Taking off: Category contrast and organizational mortality in the UK airline industry,1919

Abstract

This study analyzes the role of Category contrast in the context of an emerging industry. The level of contrast, the degree to which a category stands out from its background, appears to be an important aspect of the legitimation process, but as of yet poorly understood. The article explores three potential roles: as providing legitimacy to the category as such, as allowing for legitimacy spillovers to peripheral members, and as strengthening density-dependent legitimation. The study investigates these potential roles using longitudinal data on organizational naming patterns in the early passenger airline industry in the United Kingdom.

Keywords

Airlines categories contrast form emergence labels legitimacy

Introduction

Analyzing and understanding the dynamics of emergent industries has been of long-standing interest to academic researchers. Indeed, by embodying new technologies (Nelson and Winter, 1982) and by offering new organizational solutions to deal with the challenges posed by social and economic change (Hannan, 1986), new industries constitute a major form of source of change in economic systems. Industrial economists describe the process of industry emergence in terms of standardization, falling prices, and growing output (Gort and Klepper, 1982; Klepper and Graddy, 1990). Conversely, organizational sociologists describe industry emergence with reference to the legitimacy of a new organizational category or logic, defined as the degree to which that category or logic is taken for granted by relevant actors (see e.g., Carroll and Hannan, 2000; DiMaggio and Powell, 1983; Haveman and Rao, 1997; Maguire et al., 2004; Marquis and Lounsbury, 2007).

While legitimacy itself has been an object of fierce discussion among organizational sociologists, a substantial convergence has emerged on taken-for-grantedness as ‘the most powerful source of legitimacy identified to date’ (Suchman, 1995: 583). Taken-for-grantedness indicates a level of cognitive legitimacy associated with the absence of questioning (Deephouse and Suchman, 2008; Zucker, 1983). Initially, institutional theorists for instance have focused on isomorphic pressures as the key force behind the taken-for-grantedness of organizational forms and practices (DiMaggio and Powell, 1983); ecologists instead have focused on the legitimating role of organizational density, defined as the number of organizations in a population (Hannan and Carroll, 1992). New insights on this matter have however originated from the work of Zuckerman (1999) and from the consideration of categorization and expected codes of conduct as critical elements of taken-for-grantedness.

The work of Zuckerman (1999) has also made clear our limited understanding of processes of expectation formation among audience members (see also Khaire and Wadhwani, 2010; Navis and Glynn, 2010; Wry et al., 2011). In this respect, the concept of contrast introduced by ecologists (Hannan et al., 2007) and developed from the insights of cognitive psychologists (Rosch et al.,1976; Tversky, 1977), appears particularly promising. Category contrast may be defined as the degree to which a class of organizations stands out from its domain – an essential feature of cognitive legitimacy for external audiences. Contrast denotes the existence of ‘clear lines of demarcation’ (Durkheim and Mauss, 1903) among classes of organizations, and has been argued to be positively associated with taken-for-grantedness and the use of defaults (Hannan et al., 2007). Interestingly – and very much in alignment with the early insights of institutional theory (e.g. DiMaggio and Powell, 1983) – category contrast increases upon the emergence of substantial similarity within a class of organizations.

A few empirical studies have started to explore the role of contrast as a moderator of the illegitimacy discounts that organizations face when crossing categorical boundaries (Kovács and Hannan, 2010; Negro et al., 2010; see also Ruef and Patterson, 2009). Little remains known however about the role of category contrast for the process of form emergence and, in particular, about the relation between density and category contrast (for exceptions see Bogaert et al., 2010; Kuilman and Li, 2009). Does contrast contribute to legitimacy above and beyond density or is density still a primary driver of legitimacy as assumed in density dependence models (Hannan and Carroll, 1992)? Provided that contrast is a source of cognitive legitimacy, which organizations benefit the most from legitimacy spillovers? Do contrast and density reinforce each other in triggering cognitive legitimacy?

To remedy this state of affairs, the present article contends that recognizing the early role of contrast in the legitimation process is conducive to a richer account of the emergence of organizational populations. We develop our argument in detail as we describe and test for various potential roles of contrast. By juxtaposing the impact of contrast to that of density, the article contributes to the extant literature by unraveling the mechanisms behind the emergence of new organizational populations. A novel measurement of category contrast – as related to organizational naming patterns – is also advanced. We study these issues in the early UK passenger airline industry during the period 1919–64.

Theoretical background

Organizational categories

Studying the emergence of a new organizational population presupposes that a common definition exists on the basis of which the members of a population can be identified. While organizational populations are routinely defined as sets of similar organizations within a social system, diverse views on the basis of such similarity-based clustering have been proposed. For instance, the members of an organizational population are expected to share common environmental dependencies, to exhibit similar patterns of activities and the same core features (Hannan and Freeman, 1977, 1984).

More recently, however, it has been proposed that a population can be more fruitfully defined as a socially constructed organizational category (DiMaggio and Powell, 1991; Hannan et al., 2007). In particular, Hannan et al. (2007) have described the process of similarity-based clustering with reference to external audiences. Audience members – which may include critics (Negro et al., 2010), distributors (Cattani et al., 2008), analysts (Zuckerman, 1999), and consumers (Hsu et al., 2009) – classify organizations on the basis of their perceived similarities and attach a label to describe the entire class to which such organizations belong. Once there is some convergence in the use of meaningful labels, a class of organizations becomes a social category.

The contrast that an organizational category exhibits in the eyes of audience members has been argued to be central to this process (see Hannan et al., 2007). Research on category contrast has progressed into two major areas so far: research on contrast as a moderator of illegitimacy discounts that organizations face when not conforming to pre-existing categories (Kovács and Hannan, 2010; Negro et al., 2010); and research on the legitimacy of new organizational categories with particular attention to the role of contrast (Bogaert et al., 2010; Kuilman and Li, 2009). This latter branch of research is of primary concern to the present study.

Role of contrast in category emergence

The notion of contrast builds on the idea that an individual organization is not necessarily either a member or an outsider to an organizational category (as assumed by earlier work in organizational ecology (e.g., Hannan and Carroll, 1992), but it rather exhibits a grade of membership (GoM), defined as its degree of typicality as a member of the claimed category (Negro et al., 2011). Indeed, organizations may be seen as only partial members of a category: in some ways they might be seen as such, but in others not. The GoM function maps organizations on a [0, 1] interval to take into account such partial memberships. What separates the notion of GoM from other dimensions on which organizational similarity could be mapped – e.g., geographical location or price – is that GoM builds on the perspective of audience members who seek to match the perceived feature values of an organization with the schema of the claimed category, which does not necessarily involve location or price.

Contrast then, a category-level concept, can empirically be represented by the population’s average GoM (Bogaert et al., 2010; Hannan et al., 2007). The notion of contrast essentially captures the sharpness of a category’s boundaries, as well as similarities among members within the category. A category with a low degree of contrast has more members that only partly fit in, i.e. they have a lower GoM. The boundaries that demarcate such a hypothetical category from other organizational categories tend to be blurred. Conversely, a category with a high degree of contrast contains a high proportion of organizations with a high grade of membership, i.e., organizations that can be regarded as highly typical and have a high degree of resemblance to the rest of the population. According to Hannan et al. (2007: 46) a high proportion of such firms implies that the category ‘stands out with greater contrast from its background’. The notions of contrast and GoM have been fruitfully used in a growing number of studies (Bogaert et al., 2010; Kovács and Hannan, 2010; Kuilman and Li, 2009; Negro et al., 2010, 2011).

According to Hannan et al. (2007), the legitimacy of a population grows monotonically with its ‘contrast’. Contrast enhances the cognitive legitimacy of a population, because it makes a population more visible and focused. One reason for this is that in high contrast populations, organizations are likely to exhibit similar features (Hannan et al., 2007). Indeed, there exists ample evidence that homogeneity among organizations can strengthen the development of a collective identity (e.g., DiMaggio and Powell, 1983; McKendrick et al., 2003). Elevated similarity among organizations facilitates also the use of defaults by external audiences (Hannan et al., 2007). High contrast thus implies that it is easier for the corresponding population to gain social recognition, as it is easier to acknowledge its distinguishing features.

While the notion of contrast has been applied to understanding the consequences of organizational and individual behavior in the context of existing categories (Kovács and Hannan, 2010; Negro et al., 2010), the role of contrast in the emergence of new industries remains relatively underexplored. Few studies have thus far provided mixed evidence about the role of contrast. Most notably, Bogaert et al. (2010), in their study of the emerging accounting industry in the Netherlands, have shown that category contrast is negatively related to mortality among accounting firms. When the population included many marginal professional associations representing diverse interests and differing professional standards, this diluted the contrast between accounting and bookkeeping or consultancy. They found that at higher levels of category contrast, individual accounting firms faced lower mortality risks. However, Kuilman and Li (2009) found that that peripheral foreign banks in Shanghai were not able to derive specific benefits from the contrast of their population. Instead, various other measures of legitimacy, including how often the population was mentioned in public media, were better able to capture this process.

Existing research therefore leaves open a few questions. First, given the lack of univocal findings, it remains unclear whether increasing contrast indeed improves the viability of members of emerging populations. Second, and moving beyond prior research, if contrast represents a source of legitimacy, which organizations benefit the most from the legitimacy spillovers? Furthermore, the relation between density and contrast remains scantly understood. Do contrast and density operate independently or do they reinforce each other in sustaining the legitimacy of an emergent form? In the next section we address these questions.

Hypotheses

According to density dependence theory, when a population is emerging, its density is negatively related to organizational mortality (Hannan and Carroll, 1992). The argument states that as density grows, the social recognition of the organizational population improves and mortality rates decline. Notwithstanding the variations of the theory concerning different levels of analysis as well as over time, the empirical evidence in favor of density as a proxy for cognitive legitimacy appears robust (see Carroll and Hannan, 2000 for a review).

Nonetheless, organizational density has been also heavily criticized (Baum and Powell, 1995) and richer ways of capturing cognitive legitimacy have been advocated (e.g., Kennedy, 2008; Mohr and Guerra-Pearson, 2010). Indeed, even among ecologists, the beneficial effect postulated by density dependent legitimation has been debated (see e.g., Delmestri and Wezel, 2011; Dobrev and Gotsopoulos 2010; Dobrev et al., 2006). Recent theoretical developments within ecology acknowledge that form emergence takes place in an institutional vacuum due to the lack of a social category from which legitimacy can be sourced (Hannan et al., 2007; Hsu and Hannan, 2005). Only once the similarities among a class of organizations and their differences from alternative forms are recognized, does a population become established (McKendrick et al., 2003; Pólos et al., 2002). The extent of contrast exhibited by a category in the eyes of audience members reflects this process.

One of the novelties of this new approach to cognitive legitimacy is to recognize that category contrast may shape legitimacy independently from density: different levels of contrast may be obtained from the same value of organizational density and, indeed, empirical evidence suggests that even small populations get legitimized (see e.g., Navis and Glynn, 2010; for a discussion see also Wry et al., 2011). Therefore, net of organizational density, greater category contrast should facilitate the identification of a class of organizations as legitimate, and help managers and business owners in mobilizing resources and in improving their organization’s viability. These arguments suggest the following:

Hypothesis 1. Net of organizational density, population-level contrast is negatively related to organizational mortality rates.

The legitimacy benefits of high contrast may however vary across organizations and in particular, low GoM organizations may benefit the most from this population-level social recognition. Frequently, low GoM firms are unknown and lack institutional linkages (Baum and Oliver, 1992). Unless other organizational-level strategies are implemented – e.g., common naming (see Baum, 1999) – low GoM firms are not easily identified as members by audiences. But when the overall legitimacy of the population is enhanced, for instance by setting up of a trade association or introducing professional standards, the low GoM firms are likely to be the ones who derive the greatest marginal benefits (Bogaert et al., 2010). This is because high GoM firms obtain legitimacy through their good fit with the population and their established links with the institutional environment. As a result, they are less dependent on legitimacy spillovers (Baum and Oliver, 1992; Bogaert et al., 2010; Navis and Glynn, 2010). The low GoM firms instead may be more dependent on their membership in the population for obtaining resources. Increases in population-level legitimacy may thus benefit low GoM firms more than high GoM firms – a reasoning parallel to that developed by Park and Podolny (2000) on the power and independence that high status firms exhibit.

Albeit not framed in terms of GoM, Baum and Oliver’s (1992) study on failure rates among Toronto day care centers (DCCs) from 1971 and 1989 provides evidence in support of this argument. They compared those day care centers connected to the institutional environment (e.g., sharing a site with a church or a public school) with those who were not, and speculated that, ‘Our research raises the possibility of institutional free-rider effects in institutionalized populations. As the DCC population matured and its members developed numerous relations with the institutional environment, those DCCs that remained detached from the institutional environment appeared to obtain survival benefits from being a member of a highly institutionalized population’ (Baum and Oliver, 1992: 556). Kuilman and Li (2009), in their study of foreign banks in Shanghai between 1847 and 1935, tested the general idea that low GoM members of this emerging population benefited more from a surging cognitive legitimacy than high GoM members, and found general support for this thesis. They employed several measures of population-level legitimacy, but out of these measures, contrast appeared to be least likely to act as a source of legitimacy spillovers from which low GoM firms can benefit.

Given the uncertain state of the art about the benefits of contrast to low GoM organizations, we propose the following hypothesis:

Hypothesis 2. Organizations that display a lower GoM benefit more from the legitimacy spillovers generated by greater population-level contrast.

While operating independently, contrast and density may also reinforce each other. Although it is nowadays argued that density may not trigger legitimacy if a population lacks contrast, the converse may be true: at increasing values of contrast, the beneficial effects of organizational density for legitimacy may be even reinforced.

Theoretically, during the process of form emergence, the chances of a cluster of organizations of becoming taken-for-granted increase with high density and high contrast (Hannan et al., 2007). When contrast is coupled with density, it should become natural for outsiders to consider the focal organizational population as cognitively legitimized. Some empirical evidence appears supportive of this claim. For instance, in a study concerning the emergence of Singaporean cooperative banks, Dobrev et al. (2006) found that density dependent legitimation was stronger when (1) this banking form cohered into a single category distinct enough from existing financial institutions to avoid violations by comparison, and when (2) this new category was less diluted by corporate name changes that reduced the perceived focus of the newly emerging category. Similarly, Delmestri and Wezel (2011), in their study of founding rates of multiplex cinemas across Europe, found that contestation took place early on in the industry and that the positive effect of density dependent legitimation became amplified by the consolidation and defense of the multiplex as a distinct organizational category – i.e., once enough contrast for the multiplex category was established. Moreover, the findings obtained by Bogaert et al. (2010: Table 3, Model 12) underscore the amplifying effect of contrast and density on legitimacy. As they put it, ‘density decreases exits when contrast is high’ (Bogaert et al., 2010: 141).

Building on these arguments we advance that category contrast amplifies the beneficial effects of density on legitimation. In this sense, category contrast strengthens the impact of organizational density, aside from any independent impact that contrast may have on mortality. At increasing levels of density, contrast facilitates form emergence by making these more numerous – and by definition more homogeneous – firms more visible, ultimately spurring taken-for-grantedness. Therefore, we propose the following:

Hypothesis 3. Greater population-level contrast strengthens the negative relationship between density and organizational mortality rates.

Data and methods

Data sources

The hypotheses were tested using data collected on every passenger airline in the UK from 1919 to 1964. The UK airline industry is an excellent setting for our purposes because the quest for legitimacy turned out to be substantial for early airlines that aimed at establishing a very innovative service. Critical audiences in this setting were consumers and investors who controlled access to much-needed resources. Second, unique relative to other studies set in this industry (Baum and Korn, 1996; Seidel, 1997) is the fact that data on the complete population are available, therefore allowing us to study the progression from the first organizational founding through its development as a taken-for-granted population.

Many observers define the ‘natural’ end of the legitimation process of the airline industry and the consideration of airlines as the creation of the safe and efficient turbojets in the 1960s. Indeed, studies employing data of this industry employed 1962 (e.g., Kelly and Amburgey, 1991) or even later (e.g., Smith et al., 1991) – although most of this research focuses on the US airline industry. We followed Halford-MacLeod’s (2007) approach by ending our observation window in 1964. This year saw a change in the institutional environment with the end of Conservative rule, closing a period in which the industry had become increasingly structured following amalgamations triggered by the 1960 Civil Aviation Act and Air Transport Licensing Board (Halford-MacLeod, 2007). More importantly, ending the observation window in 1964 also excludes the competitive complications related to deregulation, the formation of international alliances, and the emergence of low cost carriers.¹

Merton Jones (1976), supplemented with data from Smith (2002), provided information on the date of each firm’s first flight, the date when an airline ceased to exist, as well as data on fleets. The data covered passenger airlines offering scheduled and/or charter services, but excluded helicopter companies. It also excluded all-cargo airlines, but not passenger airlines partially involved in the cargo business. The final dataset included 2293 airline-year observations, representing 298 airlines, 152 of which were disbanded during our window of observation. Among those that disbanded, the average age was 4.9 years.

The emergence of the UK airline industry

As noted, the emergence of the UK airline industry is one characterized by significant legitimacy struggles. The difficulties of many airlines in trying to keep afloat in the 1920s and 1930s were such that ‘few if any British air operators ever made a real profit in the inter-war years’ (Dyos and Aldcroft 1969: 373). Weak demand for civil aviation and high fares left much of the capacity underutilized. Most importantly, there was a lack of public recognition, acceptance, and trust in the air transport industry. A 1938 editorial in Flight summarized the above-mentioned problem as follows:

Numerous companies were formed and a number of air services were inaugurated. Their history has, on the whole, been a sad one. Many were under-capitalised, the personnel were lacking in experience, the aircraft used were unsuitable, and the complete lack of navigational aids made regularity or punctuality impossible. The travelling public were not ‘ready,’ and in the majority of cases the methods of airline operation did little to create confidence in this new method of travel. It was a period of the survival of the fittest. (Flight, 27 January 1938: 84)

Moreover, as one of the correspondents of the Flight magazine put it a few years earlier,

The growth of aviation depends largely upon the airmindedness of British people. A desirable form of aero-mentality must therefore be engendered, not only in those who fly or want to fly, but also in all those who stay on the ground and do not want to fly. Even indifference is an unfavorable attitude. (Flight, 19 December 1930: 1464)

So, notwithstanding an apparent interest in aeronautical matters, a general lack of airmindedness among the general public remained up until the ensuing of the Second World War. Airmindedness is a fundamental concept in understanding the development of the social category of airline producers and, in our reading of the industry, it is closely intertwined with the concept of taken-for-grantedness. Indeed, as Palmer (2006: 2) put it, ‘airmindedness’ refers to ‘the particular set of cultural traditions, symbols, and markers that, combined with existing political culture and social institutions, constitute a given nation’s response to the airplane’. As mentioned above, a lack of airmindedness for instance greatly constrained the use of aviation by the public before the Second World War. As a result, flying, whether as passengers or pilots, remained an activity of a small community of early enthusiasts during these years.

After the war, this situation changed substantially and the airline industry progressively became taken-for-granted for the mass audience. Flying and the formation of new passenger airlines had become substantially more common and accepted compared to the years prior to the Second World War. First, fuelled by a surfeit of suddenly redundant military aircrew and the availability of surplus aircraft at knock-down prices after the war, the number of airlines increased to peak in 1949. This was followed by a brief shake-out and a subsequent stabilization in airline numbers (see Figure 1). Ecological theory holds that such a pattern signals that a population has become taken-for-granted (Carroll and Hannan, 2000: 222–8; McKendrick et al., 2003). Not surprisingly, airline activity increased tremendously after the war. As Figure 2 shows, the annual number of aircraft miles flown by airlines in Britain quadrupled from 33 million in 1946 to 127 million by the end of our observation window (Mitchell, 1988).

Figure 1.

Number of passenger airlines in the UK, 1919–64

Figure 2.

Number of aircraft miles flown by UK airlines, 1919–64

Second, industry historians stress the importance of the Second World War in the development of the industry and that a major by-product of the war was that air travel gained significant legitimacy. For example, according to Fearon (1985),

War needs had led to a considerable improvement in many airports, in navigational aids such as radar, in safety and in the number of pilots … the public had become more ‘air-minded’ as [during the war] they were constantly reminded of the power of the aeroplane. (Fearon, 1985: 35)

Independent variables

The measurement of category contrast is in its infancy. The limited research available has employed the concentration of membership across industry associations (Bogaert et al., 2010), national origins (Kuilman and Li, 2009) or production styles (Kovács and Hannan, 2010; Negro et al., 2010). In the present article we propose a novel way to operationalize contrast: from the category labels incorporated in organizational names. Organizations use names to position themselves in the competitive arena and to manage the relationships with external audiences, such as customers (Chuang and Baum, 2003; Ingram, 1996). Name changes, albeit not without risk, can also signal a novel identity of the organization to existing audiences and/or broaden the organizational appeal (Dobrev et al., 2006; Lee, 2001). Dobrev et al. (2006), in their study of Singaporean banking, argued and empirically demonstrated that cumulative name changes with respect to the category label diluted the distinct identity of the emerging population of cooperative banks. The paper of Dobrev et al. (2006) thus utilized names to infer population-level processes rather than only studying firm-level effects (as for instance done by Ingram, 1996; see also Chuang and Baum, 2003). Labels and names are especially important in domains in which clear-cut categories are lacking and in which audiences have to make sense of the emerging organizational landscape and rely strongly on such basic markers (Dobrev et al., 2006; Hannan et al., 2007). In the context of our study, labels such as ‘airline’, such as in the name Culliford Airlines, are important elements of names as they represent implicit claims of membership in the nascent airline domain by founders and business owners. The early UK airline industry displayed substantial variation in corporate names which included labels such as Airways, Flying Boats, Air Ferries, and Aerial Transport. This variation serves as a basis for our contrast measure. The variety of industry labels used by airlines paralleled the diverse terminology employed in the media (for a discussion see also the section containing the sensitivity analyses). For example, in a 25 September 1920 article, the ‘aeronautical correspondent’ of the London-based newspaper The Times described the new passenger airlines that had emerged in the preceding year as ‘air expresses’, ‘airways’, as well as ‘aerial transport companies’. We assume that similar patterns can be found among alternative audiences, such as investors and consumers, i.e., that diverse types of audience members are comparable in the ways they perceive, pay attention to, and make sense of labels.

For testing Hypothesis 2, each organization’s grade of membership (GoM), was calculated by first computing the proportion of companies using a particular label in their names in a given year. Systematic analysis of company names yielded a set of seven non-unique labels.² Together, these seven labels accounted for the 93.5% of all airline-year observations. These labels were (1) air (e.g., Air Europe), (2) airways (e.g., Chieftain Airways), (3) airlines (e.g., Culliford Airlines), (4) aero/aerial (e.g., Aerial Enterprises), (5) fly/flight (e.g., Go Fly), (6) avia/aviation (e.g., Giro Aviation), and (7) sky (e.g., Skytravel).³ The similarity of airline i with label ε to other airlines’ labels, was then calculated as:

L S_{i t} = \frac{(n_{i t} (ε) - 1) + \sum (α_{i ε} * n_{t} (ε))}{N_{t} - 1}

where n(ε) is the total number of airlines using a particular label ε in year t, and N is the total number of airlines in the population in that year (cf. Hannan et al., 2007). One was subtracted from both n(ε) and N to exclude the airline i from the count. Recognizing that some labels (such as ‘airlines’ and ‘airways’) are more similar to each other than others (such as ‘airlines’ and ‘sky’), a weight, α_iε, was included representing the semantic relatedness of the label used by airline i and any other label ε, on a scale from zero to one (Pederson et al., 2004). The weights were computed by means of a ‘gloss vector’ measure based on the second-order co-occurrence vectors from their glosses or WordNet (Miller, 2005) definitions of the concepts. The relatedness of two concepts was then defined as the cosine of their gloss vectors⁴ (Patwardhan and Pedersen, 2006). The output of this procedure has been shown to outperform other measures of semantic relatedness in terms of representing actual human perception. It ensures that an organization with the label ‘airline’ operating in a population where the consensus revolves around the label ‘airways’ would have a higher label similarity than one with the label ‘jet’ in the same population. In cases where an airline was the only one using a specific name label, its label similarity would be zero. For the airlines that did not use any of the seven labels studied, their label similarity was set equal to zero. In a hypothetical case where all airlines employed the same label, each would have a score of one.

Contrast then was measured as the average GoM among the airlines operating in a particular year (Hannan et al., 2007). This variable is employed to test Hypothesis 1. Figure 3 displays the evolution of the contrast measure during our window of observation.

Figure 3.

Contrast of UK passenger airlines, 1919–64

To test Hypotheses 3, a density measure was created (Hannan and Carroll, 1992), defined as the total number of airlines in operation in the UK in a given year, and was interacted with the contrast variable. Density was lagged by one year to address reverse causation.

Control variables

The size of an airline in a given year was represented by the aggregate seating capacity of its fleet in that year. Organization size was thus treated as a time-varying covariate that changed annually with the fleet’s composition. Calculating size in this way, however, created a measurement problem for those firms which had aircraft designed for cargo activity. For these aircraft, the seating capacity of an equivalent passenger aircraft type was used. Not taking into account these cargo aircraft would underestimate the true size of the company, thus, a dummy variable was added which was set equal to one for those firms that were partially involved in cargo, zero otherwise. Organization size was log-transformed to normalize its distribution. In addition to the absolute measure of size, models were evaluated in which size relative to the largest firm in the industry was used (Hannan et al., 1998). It has been argued that relative size may better capture a firm’s position in evolving size distributions (Carroll and Hannan, 2000). This variable showed significant predictive power, but models based on absolute size generally fitted better. Given that the two measures showed a significant correlation of 0.64 (p < .001), only absolute size was eventually used. Since the data did not allow measurement of competition over specific routes (Gimeno and Woo, 1996; Seidel, 1997), competitive pressure was controlled for with one alternative measure, aside from the population density measure already discussed. This alternative measure represents the number of airlines competing at each firm’s home base airport (airport density), reflecting local competition for critical resources such as flight crew and airport slots.

Moreover, flying international routes is likely to expose the focal airline to competition with foreign airlines. To control for this, a dummy variable marking airlines that were active internationally was included. We controlled for name changes (Dobrev et al., 2006; Glynn and Abzug, 2002), by adding a firm-level indicator variable that marks any change that was made to a firm’s name in the prior year. If an organization originated from outside the airline industry (de alio) rather than having been founded as an airline from the start (de novo), this was also controlled for (one for the de alio firms, zero otherwise). Among de alio organizations common origins were shipping and railroad companies, as well as travel organizations. A final firm-level variable indicated the airlines which were government owned (one if the government had a stake, zero otherwise).

Following work that has analyzed the impact of airline accidents (Haunschild and Sullivan, 2002), we controlled for the total number of airline accidents taking place on British soil in each year. The size of the British population was included to proxy for the market’s munificence and its carrying capacity. Oil prices (in pounds and corrected for inflation) were also included as a control variable, because higher oil prices might be expected to affect each airline’s performance both directly by adding to operating costs, and indirectly by reducing consumers’ disposable income and thus their demand for air travel. As a proxy for the purchasing power in the passenger market, we added a measure representing GDP, and in order to control for population-level structuration we added population age, defined as the number of years passed since 1919 (both logged to normalize their distribution). Finally, the effect of the Second World War, WWII, was controlled for by creating a period dummy coded as 1 during the period 1940–5, zero otherwise. Table 1 shows descriptive statistics for all the variables and their pair-wise correlations.

Table 1.

Descriptive statistics

	Mean	SD	Min	Max	1	2	3	4	5	6	7	8	9	10	11	12	13
1. Government	0.04	0.19	0	1
2. Cargo	0.38	0.49	0	1	−0.15
3. Accidents	4.63	2.53	0	10	−0.07	0.05
4. De alio	0.19	0.39	0	1	−0.09	0.06	0.04
5. Name change	0.01	0.11	0	1	−0.02	0.01	−0.07	−0.04
6. International routes	0.68	0.47	0	1	0.13	0.29	−0.03	0.03	−0.02
7. Airport density	2.56	3.37	0	18	−0.04	0.19	0.10	−0.05	0.00	0.07
8. WWII	0.07	0.26	0	1	0.03	−0.02	0.01	−0.03	−0.03	−0.04	−0.1
9. Human population	3.91	0.04	3.78	3.98	−0.13	0.19	0.08	0.07	0.05	0.11	0.11	−0.2
10. Oil prices	14.83	2.6	6.11	27.89	−0.02	0.08	0.21	0.06	0.00	0.10	0.14	−0.33	0.53
11. Firm size	3.37	1.72	−0.56	9.23	0.40	0.16	0.00	0.1	0.03	0.35	0.12	−0.08	0.24	0.16
12. Density	64.45	20.87	0	102	−0.20	0.14	0.40	0.07	−0.02	−0.03	0.24	−0.34	0.55	0.35	0.06
13. GoM	0.55	0.21	0	0.76	−0.10	0.04	−0.05	0.03	0.04	0.02	0.10	0.14	0.15	−0.08	0.03	−0.05
14. Contrast	0.55	0.07	0.21	0.66	−0.07	0.08	−0.11	−0.03	0.05	0.03	−0.09	0.38	0.40	−0.23	0.10	−0.14	0.37
15. Population age	30.91	9.13	0	45	−0.03	0.19	0.30	0.07	0.05	0.11	0.11	-0.23	0.82	0.47	0.46	0.30	0.18	0.48
16. GDP	12.63	0.19	12.01	12.96	−0.02	0.18	0.25	0.08	0.07	0.13	0.07	-0.05	0.79	0.43	0.51	0.16	0.21	0.56	0.79

Modeling framework

Event-history analysis (Tuma and Hannan, 1984) was applied to estimate the mortality hazard of each airline. The mortality hazard was formally defined as:

µ (u) = \lim_{Δ u \to 0} \frac{\Pr (f a i l u r e (u + Δ u) | n o f a i l u r e (u))}{Δ u}

which reads as the likelihood that an airline fails and exits from the market between its age u and u+Δu, provided that it did not exit at or prior to u. Of course, an airline can exit from a market in different ways, some of which may not necessarily be classified as failure (e.g., by M&A or nationalization). In line with previous research (Carroll and Hannan, 2000: 45), only pure disbanding was treated as a failure event. We used an exponential specification (Tuma and Hannan, 1984) in modeling the failure rate, which is a product of an unspecified constant baseline hazard, µ₀(u), and a vector x specifying the influences of covariates.

µ (u) = µ_{0} (u) \exp (β^{'} x)

Results not reported here (but available upon request) indicate that qualitatively similar results were obtained when employing piecewise exponential models.

Results

Table 2 presents the maximum likelihood estimates of the proportional hazard models of the mortality of 298 British airlines during the period 1919–64. Model 1 contains control variables only. Model 2 examines the impact of contrast. Model 3 adds the interaction term of contrast and GoM. Models 4 then adds the interaction between density and contrast.

Table 2.

Exponential models of failure rates of UK airlines. 1919–64

	Model 1	Model 2	Model 3	Model 4
Cargo	−0.18	−0.17	−0.16	−0.17
Cargo	(0.20)	(0.20)	(0.20)	(0.20)
Accidents	−0.03	−0.03	−0.05	−0.02
Accidents	(0.04)	(0.04)	(0.04)	(0.04)
De alio	−0.15	−0.11	−0.17	−0.13
De alio	(0.26)	(0.25)	(0.26)	(0.25)
Name change	1.32**	1.33**	1.34**	1.31**
Name change	(0.46)	(0.48)	(0.45)	(0.47)
International	−0.59**	−0.56**	−0.58**	−0.58**
International	(0.19)	(0.19)	(0.19)	(0.19)
Airport density	0.04*	0.04*	0.04*	0.04*
Airport density	(0.02)	(0.02)	(0.02)	(0.02)
WWII	0.09	0.10	0.20	0.61
WWII	(0.69)	(0.73)	(0.76)	(0.80)
Human population	5.98	6.40	6.57	10.03*
Human population	(4.01)	(4.80)	(5.00)	(5.02)
GDP	−14.39**	−15.24**	−15.20**	−15.77**
GDP	(3.43)	(3.43)	(3.47)	(3.40)
Government	−15.00**	−14.48**	−14.58**	−14.31**
Government	(0.61)	(0.64)	(0.64)	(0.64)
Oil prices	0.07	0.11*	0.10	0.04
Oil prices	(0.04)	(0.05)	(0.05)	(0.05)
Firm size	−0.17**	−0.19**	−0.20**	−0.19**
Firm size	(0.06)	(0.06)	(0.06)	(0.06)
Population age	0.02	0.24	0.25	0.38
Population age	(0.06)	(0.23)	(0.23)	(0.27)
Density	−0.00	−0.01*	−0.01*	−0.04**
Density	(0.18)	(0.00)	(0.00)	(0.02)
GoM	−0.95**	−0.87**	−1.06*	−0.85*
GoM	(0.32)	(0.33)	(0.46)	(0.33)
Contrast		−5.33*	−5.64*	−16.21**
Contrast		(2.69)	(2.80)	(5.47)
Contrast x GoM			−3.32
Contrast x GoM			(4.91)
Contrast x Density				−0.22*
Contrast x Density				(0.10)
Constant	−99.30**	−95.96**	−93.63**	−86.59**
Pseudo-log likelihood	−303.43	−301.46	−301.29	−298.91
Observations	2293	2293	2293	2293

Note: Robust standard errors in parentheses.

significant at 5%; ** significant at 1%.

The estimates of the control variables turned out as expected: increases in GDP, organizational size, governmental subsidies, and involvement in international routes all contributed to lower the hazard of mortality of UK airlines. Two other results concerning the control variables are worth mentioning: first, name change is positively associated to organizational mortality; second, the higher the grade of membership – i.e., the greater the naming conformity of the focal organization with respect to the general trend of the industry (i.e., GoM) – the lower the hazard of organizational mortality. We interpret these findings as initial evidence that naming patterns critically affected organizational survival in our context.

Whereas Model 1 indicates an insignificant impact of density throughout our observation window, Model 2 reveals that when contrast is added the coefficient of density turns statistically significant. We interpret this finding as suggesting that, indeed, the effect of density dependent legitimation may vary across the values of category contrast. As expected and in line with Hypothesis 1, increasing contrast lowered the mortality rate of UK airlines – net of organizational density. It is worth noticing that Model 2 exhibits a significantly better fit to the data than Model 1 (likelihood ratio χ² = 3.94; Δd.f. = 1; p < .05).

Against the arguments advanced by Hypothesis 2, firm-level GoM did not interact with population-level contrast. This finding however appears in line with that of Kuilman and Li (2009), who also were not able to find support for this thesis, and seems to run against the result obtained by Bogaert et al. (2010): during a legitimacy process the benefits of contrast did not significantly vary across high and low GoM organizations – at least as far as naming patterns are concerned. Contrast therefore did not appear to be a source from which peripheral organizations in particular may derive legitimacy.

Model 4 adds the interaction between density and contrast. The results obtained indicate that the negative relationship between contrast and mortality is reinforced at higher values of density, as we predicted. The model also improved significantly over Model 2 (likelihood ratio χ² = 5.08; Δd.f. = 1; p < .05). Figure 4 provides a visual plot of the interaction between density and contrast and shows that in presence of high contrast, increases in density lead to proportionally larger declines in mortality rates. We interpret this finding as indicating that the simultaneous existence of high density and elevated category contrast provides the strongest results in terms of cognitive legitimacy and, therefore, of organizational survival.

Figure 4.

Convergence between the label usage among UK passenger airlines and label usage in The Times (London), 1919–64

Sensitivity analyses

Several additional analyses were run to clarify the mechanism through which category contrast influenced organizational survival. As our theory makes reference to audience members and to the effects of labels on easing their cognition and the identification of a similarity-based cluster among airline producers, analyses not reported here investigated whether contrast exhibited any impact on the appeal of airliners in the eyes of audience members. The results obtained (available upon request) indicate that category contrast positively affected the growth rates of the yearly number of airline passengers as well as the annual number of organizational entries observed in our sample.

We also explored the extent to which our measure of contrast (based on the labels incorporated in organizational names) converges with the usage of the labels by external audiences. For this purpose, we constructed a measure of the degree to which labels in corporate names matched the labels used in newspaper The Times (London).⁵ Figure 5, which plots this measure against historical time, illustrates a low degree of convergence early on but an increasing convergence as time passes. Taken together, we interpret this trend of results as providing further evidence concerning the role of category contrast for the emergence of the population of UK airlines and in sustaining the appeal of their offerings.

Figure 5.

Interaction between density and contrast

Discussion

A significant body of research has contributed to our understanding of cognitive legitimacy and taken-for-grantedness of organizational categories. Institutional theorists for instance examined how taken-for-granted categories emerge as a result of processes of institutional isomorphism (DiMaggio and Powell, 1983) or through new institutional logics (Haveman and Rao, 1997), possibly driven by the actions of institutional entrepreneurs (Maguire et al., 2004) who defy competing logics (Marquis and Lounsbury, 2007). Ecologists, conversely, have studied cognitive legitimation as related to organizational density – i.e., the number of organizations within a population. The present article builds upon recent theoretical developments within ecology (Hannan et al., 2007) to study the effects of category contrast on organizational survival. The concept of category contrast appears especially interesting for the study of cognitive legitimacy as it is sensitive to the processes of convergence and perceived homogeneity described by several institutionalist scholars. Our measurement of contrast with respect to company names meshes also well with the claim that processes of institutionalization hinge upon the development of a shared label (Glynn and Abzug, 2002; Hannan et al., 2007; Ingram, 1996; Lee, 2001; Meyer and Rowan, 1977).

To strengthen our appreciation of cognitive legitimacy, we juxtaposed the effects of contrast to those triggered by organizational density and elaborated on the relationship between these two concepts. Our empirical tests – carried out in the population of UK airlines during the period 1919–64 – provided substantial support to the theoretical arguments advanced. The exception is the lack of support for the argument that low GoM firms would be more likely to benefit from the legitimacy spillovers generated by greater levels of contrast, a finding however aligned with that of Kuilman and Li (2009). Our speculative interpretation here is that having greater levels of contrast makes categorical boundaries sharper and, while enhancing a category’s overall legitimacy, it simultaneously makes clearer which organizations have lower levels of fit with the category and therefore should actually not be considered by audiences. As a result, legitimacy transfer to peripheral organizations may be hampered with greater levels of contrast.

The present article contributes to the organization ecology literature for the following reasons. First, it provides further evidence about the role of contrast in the emergence of new industries and on the spillovers of legitimacy that it triggers. While a limited set of papers have already explored this topic (e.g., Bogaert et al., 2010; Kuilman and Li, 2009), the findings obtained so far beg for replication. More importantly, none of the existing papers has attempted to approach the concept of contrast with respect to naming and labels – an element essential to the theoretical formulation of Hannan et al. (2007). Second, our article provides new evidence about the relationship between density and contrast. While the old tradition in ecology emphasized density as a key trigger of cognitive legitimacy and taken-for-grantedness (see Carroll and Hannan, 2000 for a review), recent developments within this tradition emphasize contrast and perceived homogeneity as fundamental to the emergence and legitimacy of a new form. Unfortunately, however, the role of density in this process and especially its relationship with contrast remains scantly understood. The results of our article provide support to the findings of Bogaert et al. (2010) who pointed to the simultaneous presence of density and contrast as necessary to taken-for-grantedness. A further open question however pertains to the role of density in the earliest stages of form emergence. Recent evidence seems to point to a competitive effect of density under this scenario – albeit in the context of existing industries in which violations by comparison may be at work, see e.g., Dobrev et al. (2006) and Delmestri and Wezel (2011). Separate analyses of our data zooming in on the earliest stages of form emergence revealed a positive and statistically significant main effect of density on organizational mortality. While it may be difficult to justify the existence of competition among a set of organizations not yet coalesced together, the struggles faced by organizations in a legitimacy vacuum may potentially explain this result. To address this puzzle, more research on the early stages of industry emergence and on processes of legitimation of new organizational categories is needed (for a similar claim see Hsu and Hannan, 2005; Khaire and Wadhwani, 2010; Navis and Glynn, 2010; Wry et al., 2011).

While we positioned this article as a contribution to ecological research, our arguments are clearly consonant with those advanced by new-institutionalist researchers. Indeed, according to this literature, the emergence of new organizational fields is driven by the collective action and institutional entrepreneurs who create policies and standards aligned with their interests (Garud et al., 2002) in conjunction with the evolution of meaning systems and discourses (Lawrence and Philips, 2004; Maguire et al., 2004). The legitimacy struggles encountered by a diffusing organizational form or organizational practice are debated (e.g., Ansari et al., 2010; Marquis and Lounsbury, 2007; Zelner et al., 2009) and the depiction of legitimacy as unfolding continuously appears questioned in this literature as well (for a review see Deephouse and Suchman, 2008). The debate on legitimation processes as either continuous or dichotomous appears especially relevant upon acknowledging the role of categorization for the achievement of taken-for-grantedness (for a discussion see also Hampton, 1998). Does contrast unfold continuously or once a minimum threshold is met do further increases negligibly contribute to cognitive legitimacy? Is a minimum level of contrast (i.e., a threshold) needed to see the effects of organizational density kicking in? Needless to say, filling the divide between ecologists and new institutionalists falls beyond the scope of the present article. But the questions posed here seem to confirm that the ‘strongly defended boundary between institutional and ecological views of organization does not demark an intellectually sensible divide’ (Carroll and Hannan, 2000: 81) – and even more so nowadays that ecologists shifted to a socially constructed perspective on industry emergence (Hannan et al., 2007: Ch. 2). As the present article hopefully witnesses, the concept of contrast represents a possible platform for initiating a constructive dialogue among these research camps.

Limitations and future research

Notwithstanding our efforts at uncovering the mechanisms at work during population emergence, issues of generalizability may be at stake. In particular, our conclusions may be specific to the industry selected (i.e., a clear discontinuity with previous organizational categories), to the measurement strategy (i.e., contrast as related to consensus in organizational names), and to the period of time observed (i.e., the early years of industry evolution). For instance, the performance consequences attached to labeling may vary by industry and labels themselves may play a different role in mature sectors for instance. In many mature sectors organizational labels serve to differentiate companies and market them in a crowded competitive arena. Thus, our narrative may be especially true for young and small firms lacking reputation in the market and heavily relying on legitimation spillovers. Future research should investigate these matters further. As for the specific industries in which naming patterns may matter the most, we speculate that emerging and mediated markets (e.g., those in which the sense-making of customers is difficult and relies on intermediaries and social cues) may be good candidates. More evidence should be collected to prove the robustness of our theoretical statements in different empirical settings.

The present study relied on the convergence among the labels adopted by organizations to draw inferences about category contrast and on the cognitive legitimacy of an organizational population. Yet, to fully appreciate the importance of the convergence in labels for cognitive legitimacy it would be interesting to further explore the interplay between the labels employed by audience members and those put forward by entrepreneurs and business owners. For example, is the usage of a particular label by outsiders (e.g., stakeholders) influenced by the labels adopted by entrepreneurs? Or, conversely, how much of the contrast we studied here is the by-product of the expectations of external constituents? More refined data are needed to address these questions and to understand how much external constituents influence labeling decisions. Related to this, it would be interesting to explore more in-depth where convergence in labels and, thus, increasing category contrast comes from in the first place (Glynn and Abzug, 1998). Models in which labeling consensus is the outcome of isomorphism and competitive differentiation (Glynn and Abzug, 1998), and in which the attributes of the adopters (e.g., status, prominence, size) are taken into account, may be explored to further understand the origins of labeling conventions.

Footnotes

Acknowledgements

We are grateful to Lyda Bigelow, Tai-Young Kim, László Pólos, Yi Tang, and seminar participants at the University of Lugano, at the Organizational Ecology conference, and at the Academy of Management meetings for providing useful comments on earlier drafts. Erik Huizen, Rajeev Ranjan, and Wang Wei provided great research assistance. Joel Baum, editor of Strategic Organization, and three anonymous reviewers are gratefully acknowledged for their constructive comments and suggestions on earlier drafts of this article.

Funding

This research was supported by the Hong Kong Research Grants Council through a Direct Allocation Grant (DAG06/07.BM11) and by the Erasmus Research Institute of Management. Filippo Carlo Wezel acknowledges the financial support received from the Swiss National Science Foundation, Research Grant 100018_135313.

Notes

Author biographies

Jeroen G. Kuilman is an Associate Professor at the Department of Organization and Strategy at Tilburg University (the Netherlands). He previously worked at the Hong Kong University of Science and Technology. His research focuses on ecological dynamics prior to market entry, the evolution of industries, and categorization at various levels of analysis. His work has been published in Academy of Management Journal, Academy of Management Review, Industrial and Corporate Change, and Organization Science. Address: Jeroen G. Kuilman, Department of Organization and Strategy, Tilburg University, PO Box 90135, 5000 LE, Tilburg, The Netherlands. [email: J.G.Kuilman@uvt.nl]

Filippo Carlo Wezel is a Professor of Organization Theory at the Faculty of Economics of the University of Lugano (Switzerland) and a senior Extramural Fellow of CentER, Tilburg University (the Netherlands). He previously held appointments at Groningen University and at Tilburg University. His research focuses on organizational diversity, inter-firm mobility, and industry emergence. His work has been published in Academy of Management Journal, Organization Science, Journal of International Business Studies, Organization Studies, Research in the Sociology of Organizations, and Journal of Business Venturing. Address: University of Lugano, Faculty of Economics, Via Buffi, 13 – 6900, Lugano, Switzerland. [email: wezelf@usi.ch]

References

Ansari

Fiss

P. C.

Zajac

E. J.

(2010) ‘Made to Fit: How Practices Vary as They Diffuse’, Academy of Management Review 35: 67–92.

Baum

J. A. C.

(1999) ‘The Rise of Chain Nursing Homes in Ontario, 1971–1996’, Social Forces 78(2): 543–84.

Baum

J. A. C.

Korn

H. J.

(1996) ‘Competitive Dynamics of Interfirm Rivalry’, Academy of Management Journal 39: 255–91.

Baum

J. A. C.

Oliver

(1992) ‘Institutional Embeddedness and the Dynamics of Organizational Populations’, American Sociological Review 57: 540–59.

Baum

J. A. C.

Powell

W.W.

(1995) ‘Cultivating an Institutional Ecology of Organizations: Comment on Hannan, Carroll, Dundon, and Torres’, American Sociological Review 60: 529–38.

Bogaert

Boone

Carroll

G. R.

(2010) ‘Organizational Form Emergence and Competing Professional Schemata of Dutch Accounting, 1884–1939’, Research in the Sociology of Organizations 31: 115–50.

Carroll

G. R.

Hannan

M. T.

(2000) The Demography of Corporations and Industries. Princeton, NJ: Princeton University Press.

Cattani

Ferriani

Negro

Perretti

(2008) ‘The Structure of Consensus: Network Ties, Legitimation, and Exit Rates of U.S. Feature Film Producer Organizations’, Administrative Science Quarterly 53(1): 145–82.

Chuang

Y. T.

Baum

J. A. C.

(2003) ‘It’s All in the Name: Failure-induced Learning by Multiunit Chains’, Administrative Science Quarterly 48(1): 33–59.

10.

Delmestri

Wezel

F. C.

(2011) ‘Breaking the Wave: The Contested Legitimation of an Alien Organizational Form’, Journal of International Business Studies 42: 828–52.

11.

Deephouse

D. L.

Suchman

M. C.

(2008) ‘Legitimacy in Organizational Institutionalism’, in Greenwood

Oliver

Sahlin

Suddaby

(eds) The Sage Handbook of Organizational Institutionalism, pp. 49–77. Thousand Oaks, CA: Sage.

12.

DiMaggio

Powell

(1983) ‘The Iron Cage Revisited: Institutional Isomorphism and Collective Rationality in Organizational Fields’, American Sociological Review 48: 147–60.

13.

DiMaggio

Powell

(1991) ‘Introduction’, in Powell

DiMaggio

(eds) The New Institutionalism in Organizational Analysis, pp. 1–38. Chicago, IL: University of Chicago Press.

14.

Dobrev

S. D.

Gotsopoulos

(2010) ‘Legitimacy Vacuum, Structural Imprinting, and the First-mover Disadvantage’, Academy of Management Journal 53: 1153–74.

15.

Dobrev

S. D.

Ozdemir

S. Z.

Teo

A. C. Y.

(2006) ‘The Ecological Interdependence of Emergent and Established Organizational Populations: Legitimacy Transfer, Violation by Comparison, and Unstable Identities’, Organization Science 17: 577–97.

16.

Durkheim

É.

Mauss

(1903) Primitive Classification, trans. Needham

Chicago, IL: The University of Chicago Press.

17.

Dyos

H. J.

Aldcroft

D. H.

(1969) British Transport: An Economic Survey from the Seventeenth Century to the Twentieth. Leicester: Leicester University Press.

18.

Fearon

(1985) The growth of aviation in Britain. Britain Journal of Contemporary History 20: 21–40.

19.

Garud

Jain

Kumaraswamy

(2002) ‘Institutional Entrepreneurship in the Sponsorship of Common Technological Standards: The Case of Sun Microsystems and Java’, Academy Management Journal 45: 196–214.

20.

Gimeno

Woo

C. Y.

(1996) ‘Hyper Competition in a Multimarket Environment: The Role of Strategic Similarity and Multimarket Contact in Competitive De-escalation’, Organization Science 7: 322–41

21.

Glynn

M. A.

Abzug

(1998) ‘Isomorphism and Competitive Differentiation in the Organizational Name Game’, in Baum

J. A. C.

Rao

(eds) Advances in Strategic Management, pp. 105–28. Stamford, CT: JAI Press.

22.

Glynn

M. A.

Abzug

(2002) ‘Institutionalizing Identity: Symbolic Isomorphism and Organizational Names’, Academy of Management Journal 45(10): 267–80.

23.

Gort

Klepper

(1982) ‘Time Paths in the Diffusion of Product Innovations’, Economic Journal 92: 630–53.

24.

Halford-MacLeod

(2007) Britain’s Airlines – Volume Two: 1951–1964. Chalford: Tempus Publishing.

25.

Hampton

J. A.

(1998) ‘Similarity-based Categorization and Fuzziness of Natural Categories’, Cognition 65: 137–65.

26.

Hannan

M. T.

(1986) ‘Uncertainty, Diversity, and Organizational Change’, in Smelser

Gerstein

(eds) Social Structures and Human Lives, pp. 73–94. Washington, DC: National Academy Press.

27.

Hannan

M. T.

Carroll

G. R.

(1992) Dynamics of Organizational Populations: Density, Legitimation, and Competition. New York: Oxford University Press.

28.

Hannan

M. T.

Freeman

J. H.

(1977) ‘The Population Ecology of Organizations’, American Journal of Sociology 82: 929–64.

29.

Hannan

M. T.

Freeman

J. H.

(1984) ‘Structural Inertia and Organizational Change’, American Sociological Review 49: 149–64.

30.

Hannan

M. T.

Carroll

G. R.

Dobrev

S. D.

Han

(1998) ‘Organizational Mortality in European and American Automobile Industries Part I: Revisiting the Effects of Age and Size’, European Sociological Review 14(3): 279–302.

31.

Hannan

M. T.

Pólos

Carroll

G. R.

(2007) Logics of Organization Theory: Audiences, Codes, and Ecologies. Princeton, NJ: Princeton University Press.

32.

Haunschild

Sullivan

B. N.

(2002) ‘Learning from Complexity: Effects of Accident/Incident Heterogeneity on Airline Learning’, Administrative Science Quarterly 47: 609–43.

33.

Haveman

H. A.

Rao

(1997) ‘Structuring a Theory of Moral Sentiments: Institutional and Organizational Coevolution in the Early Thrift Industry’, American Journal of Sociology 102(6): 1606–51.

34.

Hsu

Hannan

M. T.

(2005) ‘Identities, Genres and Organizational Forms’, Organization Science 7: 211–20.

35.

Hsu

Hannan

M. T.

Koçak

Ö.

(2009) ‘Multiple Category Memberships in Markets: An Integrative Theory and Two Empirical Tests’, American Sociological Review 74: 150–69.

36.

Ingram

(1996) ‘Organizational Form as a Solution to the Problem of Credible Commitment: The Evolution of Naming Strategies among U.S. Hotel Chains, 1890–1980’, Strategic Management Journal 17: 85–98.

37.

Kelly

Amburgey

T. L.

(1991) ‘Organizational Inertia and Momentum: A Dynamic Model of Strategic Change’, Academy of Management Journal 34: 591–612.

38.

Kennedy

M. T.

(2008) ‘Getting Counted: Markets, Media and Reality’, American Sociological Review 73: 270–95.

39.

Khaire

Wadhwani

R. D.

(2010) ‘Changing Landscapes: The Construction of Meaning and Value in a New Market Category – Modern Indian Art’, Academy of Management Journal 53(6): 1281–304.

40.

Klepper

Graddy

(1990) ‘The Evolution of New Industries and the Determinants of Market Structure’, Rand Journal of Economics 21: 27–44.

41.

Kovács

Hannan

M. T.

(2010) ‘The Consequences of Category Spanning Depend on Contrast’, Research in the Sociology of Organizations 31: 175–201.

42.

Kuilman

J. G.

J. T.

(2009) ‘Grades of Membership and Legitimacy Spillovers: Foreign Banks in Shanghai, 1847–1935’, Academy of Management Journal 52: 229–45.

43.

Lawrence

T. B.

Philips

(2004) ‘From Moby Dick to Free Willy: Macro-cultural Discourse and Institutional Entrepreneurship in Emerging Institutional Fields’, Organization 11: 689–711.

44.

Lee

P. M.

(2001) ‘What’s in a Name.com? The Effect of “.com” Name Changes on Stock Prices and Trading Activity’, Strategic Management Journal 22: 793–804.

45.

McKendrick

D. G.

Jaffee

Carroll

G. R.

Khessina

O. M.

(2003) ‘In the Bud? Disk Array Producers as a (Possibly) Emergent Organizational Form’, Administrative Science Quarterly 48: 60–93.

46.

Maguire

Hardy

Lawrence

T. B.

(2004) ‘Institutional Entrepreneurship in Emerging Fields: HIV/AIDS Treatment Advocacy in Canada’, Academy of Management Journal 47: 657–79.

47.

Marquis

Lounsbury

(2007) ‘Vive la Résistance: Competing Logics in the Consolidation of Community Banking’, Academy of Management Journal 50: 799–820.

48.

Merton Jones

A. C.

(1976) Independent Airlines Since 1946. Liverpool: Merseyside Aviation Society.

49.

Meyer

J. W.

Rowan

(1977) ‘Institutional Organizations: Formal Structure as Myth and Ceremony’, American Journal of Sociology 83: 340–63.

50.

Miller

(2005) WordNet 2.1. Available at: wordnet.princeton.edu/ (accessed 1 August 2008).

51.

Mitchell

B. R.

(1988) British Historical Statistics. Cambridge: Cambridge University Press.

52.

Mohr

J. W.

Guerra-Pearson

(2010) ‘The Duality of Niche and Form: The Differentiation of Institutional Space in New York City, 1888–1917’, Research in the Sociology of Organizations 31: 321–68.

53.

Navis

Glynn

M. A.

(2010) ‘How New Market Categories Emerge: Temporal Dynamics of Legitimacy, Identity, and Entrepreneurship in Satellite Radio, 1990–2005’, Administrative Science Quarterly 55: 439–71.

54.

Negro

Hannan

M. T.

Rao

(2010) ‘Categorical Contrast and Audience Appeal: Niche Width and Critical Success in Winemaking’, Industrial and Corporate Change 19: 1397–425.

55.

Negro

Hannan

M. T.

Rao

(2011) ‘Category Reinterpretation and Defection: Modernism and Tradition in Italian Winemaking’, Organization Science 22(6): 1449–63.

56.

Nelson

R. R.

Winter

(1982) An Evolutionary Theory of Economic Change. Cambridge, MA: Harvard University Press.

57.

Palmer

S.W.

(2006) Dictatorship of the Air: Aviation Culture and the Fate of Modern Russia. Cambridge: Cambridge University Press.

58.

Park

D. Y

Podolny

(2000) ‘The Competitive Dynamics of Status and Niche Width: US Investment Banking, 1920–1949’, Industrial and Corporate Change 9: 377–414.

59.

Patwardhan

Pedersen

(2006) ‘Using WordNet Based Context Vectors to Estimate the Semantic Relatedness of Concepts’. Paper presented at Proceedings of the EACL 2006 Workshop Making Sense of Sense – Bringing Computational Linguistics and Psycholinguistics Together, Trento, Italy.

60.

Pederson

Patwardhan

Michelizzi

(2004) ‘Wordnet::Similarity – Measuring the Relatedness of Concepts’. Paper presented at Proceedings of the Nineteenth National Conference on Artificial Intelligence, AAAI-04.

61.

Pólos

Hannan

M. T.

Carroll

G. R.

(2002) ‘Foundations of a Theory of Social Forms’, Industrial and Corporate Change 11: 85–115.

62.

Rosch

E. H.

Mervis

C. B.

Gray

W. D.

. (1976) ‘Basic Objects in Natural Categories’, Cognitive Psychology 8: 382–439.

63.

Ruef

Patterson

(2009) ‘Credit and Classification: The Impact of Industry Boundaries in 19th Century America’, Administrative Science Quarterly 54: 486–520.

64.

Seidel

M. D. L.

(1997) ‘Competitive Realignment in the Airline Industry: A Dynamic Analysis of Generalist and Specialist Organizations under Different Route Network Structures’. Unpublished doctoral dissertation, University of California at Berkeley.

65.

Smith

(2002) The Airline Encyclopedia, 1909–2000. Lanham, MD: Scarecrow Press.

66.

Smith

K. G.

Grimm

C. M.

Gannon

M. J.

Chen

M.-J.

(1991) ‘Organizational Information Processing, Competitive Responses, and Performance in the U.S. Domestic Airline Industry’, Academy of Management Journal 34: 60–85.

67.

Suchman

(1995) ‘Managing Legitimacy: Strategic and Institutional Approaches’, Academy of Management Review 20: 571–610.

68.

Tuma

N. B.

Hannan

M. T.

(1984) Social Dynamics: Models and Methods. Orlando, FL: Academic Press.

69.

Tversky

(1977) ‘Features of Similarity’, Psychological Review 84: 327–52.

70.

Wry

Lounsbury

Glynn

M. A.

(2011) ‘Legitimating Nascent Collective Identities: Coordinating Cultural Entrepreneurship’, Organization Science 22: 449–63.

71.

Zelner

B. A.

Henisz

W. J.

Holburn

G. L. F.

(2009) ‘Contentious Implementation and Retrenchment in Neoliberal Policy Reform: The Global Electric Power Industry, 1989–2001’, Administrative Science Quarterly 54: 379–412.

72.

Zucker

L. G.

(1983) ‘Organizations as Institutions’, Research in the Sociology of Organizations 2: 1–47.

73.

Zuckerman

E. W.

(1999) ‘The Categorical Imperative: Securities Analysts and the Legitimacy Discount’, American Journal of Sociology 104: 1398–438.

Taking off: Category contrast and organizational mortality in the UK airline industry,1919–64

Abstract

Keywords

Introduction

Theoretical background

Organizational categories

Role of contrast in category emergence

Hypotheses

Data and methods

Data sources

The emergence of the UK airline industry

Independent variables

Control variables

Modeling framework

Results

Sensitivity analyses

Discussion

Limitations and future research

Footnotes

Acknowledgements

Funding

Notes

Author biographies

References