This file is also available in Adobe Acrobat PDF format
SELF-EMPLOYMENT DYNAMICS IN ADVANCED ECONOMIES
Walter Müller and Richard Arum
IN THE SECOND half of the twentieth century, social scientists typically viewed self-employment as an obsolete remnant of past forms of economic organization: its small-scale mode of production was expected to disappear under the dominating logic and competitive pressures of capital accumulation and mass production. Contrary to such predictions, recent developments have demonstrated that self-employment has resisted these pressures. Rather than diminishing its role in contemporary advanced societies, self-employment has grown in many national settings over recent decades. However, recent economic and social changes have not only reinvigorated self-employment, but also affected its character. Both developments require that social stratification researchers rethink their assumptions and reconceptualize their approach toward examining this activity.
Even if self-employment has merely persisted, it still represents a highly particular way to gain one's living in modern economic contexts. For most advanced societies, indeed, a cross-sectional picture immediately demonstrates that the self-employed constitute only a small fraction of the labor force. At any given time, only a minority fail to follow the normative practice of selling their labor capacity to an employer, but rather use their human, social, and economic capital for a business of their own and at their own risk. The first basic question then still is: What are the factors leading individuals into self-employment and then inducing them to remain active in these kinds of endeavors?
A self-employment rate of about 10 percent of the nonagricultural labor force is typically found in most advanced societies. However, in some societies this rate is much higher, in particular in the south of Europe and in Asian countries. In these countries, the rate of self-employment may be more than double the usual size, and in some national settings it accounts for 25 to 30 percent of the labor force. In earlier decades the high cross-national variation in the share of self-employment often was attributed to delays in economic development with convergence expected in the course of modernization. Today such explanations have lost much of their plausibility as even in economically highly advanced countries, such as Italy or Japan, self-employment has remained at very high levels. One should expect that societies with these widely varying levels of self-employment also vary in other ways, such as in the organization of work or even more generally in principles of social organization. How, then, can we account for the continued strong variation between countries in the presence of self-employment today? What are the implications of different economic and social contexts for the level and character of self-employment?
In the past, sociologists often assumed that self-employment to a large extent could be equated with the petty bourgeoisie, considered as a particular social class with its own specific social basis and principles of class formation and reproduction. Even though it has been characterized as the "uneasy stratum" (Bechhofer and Elliott 1981) and its internal heterogeneity has been widely acknowledged, it was assumed that this class mainly consisted of small proprietors, shopowners, skilled craftspeople, nonprofessional service providers, and others with similar working conditions who were running their own business. The bulk of the self-employed did indeed belong traditionally to these groups. Besides not working for wages or salaries, they shared a number of further common characteristics, notably a high level of intergenerational inheritance, a particular "moral economy" (Bechhofer and Elliott 1981:189), and specific political attitudes and orientations. Significant changes in the technology of production and the economic opportunity structure have recently occurred in many countries, including a revolution in computing and new communication technologies; expansion of female employment; the growth and persistence of mass unemployment; the decline in goods-producing industries and concentration of the distribution of goods in larger units; the rise of the service sector and increased professionalization; the growth of flexible production; and the spread of nontraditional work arrangements. For all these developments, arguments have been advanced that have repercussions on self-employment. What has been the impact of economic change on the character of self-employment, the composition of the group, its internal homogeneity, and its economic strength and stability?
Self-employment and small firm production have always constituted a context of work very different from employment in large, bureaucratically organized private companies or public administration. Specific social relations have usually been involved in the traditional petty bourgeois work unit: "The 'boss' works alongside his 'labour force.' Often the 'labour' is recruited from the networks of kinship, friendship and neighbourhood and thus there are 'non-contractual' elements in the relationship from the start." These authors also suggest that "in a society where large impersonal institutions mould much day-to-day living, this personalising of work relationships may be welcomed by both sides" (Bechhofer and Elliott 1981:194). Perhaps most significant in this traditional small-scale economic form of organization, and rather distinct in the modern context in which work tends to be separated and isolated from the family, has been the often intimate union of family and work in the self-employed venture. Indeed, self-employment has often been a family business, strongly dependent on the work of a married couple and their children. Women today, however, are more inclined to pursue their own independent careers and projects--even though to a varying extent in different societies--and children are certainly no longer easily kept as cheap family labor. Has thus the family-embeddedness of self-employment eroded? Has the separation between spheres of life also invaded self-employment and transformed its social character? Is the varying extent to which individuals in different societies have access to family resources precisely one of the factors that lies behind the uneven level and character of self-employed activity?
In research on social stratification, the family relatedness of self-employment has been particularly studied in the view of intergenerational inheritance of self-employment. It is regularly found that entry into self-employment is often a matter of directly inheriting a parental business, particularly so in agricultural self-employment where the family farm is passed from generation to generation. Also in this sense, self-employment has been characterized as a remnant of a social order, in which the intergenerational reproduction of social inequality still occurs primarily via direct inheritance, in contrast to access to positions of dependent work, where the transmission of advantage from generation to generation occurs mainly through socially unequal participation in education. Is contemporary self-employment still characterized by the traditional mode of socially inherited reproduction, or have more recent forms of self-employment moved in a direction whereby acquired human capital resources have gained a significant role compared with both transmission of material capital and the provision of social capital?
These are some of the core substantive questions that we pursue in this book with eleven national case studies on entry into and survival in self-employment. The national case studies provide both rich historical and contextual information to understand the specific nature of self-employment and the conditions of its recent development. The book concludes with a chapter presenting a comparative meta-analysis of cross-national results.
In this introductory chapter, we provide background information essential to understanding the topic, advance a theoretical framework for a cross-national study of self-employment dynamics, discuss the research design for the analysis, and present an outline of the structure of the book. In particular, we will discuss the reemergence of self-employment; the extent to which renewed analysis of self-employment is critical to understanding social inequality; the contemporary historical context of changes in work and employment relationships that have affected self-employment characteristics; a theoretical framework that emphasizes how resources, opportunities, and constraints structure individual-level decision making related to self-employment dynamics; and relevant dimensions of institutional and societal cross-national variation that affects self-employment outcomes.
Self-employment can no longer be dismissed as an economic activity on the verge of withering away in response to processes of capital accumulation or through competition with large firms. In the last quarter of the twentieth century, nonagricultural self-employment in most advanced economies reversed a historic pattern of decline and significantly increased its relative economic presence. In the United States, for example, the share of self-employment in the nonagricultural economy steadily declined from 12.0 percent in 1948 to 6.9 percent in 1970, before reversing course and increasing to 7.7 percent by 1993.1 In many other countries, such as England and Japan, a relatively similar historic pattern occurred. An analysis of twenty-eight Organization for Economic Cooperation and Development (OECD) countries found not only that "self-employment has tended to increase its share of non-agricultural civilian employment over the past three decades," but that this was particularly true from 1979 to the early 1990s when self-employment grew in these countries at an average annual rate of 2.3 percent per year (compared to 1.4 percent for civilian employment in general), and the overall share of self-employment in nonagricultural employment increased on average from 9.8 to 11.9 percent. In many postsocialist countries, changes were even more dramatic: self-employment--a practice that at one time had been virtually prohibited--not only reemerged in many of these countries but often was actively encouraged by new state policies promoting an economic transition to capitalism.2
While specific estimates differ with respect to sources of data (e.g., individuals responding to government officials tend to underreport self-employment activity), results are fairly robust in supporting the substantive conclusion that following unanticipated increases in self-employment rates, particularly from the late 1970s through the early 1990s, the final part of the twentieth century was then characterized by relative stability in levels of this activity. From this recent pattern, there is thus no reason to anticipate an ever-expanding renaissance of self-employment; however, neither is there any longer a reason to expect a withering away of this economic form. Self-employment's unexpected reemergence and continued persistence in advanced economies thus demands renewed examination. Self-employment has returned: social scientist must now, once again, reckon with this resilient social form in our midst.
An increase in self-employment not only affects how researchers and policymakers evaluate this activity and how much attention is focused on it but, as importantly, has significant implications for understanding the life experiences and economic position of individuals more generally. Many current forms of self-employment are significantly less stable than traditional petty bourgeois activity of the past. Increased occupational instability and increasing overall rates of self-employment are associated with spells of self-employment becoming more common over an individual's life-course. For example, in the United States (where self-employment occurs at relatively modest levels), more than 40 percent of men by their early fifties have engaged in self-employment at some point in their past.3 This figure is considerably higher than estimates from earlier periods, such as Lipset and Bendix's (1959: 102) estimate of "somewhere between 20 and 30 percent" of individuals. Given the high and apparently increasing propensity for self-employment activity to occur, the majority of individuals in advanced economies are related to, and likely to know well, individuals who have become self-employed.4
SOCIAL STRATIFICATION RESEARCH AND SELF-EMPLOYMENT
Studying self-employment's reemergence and internal transformations is particularly interesting to social stratification researchers in that these changes are closely related to several core intellectual concerns found within scholarship in this area. Social stratification researchers, for example, have long been interested in the character of employment activity in terms of inequality and stability, the social conditions associated with individuals engaging in nondependent enterprises (as this social status has historically had a disproportionate impact on political outcomes), and the extent to which an individual's educational achievement is capable of replacing direct ascriptive inheritance of occupational positions in modern societies.
Prior stratification researchers have paid varying attention to conceptualizing self-employment in occupational terms. Peter Blau and Otis Dudley Duncan's The American Occupational Structure (1967), for example, has three pages of indexed references to self-employment; David Featherman and Robert Hauser's Opportunity and Change (1978) has only two entries. Blau and Duncan identify self-employment as occurring in three of their seventeen occupational categories: self-employed professionals, proprietors, and farmers. Self-employed skilled and unskilled workers who associate with occupations other than proprietors (such as carpenters, bookkeepers, and drivers of motor vehicles) are subsumed into occupational categories dominated by dependent employees with similar skills.
European approaches within social stratification research differ from the Blau and Duncan classification schema in theoretical conceptualization of self-employment. Erikson, Goldthorpe, and Portocarero (1979) devised a now widely used class schema that assigns professional self-employment as well as self-employed large proprietors to occupational categories dominated by individuals associated with professional and managerial occupations in dependent employment.5 Other self-employment is divided into small proprietors either with or without employees (IVa, IVb, respectively) and farmers (IVc). In most applications of the schema, categories IVa and IVb are collapsed into a common category of proprietors. The EGP schema thus in some respects inverts the Blau and Duncan classification system's treatment of self-employment by assigning professional self-employed to their related dependent occupational positions and defining the petty bourgeoisie as a residual category of non-professional and nonmanagerial self-employed.
In spite of these differences, however, both American and European approaches share the common feature of assigning certain forms of self-employment to class categories that include self-employment and dependent employees and that provide no explicit attention to "residual" self-employed occupational categories--that is, many forms of self-employment are simply excluded from focused analytical examination. In the Blau and Duncan classification, the nonprofessional self-employed are not considered; in the EGP schema, the professional self-employed are not identified as self-employed. Both these prior conceptualizations have thus become increasingly problematic as conceptual frameworks as self-employment has become increasingly heterogeneous: "the partial renaissance of self-employment"--as the OECD report cited above has labeled this economic phenomenon--is occurring precisely in those areas that have been analytically obscured by prior stratification work.
While social stratification researchers have differed in their conceptualization of self-employment as a category (e.g., in whether self-employed professionals or laborers should be included in this social grouping), there has been a general consensus among these scholars that the activity must be understood in some fashion as distinct from dependent employment. Self-employment differs from dependent employment in that it allows greater autonomy for individuals, but it also increases individual's exposure to the rewards and costs associated with direct involvement in and dependence on sales of products and services in a competitive and unstable market. As stratification researchers in particular are interested in inequality, they also should consider heterogeneity within self-employment--that is, not just mainly focus on traditional petty bourgeois craft production and proprietorship, but also take account of the potential growth of unskilled self-employment and self-employed professionals engaged in freelancing or other semi-autonomous jobs.
Although the self-employed as a social grouping have often been relatively small in terms of their demographic proportions, they have traditionally been the focus of inordinate attention as they have been recognized as possessing the potential to demonstrate a historically critical role in determining societal-level political outcomes. The autonomy and market insecurity of self-employed activity have been argued to combine to provide both greater freedom and motivation for this social grouping to attend to political matters. The material conditions associated with self-employment have thus been claimed to generate distinct social identities, economic interests, and political orientations relative to individuals in dependent employment (Bechhofer and Elliott 1985). Marx and Engels, for example, in the Communist Manifesto noted: "The lower middle class, the small manufacturer, the shopkeeper, the artisan, the peasant, all these fight against the bourgeoisie, to save from extinction their existence as fractions of the middle class. They are therefore not revolutionary, but conservative. Nay more, they are reactionary for they try to roll back the wheel of history."6 Subsequent sociological research in general has tended to provide fairly similar evaluations. Seymour Martin Lipset, for example, argued that the petty bourgeoisie were the group "most prone to support fascist and other middle class extremist ideologies."7 More recently the analysis by Michael Hout and colleagues of postwar U.S. voting suggested that "owners and proprietors moved from indifference to strong Republican support, especially since 1972."8
The political orientations of the self-employed have been understood as emerging from their contradictory class position in society. Erik Olin Wright, for example, theorized contemporary class location as involving several dimensions: a relationship to ownership, workplace authority, and education-based skills or credentials. While Wright's actual discussion of self-employment has tended to focus on the first two dimensions (ownership and workplace authority relationships associated with size of firm), consistent with this approach would be a consideration of whether the self-employed have professional qualifications and whether they are skilled or unskilled. Given that the self-employed vary along these three dimensions, it follows that their class interests are ambiguous or "contradictory": that is, their interests are only partially aligned with either capitalists or workers. In addition to occupying a contradictory class location with a potential for unpredictable political alliances and support for social movements, this social grouping has also been argued to take an active and disproportionately large role in the functioning of civil society through relatively high propensities for involvement with voluntary associations related to neighborhoods, ethnic groups, and economic advocacy organizations.
While the petty bourgeoisie have manifested a central role in contemporary political narratives, both Marxism and Modernity theories (e.g., Treiman 1970) have predicted the disappearance of this social activity and have emphasized how, as the traditional petty bourgeoisie disappear, so too does direct inheritance of occupational position. The declining presence of self-employment has thus been critical to stratification researchers' contention that education increasingly plays a critical role in mediating the relationship between origins and destinations. Blau and Duncan (1967: 41) note that self-employed occupational categories manifest the most intergenerational "occupational inheritance and self-recruitment"; the occupations are also assumed to have high degrees of stability over the life course due to "stronger occupational investment and commitment than mere employment." These arguments thus assume that self-employment is disappearing, that self-employment is stable over the life course, and that education is relatively unimportant for self-employment activity. The former assumption is clearly no longer valid, and, given the growing heterogeneity within self-employment, the latter two contentions require further empirical investigation.
In the past decade social stratification researchers, confronted with changing empirical realities, have begun to grapple anew with self-employment and to reconsider earlier assumptions about intergenerational inheritance of the status as well as the inevitable historical decline and disappearance of the activity. Robert Erikson and John Goldthorpe (1993: 222), for example, noted that "in the context of advanced industrialism this decline is often checked or indeed reversed" and that while "the petty bourgeoisie reveals a moderately strong propensity for inter-generational immobility, which we associate with the possibility of direct inheritance of property . . . in other respects, it must in fact be regarded as a rather open class."
Research in the past several years has also attempted even closer examination of assumptions about stability, inheritance, and homogeneity of the self-employed by separately studying paths into self-employment and factors related to survival in such enterprises (as opposed to identifying simple patterns of associations from cross-sectional data), as well as by distinguishing self-employment categories on the basis of occupational differences. Specifically, researchers have treated all self-employed as distinct from employees and have begun to distinguish among the self-employed in a clearer fashion. They have also adopted event-history modeling of longitudinal data estimating factors affecting entry and exit. Arum (1997) proposed a simple occupational distinction between professional and nonprofessional self-employment for event-history analysis of this sector in the United States. Subsequently, self-employment has been further explored by distinguishing between professional and managerial proprietors, between skilled and unskilled nonprofessionals, as well as through examining the entry and exit dynamics of these and related occupational distinctions (see, e.g., Shavit and Yuchtman-Yaar 2001; Laferr`ere 2001; McManus 2000b).
We believe that an improved understanding of self-employment is possible, however, not simply by incorporating occupational differentiation more fully into an analytical framework, but more critically through explicitly recognizing individual-level decision-making processes whereby individual characteristics and structural factors affect choices related to entering and exiting self-employment. Specifically, involvement in self-employment implies a process whereby individuals actively decide--after considering the perceived relative costs and benefits attached to distinct paths--whether to enter self-employment or subsequently to remain self-employed (Abell 1996). We posit that most individuals in advanced economies begin their labor market activity with a "taken for granted" assumption of involvement in dependent employment. In almost all of the countries examined in this book (with the exception of Japan and Italy), the vast majority of self-employed individuals have not "inherited" self-employment from parents who were involved in similar enterprises. The decision to become self-employed in these settings thus must typically involve active consideration of relative incentives, resources, opportunities, and constraints, as would an individual's decision to remain self-employed. It is worth noting, however, that in most countries a significant percentage of employed individuals actually demonstrate an expressed preference for self-employment activity: in Italy and the United States, approximately 60 percent of employees claim that they would choose self-employment if given a choice between dependent employment and self-employment; in West Germany, Great Britain, and the Netherlands, the percentages of employees expressing a desire for self-employment are 47, 43, and 33 percent, respectively.9 Through examining the structural factors affecting these choices, we are able to learn about self-employment dynamics and recognize factors that are related to producing changes in the characteristics of self-employment over time.
CHANGES IN WORK AND EMPLOYMENT STRUCTURE
While abstract theorizing about globalization has often been poorly connected to empirically observable social change and thus at times ill-suited as a basis for informing social stratification research (Goldthorpe 2001), concrete and identifiable changes to work arrangements have indeed occurred in many countries. Long-term employment contracts have in certain firms, industries, and countries often been replaced by nontraditional work arrangements (such as sub- and temporary employment contracting as well as other techniques that have redefined employer-employee relationships and commitments). While it is quite likely that these changes are related more to the spread of neoliberal economic policies than any actual increase in globalization of economic trade per se, these changes--regardless of their origins--manifest themselves particularly in, and at times emerge from, self-employment. Understanding the logic of self-employment and the contemporary character of it thus can also inform our understanding of the direction of social developments and inequality more generally. This study, by providing new insights into the growing variation in self-employment and the individual and institutional factors that have produced these changes, is suggestive of larger societal processes affecting social inequality more generally.
In related scholarship on regional economies and small business growth, researchers have advanced arguments about the advantages of small firm flexibility and raised expectations that self-employment and small businesses could have a critical role in job creation (e.g., Piore and Sabel 1984). Thus, entrepreneurship has been argued to be highly dynamic and critical for future economic growth. Because of these beliefs, governments have created specific policies affecting this sector in relationship to larger policies connected to labor market regulation. Ironically, as labor market regulation makes self-employment less attractive to enter (in terms of producing relatively high rewards within dependent employment), it simultaneously has created increased opportunities for entrepreneurial activity by producing significant incentives for employers to outsource production to increase flexibility and reduce internal labor costs. In addition, specific policies have been advanced to promote entrepreneurial behavior to stimulate job growth and utilize self-employment to reduce unemployment associated with labor market rigidities or economic downturns.
Structural conditions conducive to self-employment have also emerged from a broad related set of economic changes that are sectoral in character--i.e., declines in agriculture and manufacturing associated with technological change, increased productivity and the diffusion of manufacturing production away from localized domestic sources, and most importantly, the expansion of the service sector, where entry barriers into self-employment are lower than in capital-intensive manufacturing. These sectoral changes have led to increased individual opportunities for self-employment.
Several other developments, commonly related to increases in globalization and the spread of neoliberal policies, also contribute to a changing composition and character of self-employment, not simply the aggregate rate of this activity. Specifically, in many societies there has been an increase in marginalized forms of labor relationships and in labor market flexibility, and a decline in lifetime dependent employment relationships (Schizzerotto and Pisati 2003; Kim and Kurz 2003; Bernardi 2003). In attempts to increase flexibility, firms tend to reduce their stable work-force and often rely on subcontracting and outsourced self-employed or pseudo-self-employed labor, in particular to adapt to demand fluctuation. With such practices, there has been an increase in the number of self-employed professionals who rely on freelancing for income (e.g., Kallenberg, Reskin, and Hudson 2000; Hakim 1998) as well as in the number of individuals engaged in low-skilled self-employment found in construction or manual services (see chapter 5). Furthermore, in many countries immigration has expanded, with immigrant groups often having high rates of self-employment of a rather marginal kind, particularly when concentrated in ethnic enclaves.
The reemergence of self-employment has thus been associated with related changes in economic structure, technologies of production, and political market interventions (e.g., increases in professionalization, the rise of the service sector, business firm commitment to flexible production, growth of information technologies, and the introduction of policies promoting entrepreneurialism and small firm growth). These changes, however, imply that traditional forms of petty bourgeois self-employment (e.g., small shopkeepers, restauranteurs, skilled craftsworkers) quite likely have continued their declining historical presence and new forms of self-employment in both professional and unskilled occupation have been largely responsible for recent increases.
In addition, in many industrialized countries female self-employment in the past few decades has increased at significantly higher rates than male self-employment and has thus become an increasingly important component of overall self-employment (OECD 2000). During certain periods in their life course, women might be particularly drawn into various forms of self-employment activity that allow more flexible work arrangements. The growth of female self-employment is likely to have produced increased heterogeneity in self-employment overall as such activity has often been associated with simultaneous involvement with significant family commitments and has thus tended to be short-lived, part-time, or marginal in other ways. Increases in female labor force activity also likely produce growing demand for marginal domestically oriented self-employment, as women who enter the workforce outsource traditional domestic responsibilities to other women with fewer occupationally marketable skills (McManus 2001; Hochschild 1997; Connelly 1992).
Self-employment is likely not only increasingly heterogeneous within countries, but also probably more heterogeneous than dependent employment across industrialized societies as it is often directly embedded in family economic and social organization, not just market and state institutions. Family institutions, and in particular the share of women's work devoted to the family and labor market, considerably differ between countries, with such practices likely affecting the development of self-employment--an activity that has always been embedded in family relationships. As changes have occurred in family structure, women's educational attainment, female labor force participation, and forms of self-employment, research is needed to explore whether self-employment's relationship to the family has persisted. The relationship of family to self-employment should be considered in terms of both the effect of family participation on the act of self-employment and the direct intergenerational familial inheritance of self-employment positions.
Researchers have argued that a portion of the increase in self-employment has been concentrated in undesirable occupations and associated with an erosion of the social position of labor (see, e.g., Arum 2001, 1997; McManus 2000b; Kallenberg, Reskin, and Hudson 2000). Self-employment, particularly new forms of transitory and marginal self-employment in unskilled occupations, thus possibly is in part related to a "dark side of flexible production" (Harrison 1994). As a result one would expect that a growing proportion of self-employment has occurred in unskilled occupations with high entry and exit rates as well as strong associations with prior unemployment. On the other hand, self-employment increases have also been associated with the growing competitiveness of small firms and self-employed professionals who have taken advantage of changes in technology to identify specific market niches conducive to entrepreneurial success. In many locations, businesses have adopted strategies of flexible production, including outsourcing and the development of extensive network relationships with small firms and independent contractors (Reich 1991; Powell 1990). These changes have likely promoted a renaissance of small firms and increased possibilities for entrepreneurial success in some locations (Piore and Sabel 1984). While social science researchers have often focused on self-employment increases in both professional-managerial and unskilled occupations, they have also predicted and identified the decline of traditional petty bourgeois self-employment (e.g., Blau and Duncan 1967).
RESOURCES, OPPORTUNITIES AND CONSTRAINTS
For this project our analysis is guided by the assumption that individuals under conditions of bounded rationality consider existing and potential resources, opportunities, and constraints when choosing to enter or leave distinct self-employment occupational paths. Resources most salient to individuals in calculating the likelihood of self-employment success are education, family support (both inheritance and spousal assistance), and work experience (i.e., occupationally based human capital). The most significant opportunities and constraints to consider are related to the occupational structure an individual faces as well as his or her particular location in such a system. For example, an individual located in certain industrial sectors such as construction would have greater opportunities and fewer constraints to move into self-employment than an individual located in manufacturing. In a similar fashion, an unemployed individual who faces significant constraints in finding dependent employment is likely to assess opportunities in self-employment more favorably.10
Education, particularly tertiary and vocational, provides individuals with human capital skills that are transferable between either self-employment or dependent employment. Prior research has suggested that in all countries education is related to self-employment, but research so far is inconsistent concerning the direction of educational effects. Generally, higher qualifications improve the likelihood of self-employment in a given area of work since self-employment compared with dependent work often requires qualifications in addition to the purely functional skills necessary for doing a given job well (such as management and planning skills, knowledge of financial affairs, understanding of market opportunities, ability in personnel guidance, and skills in public or customer relations). However, while several of these abilities can be trained, they often are not learned in schools and rather are acquired in family or other contexts. Prior research has found that in some countries, such as Germany and postsocialist Russia, education has a quasi-linear positive relationship to self-employment (Gerber 2001b, Luber et al. 2000). In other countries, such as the United Kingdom and Israel, education has more curvilinear effects, with both low and high levels of education associated with increased likelihood of self-employment activity (Shavit and Yuchtman-Yaar 2001; Luber et al. 2000; Meager, Kaiser, and Dietrich 1992). Meager et al. (1992) have also identified positive effects of vocational education on male self-employment in Germany.
While education in general is positively associated with self-employment entry, exceptions to this trend exist. Countries with high levels of self-employment, such as Italy, Portugal, and Greece, tend to have negative associations between education and self-employment, as much of self-employment is in low-skilled occupations (Blanchflower 2000). Research on national minorities and the prevalence of ethnic entrepreneurial enclaves has also suggested that self-employment can be used as an alternative strategy for social advancement when access to educational attainment is blocked or difficult to obtain (see, e.g., Wilson and Portes 1980; Aldrich and Waldinger 1990; Light 1992; Shavit and Yaar 2001). Individuals from lower social origins (e.g., ethnic minorities) or those who otherwise have less ability and aptitude for educational achievement, it has been argued, seek professional/managerial self-employment as an alternative route for acquiring successful occupational status given otherwise inadequate possession of educational credentials required for high attainment in dependent employment (Shavit and Yuchtman-Yaar 2001). From this perspective, individuals make a rational choice that, given their personal attributes, pursuing further educational attainment is unwarranted relative to labor force participation in a specific area (see, e.g., Breen and Jonsson 2000; Breen 1999; Manski 1993); related and supplementary factors can then lead these individuals into self-employment activity.
Work-Experience and Employment Position
Most individuals start their working life as wage or salary workers; very few work in autonomous enterprises when they first enter the labor market. They need work experience and must accumulate capital and financial resources for self-employment success, but they will also want to have sufficient years of expected future earnings to make initial investment in their self-employment projects worthwhile. Individuals at mid-career are therefore particularly likely to enter self-employment. Prior research suggests that work experience, given its association with skill acquisition, and knowledge of markets necessary for self-employment success--measured directly or indirectly through proxies such as age--have strong positive, curvilinear effects on self-employment likelihood in most advanced economies (Blanchflower 2000; Luber et al. 2000).
The early work career, however, and the characteristics of the position in the labor market an individual occupies at a given point are highly influential for the likelihood to enter self-employment and for the kind of self-employment entered because they define the opportunity structure and constraints that individuals face. Strohmeyer and Leicht (2000) show how training in a small firm strongly increases the probability of later becoming self-employed, most likely because the worker learns the skills required for operating a business. In terms of industrial sectors, we know from existing research that self-employment is particularly likely in certain industrial areas (e.g., construction and services) and less likely in sectors with large requirements for economic capital, such as manufacturing. We also expect that when individuals assess opportunities and relative incentives to become self-employed as opposed to remaining dependently employed, they will be unwilling or reluctant to move downward in terms of an occupational hierarchy. Individuals will utilize resources associated with their prior occupational position as much as possible to maintain or improve their occupational location. Unemployed individuals have a relatively higher likelihood to move into self-employment than dependent employees with similar characteristics as they are already actively on the market searching for employment positions. They must give up only the potential rewards of prolonging their job search to find even better alternatives in the future than those opportunities suggested by entering self-employment immediately. Employed individuals, on the other hand, must give up a job and continuous earnings tied to their present position when moving into self-employment. The opportunity costs will thus tend to be much higher for employed than for unemployed individuals, and the latter should be more likely to enter self-employment than the former. While unemployment likely affects an individual's decision to become self-employed, this does not necessarily imply a positive relationship between aggregate-level unemployment and self-employment rates. As the overall rate of unemployment increases, the pool of individuals particularly prone to entering self-employment rises; unemployment rates, however, also likely affect two other related phenomena--the likelihood of a dependent employee being willing to leave a job to start self-employed activity, and the likelihood that an existing self-employed enterprise will be able to sustain itself and survive in relatively undesirable market conditions.
Family support, in the form of both occupational inheritance and spousal assistance and involvement in the enterprise, is a significant individual-level resource that could affect a person's decisions both to become self-employed and to remain self-employed. Previous research in varied national settings has identified the effect of parental self-employment background on the likelihood of respondent's self-employment. Strong inheritance of self-employment is a common finding in most studies of intergenerational class mobility. Self-employment is more likely when parents have been similarly employed for a number of reasons. First, there are cases of direct inheritance of small family businesses. Second, parental involvement in self-employment could provide access to financial collateral and network contacts, thus reducing liquidity constraints (Laferr`ere 2001). Finally, and perhaps most importantly, familial involvement in self-employment provides "a taste for self-employment, a knowledge of enterprise from the inside" and a general "socialization of youths to the norms, values and skills ('entrepreneurial capital') that are particular to the self-employed class" (Laferr`ere 2001:22; Shavit and Yuchtman-Yaar 2000:61; Aldrich, Renzulli, and Langton 1998). The significance of family resources and traditions is bound not just to the intergenerational transmission of self-employment. Self-employment is often more generally embedded in familial relationships than is the case with dependent employment. Existing research has highlighted the degree to which self-employment frequently thus not only is inherited but often manifests itself as a family business involving the spouse, and sometimes children and relatives (Aldrich and Zimmer 1986). Family conditions, notably family status and self-employment of the spouse, thus also must be assessed as a significant determinant of self-employment dynamics.
Women--since they differ from men in their relative resources, opportunities, and constraints--will have distinctly different patterns and determinants related to self-employment. Researchers have long focused on the distinct character of male and female self-employment (see McManus 2001 for a review of this topic). Female self-employment tends to occur--like dependent employment--in sex-segregated occupations and particular industrial sectors, such as service branches (Lohmann and Luber 2000; Luber et al. 2000; Wharton 1989). Since family responsibilities are allocated unequally, pronounced differences appear in the effects of various family characteristics on male and female self-employment (McManus 2000a; Boden 1999; Carr 1996; Loscocco and Leicht 1993; Kallenberg and Leicht 1991). Female self-employment is of shorter duration and is typically less stable than male self-employment (Lin, Picot, and Compton 2000). Female self-employment is often part-time and has low income associated with it, although heterogeneity produces large variation in characteristics associated with this activity (Kallenberg, Reskin, and Hudson 2000; Manser and Picot 1999; Hakim 1998; Arum 1997; Devine 1994). Women with primary responsibility for childcare have incentives to utilize self-employment to supplement family income and thus tend to have greater involvement in marginal, part-time, or unskilled self-employment associated with flexible work arrangements.
It is well known that the extent of self-employment varies between countries. We conjecture that such differences are not just a result of different levels of development and a country's economic modernization, but are conditioned by varying institutional arrangements that make entry and survival in self-employment more or less likely. Such conditions will also make self-employment more or less viable and attractive for different population groups. Countries will thus vary not only in the extent of self-employment but in the characteristics of self-employment as well. Through a dynamic and comparative design, this book contributes to a better understanding of both the general nature of self-employment and its variation in different institutional contexts.
Countries selected for the project differ in many socioeconomic and political aspects. In our discussion, we find it useful to focus on two illustrative dimensions of country-level differences: labor market regulation and family-based social capital. Figure 1.1 proposes how the eleven countries in our project might be theoretically placed within a two-dimensional chart based on these factors. We have operationalized labor market regulation by relying on a measure of employment protection legislation developed by the OECD. The measure is a composite index generated through examination of legislation affecting regular employment, temporary employment, and collective dismissals. Published information existed for nine out of the eleven countries in our study--Russia and Taiwan being the exceptions. For those two countries, authors of the relevant chapters consulted with colleagues who had expertise in the area of country-specific labor market regulation and generated estimates for this measure that approximated the values OECD researchers likely would have assigned if these countries had been included in the original study.11 We have italicized these two countries in the figure to indicate that these are author-estimated values on the OECD Employment Protection Index, generated solely for comparative purposes in our project. In the figure, we have also used two arrows to identify three countries with rapidly changing regulatory regimes. Taiwan prior to 1996, during the period when observational data used for the case study described in this volume were collected, had only rudimentary employment protections in place; in the late 1990s the country implemented a range of new labor market regulations with the current Employment Protection Index now estimated at 1.9. A second arrow in the chart indicates the degree to which labor market regulations in both Russia and Hungary were significantly relaxed in the 1990s, as the economies underwent an economic transition, and again during the period under observation in the case studies.
We have operationalized family-based social capital very simply with a widely available--or easy to generate--measure: the percentage of adults aged 25-59 who are living in a household with their parents.12 While adults live with parents in their households for a complex set of reasons (including not just cultural differences but economic necessity, housing shortages, and societal age structure), we believe this rudimentary measure is still useful for our purposes as it is suggestive of how intergenerational family relationships vary significantly across societies. Our concept of family-based social capital is distinct from broader notions of social capital that can be found in many aspects of social organization. While social capital in general has been utilized as a concept to discuss quite diverse phenomena both within our country-specific chapters and within the field of sociology more broadly, in our cross-national comparisons we are interested not in general social relationships associated with interpersonal networking or social integration of communities, but in a more narrow definition that serves to identify variation in the role of the family unit in societal economic organization.
It is worth cautioning readers that placement of countries in this figure is thus based on crude empirical measurement of our underlying concepts. Nevertheless, while the figure is not a definitive social mapping of these differences, the illustration is suggestive of a strategy useful for conceptualizing relevant factors related to cross-national variation. With these caveats in mind, the figure identifies variation along two dimensions, with countries farther to the right on the x-axis having increased labor market regulation and countries higher on the y-axis indicative of societies that are organized more extensively on principles of family-based social capital. Societies that have greater emphasis on family-based social capital would likely be characterized not just by a higher prevalence of adult children living at home with their parents, but also by the increased salience of extended kin networks and stronger family ties in general. We do not wish to reify these societal differences nor to imply that they are produced simply by national cultural differences as opposed to other economic, political, and institutional factors. Welfare-state policies, for example, likely emerge from and structure the extent to which family-based social capital is salient. Welfare-state regimes also vary in the degree to which they promote patriarchal family structures or, on the contrary, defamilialization (see, e.g., Esping-Andersen 1999). In addition, families in some countries serve to compensate for weak welfare states, but in other settings, individuals are left to the vagaries of market forces.
Labor market regulation also likely structures the prevalence and forms of self-employment activity across countries. Variation in labor market regulation is closely related to cross-national differences in how difficult it is to incorporate a self-employed business as well as the relative incentives to engage in self-employment that are associated with taxation policies. In the late 1990s the Union of Industrial and Employers' Confederations of Europe (UNICE) estimated differences across countries in the formal requirements associated with establishing a new business. Countries in our study, which were also included in the UNICE report, varied in terms of these requirements in a manner similar to our discussion of labor market regulation and political-economic ideal types. In the neo-liberal economies of Australia, the United Kingdom, and the United States, new businesses could be established with one procedure in less than a week and with a minimum amount of financial costs associated with regulatory compliance. In the corporatist states of France, Germany, and the Netherlands, as well as Italy and Japan (the other countries in our study that are described as having higher degrees of labor market regulation), on average 8.4 procedures were required to establish a new business, the average time necessary to comply with regulations was 9.4 weeks, and administrative procedural costs on average were 5.7 times higher than in the neoliberal economies.13
Cross-national variation in institutional relationships affecting the establishment and regulation of small businesses also occurs through variation in taxation policies adopted by state authorities, which create incentives for the establishment and incorporation of businesses as well as for self-employment in general. Relevant tax policies to consider in this regard include differences in tax liability for unincorporated and incorporated businesses; variation in requirements for retirement and pension contributions (for employers, employees, and the self-employed); special tax regulations affecting family business inheritance and transfer; and variation in tax rates for individual, as opposed to corporate, income and profits.14 While cross-national variation in regulatory procedures and taxation policies is significant, detailed discussion of these factors has been relegated to individual country-specific chapters. For purposes of cross-national comparison, we simplify the extraordinary legal and regulatory complexity of the institutional environment by discussing a summary measure of the degree of general labor market regulation affecting both self-employment and dependent employment in a society.
Three commonly discussed and distinct political-economic ideal types are clustered in figure 1.1. In the lower right-hand side of the illustration, Western European corporatist states with relatively high labor market regulation and comparatively low reliance on family-based social capital are represented by France, Germany, and the Netherlands. These societies have average scores of 2.5 on the OECD Employment Index and 4.5 percent of adults aged 25-59 living with parents in their households. The Anglo-American neoliberal economies of Australia, the United Kingdom, and the United States are clustered in the lower left-hand corner of the figure, being relatively low on both labor market regulation and family-based social capital (average score of only 0.9 on the OECD Employment Index and 3.6 percent of adults aged 25-59 living with parents in their households). In the middle of the figure, economies that are undergoing postsocialist transformations are represented by Hungary and Russia. (continued)
Return to Book Description
File created: 8/7/2007
Questions and comments to: email@example.com
Princeton University Press