loading

Alpha

Variance extracted

Composite reliability

.80

.50

.87

.87

.50

.88

.63 .79 .74

.68 .69

.67 .73

.59

.65 .70

.81 .70 .64 .72 (continued on next page)

1350

C. Yilmaz et al. / Journal of Business Research 58 (2005) 1340–1352

Appendix A (continued) Source

Denison (2000)

Denison (2000)

Denison (2000)

Denison (2000)

Scale

Standardized loading

Alpha

Variance extracted

Composite reliability

(8) Managers should make most decisions without consulting with subordinates. In this firm. . . Cultural strength: (1) There is a clear and consistent set of values that governs the way we do business. (2) There is an ethical code that guides our behavior and tells us right from wrong. (3) There is a strong culture. (4) It is easy to reach consensus, even on difficult issues. (5) There is a clear agreement about the right way and the wrong way to do things. (6) People from different parts of the organization share a common perspective. (7) There is a good alignment of goals across levels. Customer orientation: (1) Customer comments and recommendations often lead to changes. (2) Customer input directly influences our decisions. (3) All members have a deep understanding of customer wants and needs. (4) The interests of the customer often get ignored in our decisions. (R) Learning orientation: (1) Information is widely shared so that everyone can get the information he/she needs when it’s needed. (2) New and improved ways to do things are continually adopted. (3) We view failure as an opportunity for learning and improvement. (4) Innovation and risk taking are encouraged and rewarded. (5) Lots of things fall between the cracks. (R) (6) Learning is an important objective in our day-to-day work. (7) We make certain that the right hand knows what the left hand is doing. (8) The way things are done are very flexible and easy to change. Your firm’s performances in comparison to competition. . . Financial and market performance: (1) Sales growth. (2) Market share. (3) Return on sales. (4) Return on assets. (5) Overall profitability. Qualitative performance: (1) Quality improvements. (2) New product development capability. (3) Employee satisfaction. (4) Employee commitment.

.82

.83

.50

.87

.70

.50

.78

.82

.50

.88

.95

NA

NA

NA

.95

NA

NA

NA

.71

.70 .65 .74 .69

.74

.72

.54

.67 .84

.72

.77

.66 .65 .75 .77 .70 .64 .67

C. Yilmaz et al. / Journal of Business Research 58 (2005) 1340–1352

1351

Appendix A (continued) Source

Scale Rate of changes in. . . Market dynamism: (1) Customer preferences. (2) Competitors’ strategies. (3) Product characteristics. (4) Technology.

Standardized loading

Alpha

Variance extracted

Composite reliability

.95

NA

NA

NA

(R): Reverse coded item; NA: not available because the construct was measured with a formative scale.

organizational members. Whereas our study reflects evidence from a single-nation sample, the findings provide valuable guidance for theory development and managerial decision making.

References Argyris C, Schon DA. Organizational learning: a theory of action perspective. Reading (MA)7 Addison-Wesley; 1978. Baker WE, Sinkula JM. The synergetic effect of market orientation and learning orientation on organizational performance. J Acad Market Sci 1999;27:411 – 27. Barkema GB, Vermeulen F. What differences in the cultural backgrounds of partners are detrimental for international joint ventures? J Int Bus Stud 1997;28:684 – 845. Barney JB. Organizational culture: can it be a source of sustained advantage? Acad Manage Rev 1986;11:656 – 66. Callantone RJ, Cavusgil ST, Zhao Y. Learning orientation, firm innovation capability, and firm performance. Ind Mark Manage 2002;31:515 – 24. Celuch KG, Kasouf CJ, Peruvemba V. The effects of perceived market and learning orientation on assessed organizational capabilities. Ind Mark Manage 2002;31:545 – 54. Chatman JA, Barsade SG. Personality, organizational culture, and cooperation: evidence from a business simulation. Adm Sci Q 1995;40:423 – 43. Chen CC, Meindl JR, Hui H. Deciding on equity or parity: a test of situational, cultural, and individual factors. J Organ Behav 1998;19: 115 – 129. Conduit J, Mavondo F. How critical is internal customer orientation to market orientation. J Bus Res 2001;51:11 – 24. Day GS. The capabilities of market-driven organizations. J Market 1994;58(October):37 – 52. Deal TE, Kennedy AA. Corporate culture: the rites and rituals of corporate life. Reading (MA)7 Addison-Wesley Publishing; 1982. Denison DR. Organizational culture: can it be a key lever for driving organizational change. The handbook of organizational culture. London7 Wiley; 2000 1–26. Denison DR, Mishra A. Toward a theory of organizational culture and effectiveness. Organ Sci 1995;6:204 – 23. Denison D, Neale W. Denison organizational culture survey. Ann Arbor (MI)7 Aviat; 1994. Deshpande R, Farley JU, Webster FE. Organizational culture, customer orientation and innovativeness in Japanese firms: a quadrant analysis. J Market 1993;57(January):23 – 7. Doney PM, Cannon JP, Mullen MR. Understanding the influence of national culture on the development of trust. Acad Manage Rev 1998;23:601 – 20. Dorfman P, Howell J. Dimensions of national culture and effective leadership patterns: Hofstede revisited. JAI Press; Greenwich, CT 1988;127–150.

Erez M, Earley PC. Comparative analysis of goal setting strategies across cultures. J Appl Psychol 1987;72:658 – 65. Fisher CJ. Corporate culture and perceived business performance: a study of the relationship between the culture of an organization and perceptions of its financial and qualitative performance. Doctoral thesis, California School of Professional Psychology. LA; 1997. Fornell C, Larcker D. Evaluating structural equation models with unobservable variables and measurement error. J Market Res 1981; 18:39 – 50. Gordon G. The relationship of corporate culture to industry, sector, and corporate performance. In: Kilman RH, Saxton MJ, Serpa R, et al., Gaining control of the corporate culture. San Francisco7 Jossey-Bass; 1985; p. 103 – 25. Hamel G, Prahalad CK. Corporate imagination and expeditionary marketing. Harvard Bus Rev 1991;(July–August):81 – 92. Harris L. Cultural domination: the key to a market-oriented culture? Eur J Market 1998;32(March):354 – 73. Hofstede G. Culture’s consequences: international differences in workrelated values. London7 Sage Publications; 1980. Hofstede G, Neuijen B, Ohayv DD, Sanders G. Measuring organizational cultures: a qualitative and quantitative study across twenty cases. Adm Sci Q 1990;35:286 – 316. Hult GTM, Ketchen DJ, Reus TH. Organizational learning capacity and internal customer orientation within strategic sourcing units. J Qual Manage 2001;6:173 – 92. Hurley RF, Hult TM. Innovation, market orientation, and organizational learning: an integration and empirical examination. J Market 1998;62(July):42 – 54. Jacobson R. The Austrian school of strategy. Acad Manage Rev 1992;17:782 – 807. Kidwell RE, Mossholder KIW, Bennett N. Cohesiveness and organizational citizenship behavior: a multilevel analysis using work groups and individuals. J Manage 1997;23:775 – 93. Kotter JP, Heskett JL. Corporate culture and performance. New York7 Free Press; 1992. Liu S, Xueming L, Yi-Zheng S. Integrating customer orientation, corporate entrepreneurship, and learning orientation in organizaitons-in-transition. Int J Res Market 2002;19:367 – 82. Newman KL, Nollen SD. Culture and congruence: the fit between management practices and national culture. J Int Bus Stud 1996;27: 753 – 779. Nunnaly JC. Psychometric theory. 2nd ed. New York7 McGraw-Hill; 1978. Pascale RT. The paradox of corporate culture: reconciling ourselves to socialization. Calif Manage Rev 1985;27:26 – 41. Posner B, Kouzes J, Schmidt W. Shared values make a difference: an empirical test of corporate culture. Hum Resour Manag 1985;24(3); 293 – 309. Robbins S, Mukerji D. Managing organizations. New Jersey7 Prentice-Hall; 1994. Saffold G. Culture traits, strength and organizational performance: moving beyond dstrongT culture. Acad Manage Rev 1988;13(4);546 – 58. Sathe V. Implications of corporate culture: a manager’s guide to action. Organ Dyn 1983;12:4 – 23.

1352

C. Yilmaz et al. / Journal of Business Research 58 (2005) 1340–1352

Schall MS. A communication-ruled approach to organizational culture. Adm Sci Q 1983;28:557 – 81. Schein EH. Organizational culture and leadership. San Francisco7 JosseyBass; 1985. Schein EH. Organizational culture. Am Psychol 1990;45(2):109 – 19. Senge PM. The fifth discipline: the art and practice of learning organization. New York7 Doubleday; 1990. Sigler TH, Pearson CM. Creating an empowering culture. J Qual Manage 2000;5(Spring):27 – 57. Sinkula JM. Market information processing and organizational learning. J Market 1994;58:35 – 45. Sinkula JM, Baker WE, Noordewier TA. Framework for marketbased organizational learning: linking values, knowledge, and behavior. J Acad Market Sci 1997;24:305 – 18.

Slater S, Narver JC. Market orientation and the learning organization. J Market 1995;59(July):63 – 74. Slater S, Narver JC. The positive effects of a market orientation on business profitability. J Bus Res 2000;48:69 – 73. Van Oudenhoven JP. Do organizations reflect national cultures? A 10nation study. Int J Intercult Relat 2001;25:89 – 107. Wagner J, Moch M. Individualism–collectivism: concept and measure. Group Organ Stud 1986;11:280 – 304. Wasti SA. Affective and continuance commitment to the organization: test of an integrated model in the Turkish context. Int J Intercult Relat 2002;26:525 – 50. Yilmaz C, Hunt SD. Salesforce cooperation: the influence of relational, tax, organizational, and personal factors. J Acad Market Sci 2001;29(4);335 – 57.