Journal of Management Studies 43:7 November 2006 doi: 10.1111/j.1467-6486.2006.00620.x

To Conform or To Perform? Mimetic Behaviour, Legitimacy-Based Groups and Performance Consequences*

Ilídio Barreto and Charles Baden-Fuller Universidade Católica Portuguesa, Lisbon; Cass Business School, London abstract The study of interorganizational imitation has been an important strand in the recent literature on institutional theory. This paper offers new insights for our understanding of mimetic isomorphism and its reliance on legitimacy: we suggest that legitimacy-based reference groups guide firms in their mimetic behaviour, that firms undertake imitation even against their own ex ante information, and that legitimacy-based imitation contributes negatively to firms’ profitability. We examine Portuguese bank branching decisions between 1988 and 1996 and find that banks imitate their legitimacy-based groups, and not only towards ex ante (firm-specific) attractive locations, but also towards unattractive locations; we also find that mimetic branching produces a negative effect on profitability. We conclude that these results show the importance of legitimacy pressures on organization decisions and the tension between the pressure to conform and the pressure to perform.

INTRODUCTION The recent literature on new institutional theory has identified a wealth of material that enriches our understanding of organizational phenomena, and particular interest has surrounded the prevalence of interorganizational imitation (Dacin et al., 2002). Recent studies have addressed mimetic behaviour on issues such as diversification decisions (Fligstein, 1991), corporate acquisition choices (Haunschild, 1993), entry in new markets (Haveman, 1993), adoption of matrix structures (Burns and Wholey, 1993), choice of investment banker (Haunschild and Miner, 1997), market positions decisions (Greve, 1998), adoption of college curriculum changes (Kraatz, 1998), plant location decisions (Henisz and Delios, 2001), choice of firms’ coverage by securities analysts (Rao et al., 2001), adoption of new organizational form (Lee and Pennings, 2002), and TMT members hiring (Williamson and Cable, 2003). Address for reprints: Charles Baden-Fuller, Cass Business School, 106 Bunhill Row, London EC1Y 8TZ, UK ([email protected]). © Blackwell Publishing Ltd 2006. Published by Blackwell Publishing, 9600 Garsington Road, Oxford, OX4 2DQ, UK and 350 Main Street, Malden, MA 02148, USA.

1560

I. Barreto and C. Baden-Fuller

Most of these studies have explored the explanation offered by new institutional theory for interorganizational imitation: mimetic isomorphism. Mimetic isomorphism is a process by which, in ambiguous and uncertain situations, organizational changes are imitated in order to gain legitimacy (DiMaggio and Powell, 1983). Despite the important stream of extant literature, some significant gaps still persist, and can be encapsulated by the following three questions: Who imitates whom (i.e. which generic framework explains heterogeneity in imitation)?; Is there selective imitation (i.e. do firms choose to imitate only those choices judged to be ‘good’, while avoiding ‘bad’ options)?; What are the financial performance effects of mimetic isomorphism; In which situations do these effects occur (i.e. does mimetic isomorphism lead to firms’ lower profitability, and if it does, are such effects restricted to mimetic-driven ‘good’ options-situations, or do they also apply to ex ante ‘bad’ options-situations)? This paper develops a set of hypotheses intended to help to fill these gaps. We suggest there is a generic legitimacy-driven framework that organizations use when imitating, and that it is based on legitimacy providers. These providers simplify and categorize the complex environment, providing reference points for firms to follow (Fiegenbaum and Thomas, 1995; Fiegenbaum et al., 1996; Porac and Thomas, 1990), and we suggest that managers use these to guide their mimetic behaviour. We show that firms imitate such ‘legitimized’ behaviour when their ex ante information indicates choices are ‘bad’ as well as ‘good’, and why both actions may lead to under-performance. This focus on legitimacy-based groups, on imitation of ‘bad’ as well as ‘good’ decisions, and on the effect of mimetic behaviour on profitability, is novel. Furthermore, the findings clearly show the trade-off managers face between the pressure to conform (to obtain legitimacy gains, as explained by ‘normative rationality’ literature) and the pressure to perform (as explained by ‘economic rationality’ literature). We use fine-grained time series and cross-section data on the branching behaviour of the Portuguese banking industry to test our hypotheses about mimetic behaviour and its performance effects. Our study covers the period of 1988–96, and the behaviour of 26 banking firms over 305 concelhos (counties) comprising all of Portugal’s administrative districts. The paper is organized as follows. We first extend institutional theory by offering new insights for an integrated approach to mimetic behaviour that is wholly centred on legitimacy. We then show why firms undertake mimetic actions even against their own (ex ante) previously held judgments; and we explain why legitimacy-driven mimetic actions contribute negatively to firms profitability. We describe our context, data, methods and tests before giving results and providing discussion and conclusions.

THEORY AND HYPOTHESES Who Imitates Whom? The new institutional theory provides an important explanation for inter-organizational imitation. In response to the question raised by Hannan and Freeman in their seminal paper on organizational ecology (1977), ‘Why are there so many kinds of organizations’, DiMaggio and Powell (1983, p. 148) asked: ‘Why is there such startling homogeneity of organizational practices?’ They proposed institutional isomorphism (a set of forces that © Blackwell Publishing Ltd 2006

To Conform or To Perform?

1561

press organizations to resemble other organizations to conform to the environment) as one possible answer. They further suggested three main processes (coercive isomorphism, normative isomorphism and mimetic isomorphism) of which mimetic isomorphism is central to our paper, and occurs when organizations respond to change by imitating others’ behaviour in a standardized way. Mimetic isomorphism (or ‘mimicry’) provides legitimacy and is more likely to happen when managers face ambiguous situations with unclear solutions (Cyert and March, 1963), since in those situations legitimacy becomes crucial for organizational survival (Dacin, 1997; Deephouse, 1996; Meyer and Rowan, 1977; Scott, 1987; Singh et al., 1986; Westphal et al., 1997). Past writers have adopted a variety of approaches to identify the process of selection which identifies who imitates whom. Some suggest that firms use network peers as ‘models’ (Haunschild, 1993; Kraatz, 1998; Westphal et al., 1997); some suggest size as a criteria (Fligstein, 1991; Haveman, 1993; Haunschild and Miner, 1997); some argue for the most successful organizations (Burns and Wholey, 1993; Haunschild and Miner, 1997; Haveman, 1993). We suggest adopting a more generic and more legitimacy-based framework. Hambrick and Mason (1984) and Dutton and Jackson (1987) stated that we should look to cognitive processes if we wish to model managerial decisions. Porac et al. (1989) provided the framework. When environments are complex and dynamic, decision-makers have to deal with large amounts of ambiguous information, and managers use cognitive simplification processes to overcome uncertainty and conquer the information processing nightmare (Huff, 1982, 1990; Reger and Huff, 1993; Schwenk, 1984). They create cognitive categories or groups of firms with relevant similarities, and identify their actions with others belonging to the same group category or group. This categorization influences ‘material decisions’ (what to produce, what to purchase, which customers to target) (Porac and Thomas, 1990). Fiegenbaum and Thomas (1995) used data from the insurance industry to show how firms within a strategic group context converged on a reference point. They argued for the choice of groups based on ‘characteristics’ or ‘reference points’ (Fiegenbaum et al., 1996). In this paper we argue for another process, centred on legitimacy-based groups. These groups are not selected on the basis of the firm’s own choice, but on what others in a position of authority say. Under legitimacy-based groupings, although firm A may believe that it shares many characteristics (such as ambition) with B, it may be classed as similar to C by legitimacy providers for other reasons (such as location). Legitimacy is a generalized perception or assumption that the actions of an entity are desirable, proper, or appropriate within some socially constructed system of norms, values, beliefs, and definitions’ (Suchman, 1995, p. 574). In this definition we can see the presence of notions like ‘generalized perception’ or ‘socially constructed system’ which rightly emphasize the dependence of legitimacy on collective observers, which we termed by legitimacy providers. Legitimacy providers are observers that have status to assess the conformity of firm behaviour to specific socially constructed standards and to do this need to categorize the firms. Faced with complex and potentially ambiguous information and to assure themselves, they tend to share their intra-industry categorization among themselves (and also with managers) in a two-way process that parallels the one specified by Reger and Huff (1993) for managerial categorization. That is, they engage in a © Blackwell Publishing Ltd 2006

1562

I. Barreto and C. Baden-Fuller

process that is based on frequent interaction due to meeting events and personal contacts, and on commonalities in information sources. These categories constructed by legitimacy-providers constitute the legitimacy-based groups, which act as a reference group since they serve to compare and evaluate firms against their group standards (Fiegenbaum and Thomas, 1995): external constituencies metric the organization’s behaviour against these benchmarks to assess conformity and provide legitimacy (Meyer and Rowan, 1977); in their actions, managers use categorization influenced by external constituencies as long as it serves as a gauge against which their actions are evaluated and justified (Dutton and Dukerich, 1991). In summary, we argue that when mimetic behavioural pressures are strong, firms will follow the actions of those signalled as legitimate by outside legitimacy providers. Hybels (1995) and Deephouse (1996) identified several groups of organizational constituencies that fulfil this role, including regulators and media. These processes can be applied to our context of branching decisions in banking (that is, the decisions as to whether and where to open up new branches). Both this industry and this type of decision seem particularly suitable to this research. First, the banking industry has recently undergone several periods of high uncertainty, due to severe and successive environmental changes. Second, there is some anecdotal evidence that mimetic behaviour is present in banking: ‘The banking industry is overwhelmed by imitation. . . . One bank goes into Internet banking; all banks go into Internet banking. One bank put branches in supermarkets; all banks put branches in supermarkets’ (Porter, 1998, p. 40). Third, we are studying a legitimacy-driven imitation process, and banking is an industry where there is already empirical evidence about legitimacy providers (Deephouse, 1996). Fourth, branching decisions represent visible, highly consequential strategic choices for banks (Miller, 1997, p. 72), enabling us to confirm imitation as a central issue in strategy. Fifth, branching decisions have some specific characteristics that (in contrast to other types of decisions) allow the analysis of new aspects of the imitation process, as we will show in this section. We apply our argument to this empirical setting. For instance, domestic private generalist banks will tend to imitate the branching decisions of other similar banks where legitimacy providers consider some domestic private generalist banks to be a specific category. Accordingly, we formulate the following hypothesis: Hypothesis 1: Banks imitate their legitimacy-based reference group in branching decisions. Is There a Selective Imitation? Are firms selective in what they imitate? That is, given that they want to engage in mimetic behaviour for a certain activity, do they choose to imitate only ‘good’ behavioural options, while avoiding ‘bad’ behavioural options? Furthermore, would they make decisions against their own judgment, and if so, will the consequences be adverse? We begin addressing these questions with a brief digression into the theoretical models of herd behaviour. In economics literature, imitation is modelled through such © Blackwell Publishing Ltd 2006

To Conform or To Perform?

1563

formal models (Avery and Zemsky, 1998; Banerjee, 1992; Bikhchandami et al., 1992; Scharfstein and Stein, 1990), and these provide three important conclusions for our research. First, uncertainty plays an important role in herding. For instance, Banerjee’s (1992) sequential model includes two modes of uncertainty, one about the appropriateness or quality of the information received by the focal firm, and another about the quality of information underlying the (observed) choices of other competitors. Second, herd behaviour can result in erroneous or economically inefficient choices with adverse consequences (Banerjee, 1992; Bikhchandami et al., 1992; Scharfstein and Stein, 1990) and offer an explanation for mimetic adoption of technically inefficient innovations (Abrahamson and Rosenkopf, 1993). Third, firms may engage in herding movements even when their own ex ante private information suggests that the actions of others may be wrong choices (Banerjee, 1992; Devenow and Welch, 1996; Scharfstein and Stein, 1990). Although our arguments come from a different route, we also suggest that mimetic isomorphism leads to outcomes that are similar to the herding models from economics: the adoption of ‘bad’ decisions and the reliance on others’ actions even to the point of ignoring the actor’s own ex ante information. Institutional theory has argued that legitimacy-driven forces may lead firms to inappropriate resource decisions, implying the existence of a ‘normative rationality’, based on social justification, in contrast to ‘economic rationality’, based on profitability (Oliver, 1997). The potential trade-off between legitimacy and performance is due to the fact that, on one hand, legitimacy-based imitation enables organizations to increase their probability of survival (DiMaggio and Powell, 1983; Meyer and Rowan, 1977), while on the other, the likelihood of survival can be obtained at expense of performance (Henderson, 1999). This means that certain legitimacy-seeking activities, although they may make negative contributions to organizational performance, will at the same time assure the future support of certain internal or external constituencies if the organization falls into particularly adverse circumstances. Thus, the main difference between the institutional approach we suggest here and the economic approach concerns the underlying mechanism which triggers mimicry. While in herd behaviour models the source of such behaviour is an economically rational purpose (informational externalities, payoff externalities, or managerial reputation), in the mimetic isomorphism framework it is the normatively rational motive (legitimacy) that can both prompt and justify mimicry. We suggest that, where there is significant uncertainty, when legitimacy pressures are strong, firms engage in mimetic behaviour even when this means taking what would previously have been considered ‘bad’ decisions. In the context of the banking industry and the decisions about opening new branches, some locations will be more attractive than others. Thus, in our setting, good decisions are imitating other banks in setting up in (ex ante) attractive locations and ‘bad’ decisions would mean following others into ‘unattractive’ locations. We therefore formulate the following hypothesis: Hypothesis 2: Banks imitate their legitimacy-based reference group in branching decisions in both (ex ante considered) attractive and unattractive locations. © Blackwell Publishing Ltd 2006

1564

I. Barreto and C. Baden-Fuller

What Are the Effects of Imitation on Profitability? The literature just reviewed suggests that mimetic isomorphism may lead to negative performance consequences. While little attention has been paid to this subject, Westphal et al.’s (1997) study of imitation in the adoption of an administrative innovation found that hospitals that tried to conform to isomorphic pressures suffered in terms of organizational efficiency. Since ‘most strategic choices will tend to have greater economic impact on an organization than administrative innovations’ (O’Neill et al., 1998, p. 98), given the arguments linked to new institutional theory above discussed, it seems likely that we will find negative profitability effects from legitimacy-based imitation. In banking, branching is a strategic choice that can have significant consequences, and accordingly we formulate the following hypothesis: Hypothesis 3: Branching imitation leads to lower profitability. Since we have hypothesized that banks imitate others in moving to both ex ante attractive and unattractive locations, another question emerges: Is there a negative profitability effect associated with both types of imitation, or is it peculiar to just one of these types? Since we are studying the spatial moves inherent in branching decisions, there is an opportunity to discuss some complementary effects: namely, we should find some rivalry negative effects since branching imitation can drive banks to compete for the same customers in the same market places.[1] Ilmakunnas’ (2002) theoretical model on strategic behaviour in service industries provides a stylized explanation that is highly relevant to our branching context. The increase in rivalry will tend to produce a negative impact on firm profitability. Cool et al. (1999) show that both actual and potential increases in rivalry generate adverse effects on profitability, and Whalen (1992) finds that rivalry has a negative effect on profitability in local banking markets. In our context, rivalry effects should be important in imitation moves towards both attractive as well as unattractive locations. Furthermore, a relatively higher propensity of actual moves should be expected towards locations previously judged attractive than to those reckoned unattractive. This potential difference could even imply a higher rivalry effect associated with attractive locations. In sum, legitimacy-based imitation can induce a poor strategic choice (for instance, to open a new branch). This ‘wrong’ choice can be even worse if the expansion is into an unattractive location. Additionally, negative rivalry effect should be added to both types of imitation (although moves towards attractive locations can be expected to suffer the effect most). Accordingly, we formulate this hypothesis: Hypothesis 4: Branching imitation leads to lower profitability, in the cases of both ex ante attractive and unattractive locations. HISTORICAL AND RESEARCH CONTEXT We analyse branching decisions in two complementary levels: the ‘basic growth choice’, that is, the decision to expand (or not) the bank’s branch network; and the ‘location © Blackwell Publishing Ltd 2006

To Conform or To Perform?

1565

growth choice’, the decision as to where to locate the new branches. We use data on branching activity for the Portuguese banking industry in the period between 1988 and 1996, for ‘basic growth choice’, and for the years 1990 and 1991, for ‘location growth choice’, which requires more fine-grained data. To give a brief historical background, we follow Silverman et al. (1997) in taking 1975, the year when democracy was restored to Portugal after the end of the dictatorship, as marking the reference point for the start of the modern era for the Portuguese banking industry. All domestic banks were then nationalized, new merged institutions were created and, until the early 1980s, the banking industry was highly regulated and had few degrees of freedom: for instance, interest rates and total lending amounts to be offered by individual banks were bureaucratically controlled. In 1984 there was the beginning of a slow liberalization: private banks were once again allowed to exist, new entries permitted and new banks were formed such as Banco Comercial Portugues. In 1988 the situation changed again in a radical manner, as European Union directives heralded a series of important EU-wide regulatory changes in the banking industry. The single license principle was established, enabling any bank authorized in any EU member state to automatically carry on a wide range of banking activities in any other EU country. The directives led to full liberalization of capital flows, as well as defining how all EU banks were to calculate the solvency ratio, eliminating national differences and placing all banks on an equal footing. The upshot of all this was to permit more foreign (other EU) banks to enter the Portuguese market, and to allow local banks to alter their branch-networks. Simultaneously, the government started to privatize some of its state-owned banks, permitting even greater freedom of action for existing players. The cumulative result of these changes was to cause general uncertainty among players in 1988 and the immediate subsequent years (Gual and Neven, 1992). The problem faced by bankers in such dynamic and complex context was to determine which choices were the best for their banks. Obviously, banks altered products, range of services, customer segmentation, promotion, and pricing practices. More surprisingly, Portuguese banks doubled the size of their branch networks over the eight years between 1988 and 1996, with the greatest expansion taken place in 1990 and 1991. In these two years, more than 740 branches were added, representing more than 40 per cent of the total initial stock as at 1988. From a macro perspective it is not at all obvious why Portuguese banks expanded their branch networks in this explosive manner, as neither local nor the EU regulators encouraged such behaviour. It is notable that every other major European country experienced either near zero or negative rates of growth in (retail) branches despite similarity of circumstances between Portugal and other European countries with respect to the density of branches, banking legacy and the general patterns of evolution. This difference in growth between Portugal and all its neighbours cannot be explained by (economic) ‘rational factors’, for there was no evidence that Portugal suffered from a lack of branches. Nor did they need to grow the physical branch network to expand their retail customer base, since every consumer enjoyed easy access to telephone banking, computer-based banking, EFTPOS (electronic funds transfer at point of sale), and Portugal’s standardized ATM network was well advanced compared to other EU countries, allowing any customer to use any bank’s ATM. Given the change in bank custom© Blackwell Publishing Ltd 2006

1566

I. Barreto and C. Baden-Fuller

ers’ habits, the use of better segmentation procedures and the significant cost differential between branches and other distribution channels, the expansion of Portuguese bank branch networks, which was questioned by local bank experts at the time, deserves examination. DATA We built a database of the number of branches held by each of the 26 banks during the period from 1988 to 1996, based on data from the Associação Portuguesa de Bancos (APB), the Portuguese banking association. We collected detailed data for each of the two years 1990 and 1991 on the location of branches for these 26 banks in each of 305 counties (summarized in the Appendix) and on the size of the population and the average household income in the same counties. We also collected standard economic data on bank ownership, bank profits, bank assets, customer segments served, and organizational age. Our sources include annual bank reports, central bank reports, and publications of the Portuguese banking association. We also undertook interviews with the major actors to verify that our perceptions of the industry were correct. METHODS AND VARIABLES Mimetic Behaviour To test Hypothesis 1, the key variable we wished to explain is branch expansion by individual banks with reference to the branching behaviour of other banks. It has been already noted that most research on inter-organizational imitation has tended to adopt indirect measures of the existence of mimetic behaviour, such as the use of the percentage of adoptions of some organizational change or innovation, or the relationship between adoption and density of previous adoptions (Haunschild, 1993). A stronger test of imitation may satisfy the three conditions Haunschild (1993, p. 567) proposed: (a) occurrence of a certain behaviour by the ‘models’ for imitation, at time period t; (b) the ability of other actors to observe the behaviour by the key economic agent; (c) replication of the referred behaviour by other organizations at time period ‘t + x’. Branching behaviour is a strategic decision, for once undertaken there are ‘sunk costs’ and ‘exit costs’ because of specialized non-reusable furnishings, long term commitments under property leases and employment protection laws making layoffs expensive. Despite the high potential costs of exiting, entry was easy – according to industry sources it only took a few months to plan and open a branch, so mimetic behaviour could take place speedily. For our imitation study, we took a one-year time lag for possible imitation effects. Ours is therefore a very stringent test of imitation behaviour (x equals 1 in Haunschild terms), looking at whether banks copied the behaviour of their rivals in the same reference group in the previous year. We observe the branching behaviour of each bank, first in terms of the ‘basic growth choice’ (the first of the two levels of branching here studied) from each of the seven years from 1990 to 1996, and then in terms of the ‘location growth choice’, the more important level, in 1991. We choose 1991 for studying this second level of branching decisions not © Blackwell Publishing Ltd 2006

To Conform or To Perform?

1567

only due to the availability of the fine-grained data required for such kind of analysis, but also because this was the most important year (within the 1988–96 period) in terms of number of new branches opened. Since we intended to test the replication of group behaviour by reference to the behaviour of its members in the previous period, the independent variables were ‘basic growth choice’ and ‘location growth choice’ made by the focal firm’s reference group in the relevant periods (that is, 1989 to 1995, and 1990 respectively). All these variables were dichotomous, corresponding to the decision to expand (or not) the branch network (‘basic growth choice’) and to open (or not) in each of the 305 counties (‘location growth choice’). The branching behaviour of our 26 banks was skewed; whilst most banks opened many branches, some opened almost none. An analysis of the detailed data on branch expansion behaviour in 1991 showed that, while six banks opened no new branches, three banks opened more than 50 branches each. For the ‘location choice variable’ we classified each bank’s behaviour in each of the 305 counties as either zero (no opening) or one (an opening), since virtually none of the banks opened more than one branch in a particular region in the period in question. It is also important to note that the data record net changes – thus if a bank moved its location from a run-down high street to a newly opened shopping centre to benefit from local agglomeration effects, this change would not appear in our data set as the total number of branches would not be changed. Legitimacy-based reference groups. Our theory requires us to identify legitimacy-based reference groups based on key legitimacy providers. Deephouse (1996) offers evidence that, in banking, media and regulators are sources of legitimacy. In searching for their views, we focused on the APB reports for July 1988 and July 1990, those of the leading economic magazine Semanário Económico for July 1990 and July 1991, and the public statements of the central bank (e.g. Diário Económico, November 1991) and the EU Commission (confirmed in the Banco de Portugal Annual Report, 1993 and 1994). These external agencies were in complete agreement about the three relevant dimensions (attributes in categorization theory): the degree to which banks specialize (i.e. generalist or specialist banks), the ownership structure (state-owned or privately owned banks), and the origin (domestic or foreign banks). Following Porac and Thomas (1990), Figure 1 illustrates how our sources saw the picture, providing a schema that sorts banks into three levels and five relevant categories: state-owned generalist banks; domestic private generalist banks; domestic private specialist banks; foreign generalist banks; and foreign specialist banks. (As all state-owned banks engaged in universal banking activities, there were no banks in the sixth, stateowned specialist, category.) We note that between 1992 and 1996, regulators and media reclassified two banks from being specialized to the category of generalists. Location attractiveness. In order to test Hypothesis 2, it is necessary to separate attractive counties from unattractive counties. We have built a dummy variable, loc_attr, that assumes the value ‘1’ if the county is ‘attractive’, and ‘0’ if the county is ‘unattractive’. For each county we obtained data on total population and on income per capita (proxied by each county’s 1989 average collected personal tax). We adjusted county © Blackwell Publishing Ltd 2006

1568

I. Barreto and C. Baden-Fuller Banks

Specialist

Generalist

Private Specialist

Stateowned Specialists*

Domestic Foreign Private Private Specialists Specialists

Private Generalist

Stateowned Generalists

Domestic Private Generalists

Foreign Private Generalists

Figure 1. Organizational categories (applying categorization theory) Notes: *This category is empty in the Portuguese case. ‘Private’ in the Portuguese setting means ‘non-state-owned’.

income by the number of branches (of all banks) in that county to get a measure of ‘attractiveness’ (income per capita ¥ population for 1989/number of bank branches in 1990) and compare it to the average for the bank in question. This measure indicated that there were 335 instances of counties being attractive locations, whereas there were 7595 unattractive locations. This simple measure picks out obviously attractive branch locations, highlights dangers of negative spillovers (too many competitors for the amount of deposits) and is quite similar to the actual measures used by banks in location modelling. Profitability Effects Measurement. Our measure of profitability is Return on Assets (ROA), calculated as net income divided by averaged assets. ROA is one of the most commonly used measures of performance in the strategy and organizational literature, appearing in the strategic choice stream (e.g. Geletkanycz and Hambrick, 1997), in recent organizational ecology and evolutionary studies (e.g. Barnett et al., 1994; Sinha and Noble, 1997), and also in strategy studies on banking industry (e.g. Ramaswamy, 1997; Wiseman and Catanach, 1997). In addition, ROA is the indicator most closely watched by banks analysts and bankers themselves (Reger et al., 1992, p. 195), allowing comparisons between firms of different sizes. ROA measures are relatively insensitive to capital structure differences and changes, unlike measures such as Return on Equity (ROE). In our Hypothesis 3, we predicted a negative impact of branching imitation on banks’ profitability. We test this hypothesis for both ‘basic growth choice’ and ‘location growth choice’. When testing the ‘basic growth choice’ (imitation) impact on profitability our data set is pooled time series-cross section and we used the ROA for each bank-year. Our variable measures number of new branches for each bank accumulated between 1988 and each of our focal years (1990 and 1991). When testing for ‘location growth choice’, our data set is cross sectional only, and we use 1991 ROA for each bank as the dependent © Blackwell Publishing Ltd 2006

To Conform or To Perform?

1569

variable. The main independent variable is prop_91: the number of new branches opened in 1991 divided by the 1990 stock. However, since we also predicted the existence of imitation at the ‘basic growth choice’ level, we had to control it. In fact, the ‘basic’ level of imitation (opening branches as other members of the same group did in the previous period), can be decomposed in two components: one, the opening of branches in the same locations as same-group members had done in the previous period; two, the branch openings in other locations than those previously chosen by same-group members. Of course the first component is related to the ‘location growth choice’ imitation, and is captured by the variable prop_91. The second component is captured by the variable imibas91, a dummy variable which assumes the value of ‘1’ for banks that opened new branches in 1991 but in different locations than their reference group did in the previous year. Hypothesis 4 requires examining whether imitation produces a negative effect on profitability in both ex ante attractive and unattractive counties. Two additional independent variables are also used: pr_att91 is the proportion of new branches opened in 1991 by each bank in the same attractive counties as members of its reference group did in 1990, and pr_una91, the proportion of new branches opened in 1991 by each bank in the same unattractive counties as members of its reference group did in 1990. Control variables. To take into consideration organizational characteristics expected to influence bank profitability in this setting, we include in the performance analysis several control variables. Like others in the literature, we used size and age. Total assets adjusted by the appropriate annual GDP deflator measured organizational size. This measure has been extensively used in recent strategy studies based on banking (Sinha and Noble, 1997; Wiseman and Catanach, 1997), in strategic choice literature (Geletkanycz and Hambrick, 1997) and in organizational ecology and evolutionary studies (Barnett et al., 1994; Kelly and Amburguey, 1991; Sinha and Noble, 1997). Organizational age was measured by how many years the bank had been operating since the start of the modern Portuguese banking era in 1975. Characteristics associated with organizational form were also included through three dummy variables – specialism, foreignness, and stateowness – allowing the consideration of important historical and behavioural effects in this setting (see Research Context section).

Statistical Data Checks We tested Hypotheses 1 and 2 by using the SPSS statistical package to perform contingency table analyses for the branching variables: ‘basic growth choice’ and ‘location growth choice’. To test Hypothesis 3, we used STATA software to run a GLS (generalized least squares) regression for the cross-section-time-series analysis for performance effects due to ‘basic growth choice’ imitation, in the period 1989 to 1996, and SPSS to run an OLS (ordinary least squares) regression for the analysis of performance effects in 1992 due to ‘location growth choice’ imitation in 1991. To test Hypothesis 4, we also ran OLS regressions to examine whether the performance effects were differentiated contingent upon location attractiveness. © Blackwell Publishing Ltd 2006

1570

I. Barreto and C. Baden-Fuller

Table I. Descriptive statistics and correlations for the data on branching and profitability for 1992 Variable

Mean

s.d.

1

2

3

4

5

6

7

1. 2. 3. 4. 5. 6. 7. 8.

0.009 0.81 0.31 0.31 0.08 0.40 0.21 0.19

0.011 0.402 0.471 0.471 0.272 0.329 0.215 0.203

-0.394 0.258 -0.413 -0.334 -0.328 -0.230 -0.287

-0.520 0.325 0.141 0.605 0.485 0.466

-0.444 0.433 -0.775 -0.663 -0.553

-0.192 0.147 0.148 0.083

-0.358 -0.287 -0.276

0.800 0.773

0.237

ROA_92 spec_1991 fgn 1991 st_own1991 imibas91 prop_91 prop1_91 pr_att91

Table II. Descriptive statistics and correlations for the pooled data on branching and profitability from 1989 to 1996 Variable

Mean

s.d.

1

2

3

4

5

6

1. 2. 3. 4. 5. 6. 7.

0.009 0.54 14.00 0.82 0.32 0.25 74.02

0.014 0.72 5.35 0.38 0.47 0.44 121.70

-0.07 -0.44 -0.19 0.02 -0.12 -0.08

0.32 0.32 -0.45 0.25 0.81

0.40 -0.36 0.28 0.19

-0.54 0.27 0.28

-0.40 -0.28

-0.08

Performance Size ¥ 10-6 Age Specialism Foreignness Stateowness Cumulative growth choice

In the contingency tables analysis we estimate whether banks are following their reference group by using the marginal probabilities as laid out by Karl Pearson, using the Pearson chi-square and the likelihood-ratio chi-square tests. The ‘random’ prediction shown in the tables is the choice that would be made if there were no mimetic behaviour. The Goodman and Kruskal tau was used to assess the relative importance of alternative types of imitation (Norusis, 1993). For the GLS model, following several studies on strategic choice (e.g. Hambrick et al., 1996), we used a two-stage cross-section heteroskedastic and time-wise autoregressive model (Kmenta, 1997), which allows us to test and correct for both autocorrelation and heteroskesdasticity which may occur in pooled data. Tables I and II present descriptive statistics and correlations for the variables included in both GLS and OLS models. As usual in these analyses, we used the advanced diagnostics methods to investigate the existence of any multicollinearity problem for the regression models (Hair et al., 1995; Norusis, 1993). First, we examined the GLS model. Work by Hair et al. (1995) and Kennedy (1992) has established the minimum recommended cut-off value for the tolerance levels of each variable as 0.10 and the maximum recommended cut-off value for variance inflation factors as 10 and for condition indices as 30. Our results were well within these parameters; the lowest variable tolerance levels were 0.233, the highest variance inflation factor © Blackwell Publishing Ltd 2006

To Conform or To Perform?

1571

Table III. Bank ‘basic growth choice’: banks that follow the behaviour shown by their reference group in the previous period in terms of branch opening (1990–96) Basic growth group (year t - 1) No growth Basic growth (year t)

No growth Growth

Total

Count (Randomly expected count) Count (Randomly expected count) Count (Randomly expected count)

Total

Growth

34 (13.7)

17 (37.3)

51 (51.0)

15 (35.3)

116 (95.7)

131 (131.0)

49 (49.0)

133 (133.0)

182 (182.0)

Notes: The behaviour of the reference group is represented by the actions taken by the majority of them. Dependence tests: Pearson chi-square = 56.885 (significance level = 0.000). Likelihood ratio = 53.875 (significance level = 0.000).

4.292 and the highest condition indices 10.07. Thus these advanced methods detected no multicollinearity problem in the GLS model. We then examined the OLS models. The tolerance levels for each variable included in both models are all well above the recommended cut-off value (0.10), with the minimum values at 0.240, and 0.266 for models 1 and 2 respectively. The variance inflation factors are all far below the recommended cut-off value (10), with the highest values at 4.168, and 3.762 for Models 1 and 2 respectively. The condition indices are also all below the recommended cut-off level (30), the highest values being 11.799, and 12.538 for Models 1 and 2 respectively. In sum, no multicollinearity problem was detected in the OLS models. RESULTS Is there imitation? Who follows whom? Our first hypothesis predicted that there is imitation in branching decisions based on legitimacy-driven reference groups. We test this hypothesis first on ‘basic growth choice’ (crude data), and then upon ‘location growth choice’ (fine-grained data). Table III shows the results for the first test, which indicates that banks follow the behaviour of their reference group when they try to respond to changing environmental conditions. Pearson chi-square and likelihood ratio tests show that there is a significant dependence between ‘basic growth choice’ of individual banks in period t and ‘basic growth choice’ of banks of same legitimacy-based reference group in the previous period (t - 1). Of 182 cases (bank/periods), 150 (82 per cent) indicate imitation, irrespective of the growth choices made. When considering the specific strategies chosen, in 116 of the total of 131 cases (89 per cent) where a growth decision was made, the decision went the same way as by banks of similar form in the previous period. The second test on existence of mimetic behaviour concerns the fine-grained data on ‘location growth choice’ decisions. Table IV presents the findings about banks following © Blackwell Publishing Ltd 2006

1572

I. Barreto and C. Baden-Fuller

Table IV. Bank location growth choice: banks that follow the behaviour shown by any member of their group in the previous period in terms of locations chosen (1991) Location growth group90

Location growth91

No growth Growth

Total

Count (Randomly expected count) Count (Randomly expected count) Count (Randomly expected count)

Total

No growth

Growth

6810 (6688.1)

771 (892.9)

7581 (7581.0)

186 (307.9)

163 (41.1)

349 (349.0)

6996 (6996.0)

934 (934.0)

7930 (7930.0)

Notes: Banks were able (and in fact did) open more than one branch in a specific location; their choice is possible in every period. Dependence tests: Pearson chi-square = 428.588 (significance level = 0.000). Likelihood ratio = 281.271 (significance level = 0.000). Goodman and Kruskal tau = 0.054 (significance level = 0.000).

the same legitimacy-based reference group, where there are 7930 growth opportunities (corresponding to 305 counties and 26 banks involved). The Pearson chi-square and likelihood ratio tests show a highly significant dependence between the ‘location growth choice’ of individual banks in 1991 (that is, the counties they have chosen to open branches in during that period) and the ‘location growth choice’ of banks of similar legitimacy-based reference groups in the previous year. The strength of the observed imitation in that year is showed by the number of cases (163), where banks followed their groups in opening new branches in the same counties, which corresponds to four times the randomly expected number (about 41), while branch openings in counties other than those chosen by their reference group in the previous period (186) represents only about 60 per cent of the randomly expected number (about 308). These two complementary tests support Hypothesis 1, confirming the existence of imitation on important strategic choices and the importance of legitimacy forces in the dynamic of imitation processes. Are banks selective when they imitate regarding where they go? Do they tend to imitate only ‘good’ decisions and try to avoid ‘bad’ decisions? Hypothesis 2 addresses this second group of questions, and predicts the existence of imitation in both (ex ante) attractive and unattractive counties, due to the predominance of legitimacy pressures over economic considerations. Table V shows the results for both attractive and unattractive counties. For both subsets, the Pearson chi-square and likelihood ratio tests show a highly significant dependence between the ‘location growth choice’ of individual banks in period ‘t’ and the ‘location growth choice’ of banks of similar legitimacy-based reference groups in the previous period ‘t - 1’. The lower part of the table shows that, in 1991, banks had been able to exploit 68 of the 335 opportunities to open new branches in ex ante attractive counties, and that they choose to follow their legitimacy-based reference group in 53 of those cases (almost 80 per cent), a very strong signal of imitation. © Blackwell Publishing Ltd 2006

Total

Location growth91

Total

Location growth91

Growth

No growth

Growth

No growth

Count (Randomly expected count) Count (Randomly expected count) Count (Randomly expected count)

Count (Randomly expected count) Count (Randomly expected count) Count (Randomly expected count)

165 (165.0)

15 (33.5)

150 (131.5)

6831 (6831.0)

171 (252.7)

6660 (6578.3)

No growth

Location growth group90

170 (170.0)

53 (34.5)

117 (135.5)

764 (764.0)

110 (28.3)

654 (735.7)

Growth

Pearson chi-square = 25.245 (significance level = 0.000). Likelihood ratio = 26.522 (significance level = 0.000). Goodman and Kruskal tau = 0.075 (significance level = 0.000).

For attractive locations: Dependence tests:

Pearson chi-square = 272.863 (significance level = 0.000). Likelihood ratio = 175.774 (significance level = 0.000). Goodman and Kruskal tau = 0.036 (significance level = 0.000).

Notes: The data show banks that follow the behaviour shown by any member of their group in the previous period in terms of location chosen (1991): For unattractive locations: Dependence tests:

Attractive locations

Unattractive locations

Table V. Bank location growth choice with location attractiveness as the control variable:

335 (335.0)

68 (68.0)

267 (267.0)

7595 (7595.0)

281 (281.0)

7314 (7314.09)

Total

To Conform or To Perform? 1573

© Blackwell Publishing Ltd 2006

1574

I. Barreto and C. Baden-Fuller

Table VI. Branching profitability (1989–96) generalized least squares regression Variable

b

p values

Constant Size ¥ 10-6 Age Specialism Foreignness Stateownness Cumulative growth choice Wald chi-square Sig. N = 208

0.028 (0.003) 0.003 (0.001) -0.001 (0.000) -0.009 (0.002) -0.007 (0.002) -0.004 (0.001) -0.000013 (0.000006) 60.92 0.000

0.000 0.008 0.000 0.000 0.004 0.000 0.041

Notes: Standard errors are reported in parentheses. All values report to year t; ‘cumulative growth choice’ in year t is the number of new branches accumulated since 1988 to year t.

The upper part of the table shows that the pattern of mimetic behaviour is also present even in a priori unattractive counties: the number of cases where banks followed their groups in opening new branches in the same unattractive counties (110) is about four times the number randomly expected (about 28), while branch openings in counties different to those chosen by their reference group in the previous period (171) represents only about 68 per cent of the randomly expected number (about 253). The findings of Table V support our Hypothesis 2, confirming that mimetic behaviour is not selective regarding choice-targets, and reinforces the relevance of normative forces as opposed to economic forces in explaining imitation. Are there negative performance effects due to imitation? Hypothesis 3 predicts that branching imitation leads to lower firm profitability. ‘Basic growth choice’ imitation effects are analysed through a longer time-period from 1989 to 1996. This pooled time-series cross section analysis allowed us to cover a large period of environmental uncertainty. The results of the Generalized Least Squares (GLS) regression presented in Table VI show a significant negative association between accumulated branch-growth and bank ROA. Each additional branch opened in this period reduced the average bank return-on-assets by 0.0013 per cent. By the end of the period of analysis, many banks had greatly expanded their branch network, and for each 100 extra branches they opened, they would lower their ROA by almost 0.13 per cent due to imitation-driven branch growth. As the average ROA for our banks in this period was 0.9 per cent, the costs of mimetic behaviour seems large. The strong significance levels of all control variables included in the GLS model confirmed the a priori importance of historical and behavioural effects that we attributed to them in deciding to control for them. Once again we use fine-grained data to test ‘location growth choice’ (1991) imitation effects and the results are given in Table VII. Given that 1991 was the year when most branches were opened, and in order to capture the eventual causal relationship, 1992 was the ideal year to examine the performance consequences of actions in 1991. We ran OLS regressions on bank profitability, using prop_91 as the independent variable (to © Blackwell Publishing Ltd 2006

To Conform or To Perform?

1575

Table VII. Branching profitability (1992) ordinary least squares regression Model 1

Model 2

b Constant Size Age Specialism Foreignness Stateowness Imibas91 Prop_91 Pr_att91 Pr_una91 F Sig. Adj. R square

0.023 0.005 0.000 0.009 -0.003 -0.013 -0.028 -0.028

(0.007) (0.004) (0.000) (0.006) (0.006) (0.004) (0.007) (0.009)

4.501 0.005 0.495

p value

b

0.003 0.188 0.382 0.141 0.602 0.006 0.001 0.007

0.021 0.006 0.000 0.009 -0.003 -0.014 -0.028

p value (0.007) (0.004) (0.000) (0.006) (0.006) (0.004) (0.007)

0.006 0.151 0.521 0.155 0.680 0.006 0.001

-0.033 (0.012) -0.024 (0.011) 3.896 0.009 0.481

0.011 0.047

Note: Standard errors are reported in parentheses.

catch the location-based imitation effect). As control variables, we used the same variables as in GLS model, and also imibas91, to control for ‘basic growth choice’ imitation effect. Model 1 from Table VII shows a significant negative coefficient for prop_91, suggesting that the higher the degree of location-based imitation, the lower the bank profitability. The significant negative coefficient for imibas91 also confirms the results of the GLS model about the relevance of the performance effects of ‘basic growth choice’level imitation. The findings of Table VI and of Model 1 from Table VII give clear support to Hypothesis 3, confirming that legitimacy-based branching imitation leads to banks’ lower profitability. Our final set of hypotheses predicts that the negative profitability effect will be present for counties which are ex ante both attractive and unattractive locations (Hypothesis 4). Model 2 in Table VII is an OLS regression with the same variables as Model 1, above referred, except for the independent variables, which are here pr_att91 and pr_una91, intended to capture the attractive-unattractive effect. The coefficient of pr_att91 is negative and significant, meaning that the higher the branching imitation towards ex ante attractive counties the lower the profitability of involved banks. The coefficient of pr_una91 is also negative and significant. Thus Hypothesis 4 receives support, and these results endorse our claim about the adverse profitability effects of legitimacy-based imitation. DISCUSSION Neoinstitutional theory has recently shown an impressive and growing body of research, and the phenomenon of mimetic isomorphism has attracted a significant share of that © Blackwell Publishing Ltd 2006

1576

I. Barreto and C. Baden-Fuller

research. This study addresses some important questions that have received surprisingly little attention: Who imitates whom? Is there a selective imitation based on organization’s own previously-held information? Which are the effects on financial performance of mimetic isomorphism? Our study of Portuguese bank branching finds that: (1) banks imitate their legitimacy-based reference group; (2) banks engage in legitimacy-driven mimetic actions towards both previously considered attractive and unattractive counties; and (3) legitimacy-driven imitation imposes a negative contribution on banks profitability, and this holds true in either case. The contributions of our findings involve both organization theory and strategy fields; it also suggests some avenues for further research. This study contributes to neoinstitutional theory in several ways. First, we offer a generic, legitimacy-driven framework that firms may use to determine who to imitate. Our findings show that banks imitate their legitimacy-based reference groups, the subsets of competing banks categorized by legitimacy-providers. This result emphasizes the central importance of legitimacy in the managerial decision-making process: since mimetic isomorphism is about legitimacy-seeking actions, it should be expected that the choice of who follows whom would be linked in some way to legitimacy. We show that legitimacy-providers – the collective observers that assess the conformity of organizational behaviour to a specific standard (Suchman, 1995) – provide the crucial links. A second major contribution to institutional theory is made with the finding that there is imitation on branching even towards ex ante (firm-specific) unattractive counties. Accordingly, this study is the first of which we are aware that shows that firms engage in legitimacy-driven imitation even against their own a priori information. This result provides a very strong endorsement for legitimacy, mimetic isomorphism, and institutional ideas. First, it represents a powerful demonstration about the relevance of legitimacy, and second, it offers novel empirical evidence about the predominance of ‘normative rationality’ over ‘economic rationality’, as predicted by neoinstitutional theorists (Oliver, 1997). This also confirms that social justification, on many occasions, can be more powerful than economic reasoning even in a for-profit context, and that legitimacy considerations are no less important in for-profit environments than in not-for-profit contexts, in contrast to suggestions in earlier research (Singh et al., 1986, p. 174). We believe that this paper gives a third contribution to institutional theory by being the first to show that legitimacy-based imitation may lead to negative contributions to firms’ profitability. Previous work (Westphal et al., 1997) has found a negative impact on efficiency from hospitals’ conformity in an administrative innovation adoption (total quality management). We show a negative profitability effect carried on by a major strategic choice imitation by for-profit organizations, for ex ante ‘good’ decisions as well as for ‘bad’ ones, providing empirical evidence for prior suggestions about the existence of a trade-off between legitimacy and profitability (DiMaggio and Powell, 1983; Meyer and Rowan, 1977), and confirming the presence, in certain circumstances, of ‘normative rationality’ over ‘economic rationality’. This paper also provides some insights into strategy literature. Its contribution to a reference groups approach, for example, is twofold. First, it responds to the call for identification of alternative reference points (Fiegenbaum and Thomas, 1995, p. 472), by suggesting legitimacy-based groups as reference groups, since they act precisely as benchmarks against which group members compare and evaluate themselves when searching © Blackwell Publishing Ltd 2006

To Conform or To Perform?

1577

for new strategic options. Second, it begins to overcome the criticism that the reference point approach addressees convergence, but does not explain what drives divergence or differentiation (Fiegenbaum and Thomas, 1995, p. 467, note 7). In fact, the legitimacybased group concept suggests that the legitimacy process may be viewed not only at an individual strategic choice level, but also at a multiple strategic choice level. The legitimacy assessment might be made by comparing each organization against its reference group along all relevant strategic choices. It is likely that some practices or strategic choices will be more important than others in providing legitimacy; imitation would be important only in some practices or choices, while differentiation would be attained through others. Thus, legitimacy-based reference groups framework can explain both convergent and differentiation strategies. Additionally, the evidence that firms pursue mimetic actions despite the resulting negative contributions to profitability is consistent with some recent theoretical developments. O’Neill et al. (1998) assert that the occurrence and persistence of inefficient or unprofitable strategies should not be seen as surprising once some assumptions from rational choice perspective are dropped – namely assumptions that outcomes can be known a priori, and that managers are fully efficient or profit-oriented. This does not imply that managers or firms do not care about efficiency or profitability goals. In fact, they may be efficiency-seekers, but not efficiency-maximizers (Roberts and Greenwood, 1997). Or, alternatively, we can assume ‘almost all the time that business firms are mostly about money, and assuming only rarely that they maximize profits’ (Winter, 2003). The identification of some limitations of the study can shed some light on possible future research on these topics, beyond the lines implicit in the above discussion. We have used here a relatively short time frame in our fine-grained analyses, since evidence on imitation is relevant for research purposes regardless of the time persistence of the phenomenon. However, future research using longer time periods may yield interesting lessons about if (and when) mimetic behaviour persists, or whether periods of imitation are followed by periods of non-imitation. On the other hand, in our profitability analyses, we use return on assets in line with its frequent use in banking settings (and because most institutions involved were not listed); however, since return on assets has the limitations of any accounting profitability measure, future research is also needed to examine these questions by using market-based measures, such as market-to-book value ratio. Finally, since we have used a single industry and a single country setting, future research might attempt to replicate these findings in and across different contexts.

CONCLUSIONS This study extends research on inter-organizational imitation by offering new insights for an integrated approach, wholly centred on legitimacy, about who imitates whom, what is imitated, and with what effects. Our results confirm the existence of a trade-off that decision-makers have to address, between the pressures to conform and the pressures to perform. Our findings also relate to a central debate in strategic management, by showing, first, the occurrence of legitimacy-driven imitation, and, second, the emergence of negative profitability effects from imitation. We join other researchers (e.g. © Blackwell Publishing Ltd 2006

1578

I. Barreto and C. Baden-Fuller

Deephouse, 1999) in their call for the need to explain and predict both differentiation and homogenization phenomena. NOTES *We acknowledge the financial support of Fundação para a Ciência e a Tecnologia (‘Sub-Programa Ciência e Tecnologia do 2° Quadro Comunitário de Apoio’) for the development of this paper. We would like to thank our colleagues at our respective schools and the Academy of Management Meetings (2000 and 2002), with particular mention of Meziane Lasfer, Erik Larsen, and Henk Volberda for helpful comments and encouragement. Finally, we thank the editor and the reviewers of Journal of Management Studies for their helpful and thoughtful reviews. [1] We could also argue for the existence of some agglomeration effects. Agglomeration economies are positive externalities arising when firms locate close to other firms. Agglomeration effects are strong in footloose industries such as research and development clusters and entertainment clusters, and also strong in micro-location decisions, such as when retailers migrate to shopping centers or hotels to particular resorts. Demand-side agglomeration effects are those where the gains are associated with increased demand (Shaver and Flyer, 2000). These effects relate mainly to information search costs reduction by consumers. Geographic clustering may decrease consumer search costs, and this is more likely in industries where consumers need to inspect goods or services personally and compare before deciding (Chung and Kalnins, 2001), as in hotels, theatres, or restaurants. The benefit of this effect is the additional aggregate demand induced by agglomeration, that is, the number of customers is expected to be higher with co-location then those that agglomerated firms would attract individually (Kalnins and Chung, 2004). Here we are not looking at the head-quarter decisions of banks, but rather at branches used to gather retail accounts. Moreover, we are not looking at whether banks congregate along a particular high-street or even if they congregate in a particular locale: our unit of analysis – the county – is a metropolitan region. In our context, bank branching is not likely to meet to a significant degree any important conditions to benefit from agglomeration economies.

APPENDIX Details of banks and their branching activity for 1990–91 Bank name

Branches (end of 1989)

Generalist private local BANIF 19 BCI 19 BCP 51 BPA 143 BTA 97 MG 32 Generalist foreign owned BARCLAYS 3 BBV 12 CHEMICAL 6 CL 15 Generalist state owned BBI 99 BCA 13 BESCL 154 © Blackwell Publishing Ltd 2006

Change in no. of branches in 1990

Branches (end of 1990)

Change in no. of branches in 1991

Branches (end of 1991)

4 6 80 14 33 5

23 25 131 157 170 37

5 59 110 47 32 15

28 84 241 204 202 52

5 0 0 3

8 12 6 18

21 10 0 1

29 22 6 19

0 2 9

99 15 163

8 4 57

107 19 220

To Conform or To Perform?

1579

Details of banks and their branching activity for 1990–91 Continued Bank name

Branches (end of 1989)

BFB 100 BFE 22 BNU 150 BPSM 150 UBP 128 CGD 401 CPP 62 Specialist private banks BIC 3 BPI 2 Specialist foreign banks CITI 2 B.BRASIL 2 BNP 2 GENERALE 1

Change in no. of branches in 1990

Branches (end of 1990)

Change in no. of branches in 1991

Branches (end of 1991)

14 2 0 38 5 14 13

114 24 150 188 133 415 75

5 6 11 6 19 38 31

119 30 161 194 152 453 106

3 0

6 2

6 0

12 2

0 0 0 0

2 2 2 1

0 0 0 0

2 2 2 1

REFERENCES Abrahamson, E. and Rosenkopf, L. (1993). ‘Institutional and competitive bandwagons: using mathematical modeling as a tool to explain innovation diffusion’. Academy of Management Review, 18, 487–517. Avery, C. and Zemsky, P. (1998). ‘Multidimensional uncertainty and herd behavior’. American Economic Review, 88, 4, 724–43. Banerjee, A. V. (1992). ‘A simple model of herd behavior’. Quarterly Journal of Economics, 107, 3, 797–818. Barnett, W. P., Greve, H. R. and Park, D. Y. (1994). ‘An evolutionary model of organizational performance’. Strategic Management Journal, 15, 11–28. Bikhchandami, S., Hirshleifer, D. and Welch, I. (1992). ‘A theory of fads, fashion, custom, and cultural change as informational cascades’. Journal of Political Economy, 100, 992–1024. Burns, L. R. and Wholey, D. R. (1993). ‘Adoption and abandonment of matrix management programs: effects of organizational characteristics and interorganizational networks’. Academy of Management Journal, 36, 106–38. Chung, W. and Kalnins, A. (2001). ‘Agglomeration effects and performance: a test of the Texas lodging industry’. Strategic Management Journal, 22, 969–88. Cool, K., Roller, L. H. and Leleux, B. (1999). ‘The relative impact of actual and potential rivalry on firm profitability in the pharmaceutical industry’. Strategic Management Journal, 20, 1, 1–14. Cyert, R. and March, J. (1963). A Behavioral Theory of the Firm. Englewood Cliffs, NJ: Prentice-Hall. Dacin, M. T. (1997). ‘Isomorphism in context: the power and prescription of institutional norms’. Academy of Management Journal, 40, 46–81. Dacin, M. T., Goodstein, J. and Scott, W. R. (2002). ‘Institutional theory and institutional change’. Academy of Management Journal, 45, 1, 45–57. Deephouse, D. L. (1996). ‘Does isomorphism legitimate?’. Academy of Management Journal, 39, 1024–39. Deephouse, D. L. (1999). ‘To be different, or to be the same? It’s a question (and theory) of strategic balance’. Strategic Management Journal, 20, 147–66. Devenow, A. and Welch, I. (1996). ‘Rational herding in financial economics’. European Economic Review, 40, 603–15. DiMaggio, P. J. and Powell, W. W. (1983). ‘The iron cage revisited: institutional isomorphism and collective rationality in organizational fields’. American Sociological Review, 48, 147–60. © Blackwell Publishing Ltd 2006

1580

I. Barreto and C. Baden-Fuller

Dutton, J. E. and Dukerich, J. M. (1991). ‘Keeping an eye in the mirror: image and identity in organizational adaptation’. Academy of Management Journal, 34, 517–54. Dutton, J. and Jackson, S. (1987). ‘Categorizing strategic issues: links to organizational action’. Academy of Management Review, 12, 1, 76–90. Fiegenbaum, A. and Thomas, H. (1995). ‘Strategic groups as reference groups: theory, modeling and empirical examination of industry and competitive strategy’. Strategic Management Journal, 16, 6, 461–76. Fiegenbaum, A., Hart, S. and Schendel, D. (1996). ‘Strategic reference point theory’. Strategic Management Journal, 17, 219–35. Fligstein, N. (1991). ‘The structural transformation of American industry: an institutional account of the causes of diversification in the largest firms, 1919–1979’. In Powell, W. W. and DiMaggio, P. J. (Eds), The New Institutionalism in Organizational Analysis. Chicago, IL: University of Chicago Press, 311–36. Geletkanycz, M. and Hambrick, D. (1997). ‘The external ties of top executives: implications for strategic choice and performance’. Administrative Science Quarterly, 42, 654–81. Greve, H. R. (1998). ‘Managerial cognition and the mimetic adoption of market positions: what you see is what you do’. Strategic Management Journal, 19, 10, 967–88. Greve, H. R. (2000). ‘Market niche entry decisions: competition, learning and strategy in Tokyo banking: 1894–1936’. Academy of Management Journal, 43, 5, 816–36. Greve, H. R. (2001). ‘Market niche entry decisions: competition, learning, and strategy in Tokyo banking, 1894–1936’. Academy of Management Journal, 44, 5, 816–36. Gual, J. and Neven, D. (1992). Deregulation of the European Banking Industry (1980–1991). CEPR Discussion Paper no. 703. London: Centre for Economic Policy Research. Guillén, M. (2002). ‘Structural inertia, imitation, and foreign expansion: South Korean firms and business groups in China, 1987–95’. Academy of Management Journal, 45, 3, 509–25. Hair, J. F., Anderson, R. E., Tatham, R. L. and Black, W. C. (1995). Multivariate Data Analysis with Readings, 4th edition. Englewood Cliffs, NJ: Prentice-Hall. Hambrick, D. and Mason, P. (1984). ‘Upper echelons: the organization as a reflection of its top managers’. Academy of Management Review, 9, 2, 193–206. Hambrick, D. C., Cho T. S. and Chen, M.-J. (1996). ‘The influence of top management team heterogeneity on firms’ competitive moves’. Administrative Science Quarterly, 41, 659–84. Hannan, M. T. and Freeman, J. (1977). ‘The population ecology of organizations’. American Journal of Sociology, 82, 929–64. Haunschild, P. R. (1993). ‘Interorganizational imitation: the impact of interlocks on corporate acquisition activity’. Administrative Science Quarterly, 38, 564–92. Haunschild, P. R. and Miner, A. S. (1997). ‘Modes of interorganizational imitation: the effects of outcome salience and uncertainty’. Administrative Science Quarterly, 42, 472–500. Haveman, H. A. (1993). ‘Follow the leader: mimetic isomorphism and entry into new markets’. Administrative Science Quarterly, 38, 593–627. Henderson, A. (1999). ‘Firm strategy and age dependence: a contingent view of the liabilities of newness, adolescence, and obsolescence’. Administrative Science Quarterly, 44, 281–314. Henisz, W. J. and Delios, A. (2001). ‘Uncertainty, imitation, and plant location: Japanese multinational corporations, 1990–1996’. Administrative Science Quarterly, 46, 3, 443–75. Huff, A. (1990). Mapping Strategic Thought. Chichester: Wiley. Huff, A. S. (1982). ‘Industries influences on strategy reformulation’. Strategic Management Journal, 3, 2, 119–31. Hybels, R. C. (1995). ‘On legitimacy, legitimation, and organizations: a critical review and integrative theoretical model’. In Moore, D. P. (Ed.), Academy of Management Best Proceedings, 241–5. Ilmakunnas, P. (2002). ‘Strategic behavior in a service industry’. Managerial and Decision Economics, 23, 69–82. Kennedy, P. (1992). A Guide to Econometrics, 3rd edition. Cambridge, MA: MIT Press. Kalnins, A. and Chung, W. (2004). ‘Resource-seeking agglomeration: a study of market entry in the lodging industry’. Strategic Management Journal, 23, 689–99. Kelly, D. and Amburguey, T. L. (1991). ‘Organizational inertia and momentum: a dynamic model of strategic change’. Academy of Management Journal, 34, 591–612. Kmenta, J. (1997). Elements of Econometrics, 2nd edition. Ann Arbor, MI: The University of Michigan Press. Kraatz, M. S. (1998). ‘Learning by association? Interorganizational networks and adaptation to environmental change’. Academy of Management Journal, 41, 6, 621–43. Lee, K. and Pennings, J. M. (2002). ‘Mimicry and the market: adoption of a new organizational form’. Academy of Management Journal, 45, 1, 144–62. Meyer, J. W. and Rowan, B. (1977). ‘Institutional organizations: formal structure as myth and ceremony’. American Journal of Sociology, 83, 340–63. © Blackwell Publishing Ltd 2006

To Conform or To Perform?

1581

Miller, T. (1997). ‘Banks branch out’. The Banker, August, 72–3. Norusis, M. J. (1993). SPSS for Windows: Base System User’s Guide. Release 6.0. Chicago, IL: SPSS Inc. Oliver, C. (1997). ‘Sustainable competitive advantage: combining institutional and resource-based views’. Strategic Management Journal, 18, 9, 697–713. O’Neill, H. M., Pouder, R. W. and Buchholtz, A. K. (1998). ‘Patterns in the diffusion of strategies across organizations: insights from the innovation diffusion literature’. Academy of Management Review, 23, 1, 98–114. Porac, J. F. and Thomas, H. (1990). ‘Taxonomic mental models in competitor definition’. Academy of Management Review, 15, 2, 224–40. Porac, J. F., Thomas, H. and Baden-Fuller, C. (1989). ‘Competitive groups as cognitive communities: the case of Scottish knitwear manufacturers’. Journal of Management Studies, 26, 397–416. Porter, M. E. (1980). Competitive Strategy: Techniques for Analyzing Industries and Competitors. New York: The Free Press. Porter, M. E. (1998). Banking Strategies. January/February, 40. Ramaswamy, K. (1997). ‘The performance impact of strategic similarity in horizontal mergers: evidence from the U.S. banking industry’. Academy of Management Journal, 40, 3, 697–715. Rao, H., Greve, H. R. and Davis, G. F. (2001). ‘Fool’s gold: social proof in the initiation and abandonment of coverage by Wall Street analysts’. Administrative Science Quarterly, 46, 3, 502–26. Reger, R. K. and Huff, A. S. (1993). ‘Strategic groups: a cognitive perspective’. Strategic Management Journal, 14, 2, 103–24. Reger, R. K., Durhaime, I. M. and Stimpert, J. L. (1992). ‘Deregulation, strategic choice, risk, and financial performance’. Strategic Management Journal, 13, 3, 189–204. Roberts, P. W. and Greenwood, R. (1997). ‘Integrating transaction cost and institutional theories: toward a constrained-efficiency framework for understanding organizational design adoption’. Academy of Management Review, 22, 2, 346–73. Scharfstein, D. S. and Stein, J. C. (1990). ‘Herd behavior and investment’. The American Economic Review, 80, 465–79. Shaver, J. M. and Flyer, F. (2000). ‘Agglomeration economies, firm heterogeneity, and foreign direct investment in the United States’. Strategic Management Journal, 21, 1175–93. Schwenk, C. (1984). ‘Cognitive simplification processes in strategic decision-making’. Strategic Management Journal, 5, 111–28. Scott, W. R. (1987). ‘The adolescence of institutional theory’. Administrative Science Quarterly, 32, 493–511. Silverman, B., Nickerson, J. and Freeman, J. (1997). ‘Profitability, transactional alignment, and organizational mortality in the U.S. trucking industry’. Strategic Management Journal, 18, Summer Special Issue, 31–52. Singh, J. V., Tucker, D. J. and House, R. J. (1986). ‘Organizational legitimacy and the liability of newness’. Administrative Science Quarterly, 31, 171–93. Sinha, R. K. and Noble, C. H. (1997). ‘The performance consequences of subfield entry’. Strategic Management Journal, 18, 6, 465–81. Suchman, M. C. (1995). ‘Managing legitimacy: strategic and institutional approaches’. Academy of Management Review, 20, 571–610. Westphal, J. D., Gulati, R. and Shortell, S. M. (1997). ‘Customization or conformity: an institutional and network perspective on the content and consequences of TQM adoption’. Administrative Science Quarterly, 42, 161–83. Whalen, G. (1992). ‘The determinants and performance effects of rivalry in local banking markets’. Quarterly Journal of Business and Economics, 31, 2, 38–52. Williamson, I. A. and Cable, D. M. (2003). ‘Organizational hiring, patterns, interfirm network ties, and interorganizational imitation’. Academy of Management Journal, 46, 3, 349–58. Winter, S. (2003). ‘Will economics ever change? Should it?’. In SidFest – A Tribute to Sidney G. Winter. The Wharton School, Philadelphia. Wiseman, R. and Catanach, A. (1997). ‘A longitudinal disaggregation of operational risk under changing regulations: evidence from the savings and loan industry’. Academy of Management Journal, 40, 4, 799–830.

© Blackwell Publishing Ltd 2006