This file is also available in Adobe Acrobat PDF format
ASSESSING JOB QUALITY IN THE AFFLUENT ECONOMY
The Paradox of Job Quality at the Millennium
Work is no mere passing show for a contemplative community of social scientists. Almost everyone gets to do it. Work itself is a major and defining part of most people's lives. It takes up a large proportion of their time on this earth, and profoundly molds their life-experiences. Writers about work are also their own subjects.
In the affluent economies of the industrialized world, life at work in the early twenty-first century has evolved in a curious and intriguing way. Workers have, with significant exceptions, been taking home increasing wages, exercising more acute mental skills, enjoying safer and more pleasant conditions at work, and spending less time there. Yet they have also been working much more intensely, experiencing greater mental strain, sometimes to the point of exhaustion. In many cases, work has come under increased and unwelcome control from above, leaving individual employees with less influence over their daily work lives and a correspondingly less fulfilling experience than before. In these ways, work in the recent era has become more demanding. Meanwhile, significant minorities of workers continue to endure great uncertainty regarding the future security of their employment. Overall, employees are getting no more satisfaction from their time at work than they used to, even though the material wealth of nations has been increasing.
The tension in these observations, all of which are to be described precisely and supported with evidence in this book, lies in the contrast between the increasing wealth in national economies and the ambiguous changes occurring in the quality of jobs. One might, on the face of it, expect increasing affluence to be devoted to improving all the important parts of life. Take, for example, an earlier period, the so-called "golden age of capitalism" that prevailed after World War II until the early 1970s. The central socioeconomic principle, which sufficed to prevent the explosion of conflict in the ever-larger workplaces that had been envisaged by Marx and others, was an implicit social compact between employers and the majority of workers: consent at the workplace, including acceptance of managers' control of the organization of work, was traded for a share, through increased wages, in the rising affluence that twentieth-century technologies were permitting. Superimposed on this social compact were the political and economic principles of Keynesianism, which rescued capitalism, at least for a while, from degenerating into a fragile mass-unemployment system by showing governments how to mitigate cumulative collapses of aggregate demand. While the U.S. economy had emerged from World War II being far more productive than European and Japanese capitalisms, modestly rising wages and improving work conditions (such as reduced hours, longer holidays, and safer work-places) became the expectation of workers in all countries partaking in advanced capitalism. America's workers were already basking in the "affluent society," and the intellectual issue for economics had become, according to J. K. Galbraith's famous study, how to understand the construction of consumer demand once absolute poverty and wants had been diminished.1 In Britain, the workforce was gradually shifting into newer, "cleaner" industries and into nonmanual occupations. With rising wages and greater security, and with most sections able to resist encroachments of managerial control, the average quality of working life was improving in that first genuine affluent age in history.2 Prime Minister Harold Macmillan was able to secure re-election in 1959 on a slogan which probably resonated as much with the electorate's working life as with its wider experience of life: "You've never had it so good."
True, sociologists were debating whether the growing affluence was genuinely diminishing the differences between the traditional middle-class and working-class work experiences, and whether the creeping growth of bureaucracies and large-scale technologies would increase alienation among workers (Goldthorpe, 1969; Blauner, 1964); and it was being maintained that the cost of male manufacturing workers' rising pay had been a sacrifice of autonomy and intrinsic work satisfactions (Goldthorpe et al., 1968). Psychologists were already investigating the microconditions of work that engendered greater or lesser levels of satisfaction and worker motivation (Herzberg et al., 1957). Economics had less to say about the quality of work life, having already drawn in more closely the boundaries of the discipline. A long-standing contribution was the proposition that higher wages had a theoretically ambiguous impact on the work that people chose to do. An increase in wages made it more costly for workers to stay away from work, and this might mean that some workers would put up with longer hours and poorer work conditions if it gave them substantially more resources to enjoy in their leisure time. On the other hand, higher wages meant that people were more affluent, which could lead them to choose more leisure time and better conditions. In practice, male workers who had been working long, burdensome hours were choosing to reduce their hours, as they had done (supported by trades unions and political pressure) since the early nineteenth century. Overall, with relatively few exceptions the prevailing view, often unstated, was that if the conditions for affluence could be maintained, the quality of working life could and would continue to be improved in parallel.3
The current era, since the end of the golden age of capitalism in the 1970s, has continued to be one of growing wealth in all the industrialized nations. But it is different from the 1950s and 1960s in two crucial respects, whose significance merit some preliminary elaboration in the ensuing sections. First, much more is known than in earlier eras about what has been happening to the quality of work life, through the arrival in the investigators' toolkit of a powerful instrument, namely the consistent social survey series. Second, there have been enormous structural changes in the global economy, with ramifications reaching down to workplaces across whole nations; these transformations lie behind the many contradictory and ambiguous changes in the working lives of ordinary people to be investigated below.
Revealing a History of the Present
The rise of the large-scale social survey, coupled with an expansion of official statistics capabilities throughout the industrialized world, has had major consequences for our understanding of life in the workplace. Simultaneously, the facility with which administrative and employment-based data can be gathered and processed has improved. Computers now enable complex databases to be analyzed with ease, and have transformed empirical research in the social sciences almost beyond recognition. Some of these social surveys reveal a great deal about trends in the quality of working life across the whole of a nation's workplaces.
This investigative method is largely unavailable to work historians of earlier periods. Of course, the social survey was not invented recently: it has a distinguished history going back at least as far as the 1880s when Charles Booth began publication of his Life and Labour of the People in London (Bulmer et al., 1991). What is new is that modern governments and other institutions have funded consistent series of high-quality social surveys, permitting the analysis of social change. It is fortunate that at least some of these series have included a focus on work. Just as the invention of a new measurement instrument normally transforms a field of science--in our times, think of genetic fingerprinting--so the embryonic development of representative social surveys is already advancing the understanding of social and economic change.4
The aim of this book is to use the instruments now available to develop a "history of the present." This phrase is intended to convey the intention to use the surveys to explain what has actually been happening in the present era, rather than to test the validity of higher-level theories. In spite of periodic disruptions, there are no nations which were industrialized thirty years ago that have not continued to grow. On average, the people of the OECD countries, the club of rich nations, had become by the time of the millennium nearly twice as rich as they had been in 1970. But what has been happening to the quality of working life in these years of affluence leading up to the third millennium? Are people using their first world affluence to buy themselves a more decent experience when they set out to earn their daily bread? If not, why not?
The informants of this history are, for the most part, the respondents to nationally representative surveys. For understanding the world of work, these informants are in an unrivaled position: they are the ones doing and experiencing the work, and for many aspects of work no other observers can generate more reliable information. Future historians of the current era will uncover many now-hidden documents, and will be able to reinterpret with hindsight the social, political, and economic forces impinging on work. But the social survey record is already in; it may be reinterpreted but cannot be expanded in retrospect.
The availability of consistent survey series, with workers' direct reports about their work, is especially felicitous, in view of widespread assertions that the quality of working life has changed fundamentally in the last decade or so. Claims, which vary in status from formal academic studies (e.g., Burchell et al., 1999; Cappelli et al., 1997) to popular social commentary (e.g., Toynbee, 2003), have often focused on the down side--a deteriorating work-life balance, increased workplace stress due to effort intensification, stagnant or declining wages for unskilled workers, rising insecurity. But counterbalancing claims of improvement are also celebrated, for example that many workers have become more "empowered," or that they are doing more skilled (and so more fulfilling) work (e.g., British Government, 1992). Taken together with rising wages, this mix of changes breaks any simple connection that one might have expected to find between rising affluence and improved quality of work life. Finding out whether these claims are supportable generalizations is one of the main contributions that social survey analyses can make: they lend an authoritative counterweight to the cult of the vignette, the nice or shocking story, which is too often the sole evidential method of popular or journalistic social science. Social surveys can also assist in tracing connections between the quality of work life and the major structural transformations of the global economy. They can yield an understanding of the proximate causes of changes in work-life quality, and by identifying the key factors can help to focus policy where it is most effective.
But survey analyses can only go so far. Though utilized here to construct a history of the present era, social surveys are not themselves good at uncovering deep processes of change--for this purpose surveys need to be used in a complementary way with ethnographic and case study methods. Nor do survey analyses displace the methods of the regular historian: they do not explain, for example, why or when computers were invented, why China entered the market for manufactured goods, or why Americans voted for President Ronald Reagan and his policies of minimum-wage reductions, and why Margaret Thatcher's policies of reducing union power triumphed in Britain. Such matters are the subject of a wider politico-economic canvass. Moreover, surveys that have relevant information about working lives are quite rare. Only a handful of suitable series of high-quality surveys exist. While low-quality surveys abound, typically purveyed by commercial concerns or interest groups, these often generate findings that are misleading because they are derived from unrepresentative samples; and they are rarely consistent over time. Still lacking are appropriate series of surveys in most developing or transitional countries--this is the prime reason for confining the analyses of this book to the advanced industrialized world (the "North").5
Analyses in these surveys mainly rely on subjective accounts by workers of their own lives. This source of information is not favored by some, but aversion is unwarranted. Data that have useful information and predictive value deserve, subject to ethical bounds, to be utilized in any science. Unbridled skepticism needs to be replaced by a systematic approach to careful validation and reliability-testing of data--the stuff of scholarly papers. Subjective reporting from individuals can reflect biases driven, for example, by respondents' unconscious need for social esteem or by poor recall of past events. Yet these biases and errors can be assessed, and a considered weight given to each set of findings. Set off against the biases are the circumstances that make for minimal measurement errors: individuals know a great deal about what happens in their own jobs--much more than is revealed, for example, in bureaucratic records.6 The countless people who were selected by chance to be in these surveys became the collective representatives of working people in their national populations; they will be telling their own story in this book. Whose story is more accurate than theirs?
The Changing World and the Everyday Workplace
If the experience of work is changing, it is not hard in broad terms to see why. The world economy has changed in several profound ways since the golden age of capitalism ran out of steam in the early 1970s. Though the end of this era was signaled finally by the collapse in 1973 of the postwar international monetary system, the roots of crisis lay deep in the contradictions of American military and economic hegemony, technological slowdown, tight labor markets, and growing shortages of raw materials (especially oil). High on the list of major transformations of modern economies since the 1970s are: the growth of service industries, and the corresponding rise of service occupations at all levels--from sales assistants to lawyers; technological changes, led by the progressive growth of computer power and software, which pervade all industries; new or reformulated ideologies of management, which come in both hard and soft versions, both designed ultimately to raise worker productivity; the progressive withdrawal of governments from active participation as producers of goods or services; and, not least, the globalization of trade and production and consequent intensification of international competition, developments which both lie behind the new technologies and management ideologies and impinge directly on working lives.
Service industries have grown more rapidly than production goods industries in the affluent economy, largely because of consumer demand. Since many services have to be delivered as they are generated, the customer becomes a major proximate controller of work flows, and there are corresponding requirements for workers to attend to their tasks at flexible times. One cannot store up nursing services, for example, when patients need care. Services are also highly heterogeneous in their skill requirements and in the extent to which they can be automated. The routine services are those that have most easily been replaced by new information and communication technologies. But these technologies create jobs as well as destroy them. Highly skilled workers, in particular, have been in greater demand across the industrial spectrum. Generic skills involved in information processing, communication, problem-solving, and team-working activities have come to the fore. However, many unskilled activities are also emphasized in the advance of the service economy--especially those nonroutine activities that are hard to automate. The affluent economy sees a return of servant occupations, there to pack bags, clean floors, and secure property--a renaissance which mocks earlier expectations that the servant class had disappeared forever in the first part of the twentieth century.
Many changes in the modern workplace can be traced to the global expansion of trade and the internationalization of manufacturing facilities. Well known is the public display of firms closing down plants located in an advanced industrial economy and opening up in a low-wage economy. The main impact of such relocations and of direct trade with low-wage economies may be yet to come, but their significance hitherto has often been exaggerated. Most trade still takes place between the advanced industrialized countries. The significance of trade, for the work-places of the North, lies with its link to competition. Trade expansion is part of an increasing competitiveness of modern industry, which puts businesses and their managers under continual pressure to defend existing markets and find new ones through aggressive cost-minimization and innovative expansion policies.
Alongside these increased pressures, however, managers in many countries have been handed more power, with the declining strength of trade unions and the appearance of more business-friendly government; at the same time they have been presented with new ideologies of control. There is rarely a consensus as to how management should run their labor processes. Strange dichotomies have arisen: managers have to choose between "soft" and "hard" human resource management (HRM) policies. Soft HRM policies are designed to engender workers' commitment and innovative practices, but they raise labor costs and lessen managers' ability to reduce those costs without breaking the "psychological contract" with workers. Hard HRM policies, by contrast, go hand in hand with "lean and mean" systems, where no implicit promises are made to workers, and where competitive pressures to reduce costs lead to a continual re-examination of production methods, using management tools like "business process re-engineering" and involving occasional downsizing of establishment workforces. The soft/hard HRM dichotomy is parallel to the putative choice for managers between a "high road" and a "low road" to profitability. The former involves high wages, high skills, security, and commitment; the latter, low wages, no training, and frequent layoffs. Managers' choices vary along this spectrum, and there are no established wisdoms to determine which approach generates the best performance.7
Managers' decisions which affect the quality of work life are now taken, however, in the context of an expansion in the range of work scenarios that emerge with the diffusion of computer and telecommunications technologies. In some apocalyptic visions, the possibilities for replacement of labor have been deemed to be so large as to herald vast reductions in employment, even the "end of work" (Rifkin, 1995). In a more considered overview of the evidence, Castells (1996) disposes of this fantasy and redirects attention to the systematic "individualization of labor in the labor process" (265). In the "network society," one sees the reversal of the previous trend toward socialization of production (larger workplaces, collective representation, and so on). The new technologies facilitate the customization of products, the development of decentralized businesses, the subcontracting of business functions, and the introduction of new business practices to control quality and performance within core businesses. These business developments present consequent demands for workers to be prepared with relevant new skills and to be willing to accept new forms of work. Indeed, contemporary technologies often directly facilitate new working arrangements such as telecommuting, working while traveling, or shift-working.
Nevertheless, the consequences of modern technologies for the quality of work life in this modern affluent economy are hard to trace in detail, and adhere to no ironclad law. There are many possible ways in which technologies can be used. Best practices are imperfectly known, because they are likely to vary from culture to culture, and because technologies and attitudes are forever changing. There may be no single optimal way to use technology. Even if there is, diffusion of awareness of best practice could be permanently blocked or could indefinitely lag behind the forces that are changing the optimum.8 The implications for workers can be ambiguous and contradictory. Moreover, though workers' lives are the objects of study, workers are also subjects, real players in the game of workplace conflict and consent. The power and the culture of managers, the institutions of the labor market that may or may not foster cooperation and diffusion of best practice, and the partisan support of governments yield an irreducible indeterminacy to the quality of working lives.9 For these reasons, the connections between technologies and working lives are only imperfectly understood by scholars.
To begin to understand the links between these changes and the quality of the work lives being experienced in modern industrial economies, it is necessary first to examine the ingredients of a "good job." Unfortunately, this is not a straightforward matter, not least because there are considerable differences of emphasis between the social sciences in this respect. The rest of this chapter sets out the approach to be used in this book. It then takes a first look at how social survey findings have been able to contribute to an understanding of changes in job quality.
What Makes a Good Job?
There is a school of thought which maintains that, if one wants to know what sort of jobs are being generated in modern industrialized economies, one need only look at their wages. Pay is said to be the single most important factor defining the quality of a job. Moreover, wages generally are positively correlated with other favorable working conditions. Therefore, it is held that a picture of wages over time would not be substantially different from any more detailed picture of overall job quality.
Yet, is pay the overriding factor in job quality? If pay is increasing, does this mean the job is improving even if other aspects of the job, such as the amount of effort required, are deteriorating? To consider these questions, and to evaluate the significance of the nonpay aspects of work, a conceptual framework is needed. Differences among writers on the quality of work life stem at least in part from their different views of how human activities are to be evaluated.
The "quality of work life" or "job quality" is constituted by the set of work features which foster the well-being of the worker. This definition is worker-centered: it refers to what is good for the worker, not what an employer or customer might want (though these can affect the worker's well-being directly or indirectly). This book will focus only on paid work and will assume that there is normally a separable domain of work that is important for a person's overall well-being but nevertheless only one part of it. Though "work" and "job" are not universal but historical categories, discussion of how the boundary between work and nonwork could be defined is omitted. Even with these limitations, the notion of job quality is neither simple nor uncontested. Different notions are rooted in the major philosophical systems that have ordered the social sciences for at least two hundred years. Broadly, there are two ways of thinking about well-being and, hence, the quality of work life: the subjective and objective approaches.
In the subjective tradition, with its origins in nineteenth-century utilitarianism, a person's well-being is constituted by the extent to which he or she is able to satisfy personal tastes or preferences. Among the major subjective approaches to well-being is neoclassical economics, the dominant paradigm which governs the way most economists think about normative issues. The neoclassical approach to job quality is found in the theory of compensating differentials, which originated with Adam Smith and the classical tradition. Economics teaches that a job is an exchange of certain approximately specified obligations (hours, effort, duties, and so on) for a package of wages and other benefits. Both the obligations and the benefits are multifaceted. Neoclassical economic theory is purposefully silent on the relative importance of different obligations and benefits for a worker's well-being. The quality of work life is constituted by the overall utility function, which depends entirely on a subjective evaluation of the set of work characteristics. The valuation comes from the individual, and is quantified through market exchange. On the assumption of competitive market conditions, different jobs are predicted to offer packages of obligations and benefits that equate overall utility across jobs for the marginal worker. If competition is restricted, some jobs could confer greater utility than others, but access to the better jobs will be barred for some workers. Leaving aside such restrictions, wage differences for equally productive workers are thus interpreted as "compensating" differentials. Higher pay acts as the incentive to attract workers to less favored jobs. However, people are heterogeneous in their preferences for benefits and their distaste for work obligations, and firms are heterogeneous in their capacity to supply benefits or to reduce the burden or risk of work obligations. A matching process occurs in a competitive market, in which workers with the least aversion to poor working conditions gravitate toward high-paying companies that would find it most costly to avert such conditions. Thereby is constructed the "hedonic wage function" that predicts a correlation between wages and less favored work characteristics, among workers with similar skill.10
To many economists, the chief attraction of this subjective approach is that it avoids having to lay down in advance, or from above, what really matters in a job. Wants and needs are indistinguishable in this paradigm. Nevertheless, in the progressive development of economics' claim to scientific status, the subjectivist approach to well-being has led economics down ever-narrower routes. Despite the individual-oriented basis of the discipline, ironically there arose an understandable but awkward distrust of the individual as a source of empirical evidence. Individual evidence about preferences, it was held, could not be used to make interpersonal judgments of well-being. This conclusion led eventually to a rejection of classical utilitarianism at the center of economics. The main criterion upon which judgments could be made became the Pareto principle, which states that welfare gains can only be said to occur if one or more persons' lot improves while nobody loses.11
The distrust of the individual as a source of evidence had another consequence greatly relevant to the task at hand. It meant that direct survey and other evidence of workers' preferenceshave rarely been collected by economists, and usually are ignored altogether. Underlying this distrust is the problem that preferences lack a set metric--either a quantitative scale or, even, any universal point of origin. It is recognized that people's assessments of their own situation and preferences are affected by norms, which themselves are unobserved and determined by societal and historical factors. In addition, survey respondents are subject to many well-known biases. For these reasons, the standard epistemological approach of neoclassical economics--which is rooted in the empiricist tradition--has led it to favor drawing inferences only from the observed behavior of individuals. It is not what people say, but what they do, that counts.
This latter criterion provides one possible justification for economics to treat wages as the premier indicator of quality. Even if, in surveys, only a minority of workers say that pay is a very important aspect of their jobs, the wage rate is typically the most significant item over which they enter into negotiations, or about which they go on strike. This could be said to reveal workers' evaluation of wages as the main item of work reward in most cases. A second justification is that pay is the monetized part of the reward, which through the market valuation process represents the basket of goods and services the workers can consume without borrowing. And, third, this market valuation also reveals that for most workers pay is the largest share of the labor costs which employers incur. The other elements of a person's job play only a minor part, or none at all, in the consumption domain.
Notwithstanding these reasons for prioritizing wages, in recent decades the power of the empiricist urge began to undermine the economists' conventional wisdom that data on preferences carry no useful information. Several economists have maintained that happiness data, based on instruments designed and validated within the tradition of empirical psychology, have application and relevance to economic theory and policy (Frey and Stutzer, 2002a, 2002b). Preferences about work as yet play only a small part in this recent tradition. But interest in "job satisfaction" has simmered for three decades and persists. The chief defense of the use of indices of job satisfaction in economics is that they have been robustly shown to be predictors of economic behavior--in particular, job mobility. Workers who say that they are dissatisfied with their job are more likely to quit (Freeman, 1978; Clark et al., 1998). Nevertheless, it is doubtful whether the use of job satisfaction data commands widespread support among economists.
Even among economists who do use job satisfaction data in their research, there is no consensus on its interpretation. The data are regarded by some as constituting a measure of the utility received from work, which is taken to equal well-being from work (e.g., Clark and Oswald, 1996; Frey and Stutzer, 2002a). Others regard job satisfaction as a useful predictor of labor market behavior, but doubt whether it captures well-being (Hamermesh, 2001; Levy-Garboua and Montmarquette, 2004). The difference of views hangs on the character of the norm against which preference and welfare judgments are formed. In particular, job satisfaction is assessed by workers in part in relation to what they expect from the job. Workers might be conditioned to expect a lot or a little from different jobs. This fact has led the doubters to conclude that job satisfaction is not a measure of utility. Nevertheless, it can be maintained that medium-term trends in job satisfaction in a population can be an indication of trends in well-being: this conclusion depends on an assumption that the norms against which satisfaction is measured change only slowly or not at all.
In a more radical departure from orthodoxy, the equation of utility with well-being, and the resulting policy implications, have been robustly criticized by heterodox economists and others drawing on moral philosophy (Hausman and McPherson, 1996, Chap. 6). In contrast to the subjective individualism of neoclassical economics, other major social science traditions have evolved objective, and frequently explicit, approaches to the quality of work life. These approaches are related to their fundamental concepts of human need and valuation, which differ from that of homo economicus at the heart of neoclassical economics.
The sociological tradition stems in part from Marx, who held that the fundamental characteristic that distinguished humans from other animals was the conscious planning of their productive activities. Building on this foundation, the sociological tradition has privileged the concept of skill as central to the quality of work life (e.g., Wood, 1982). The Tayloristic philosophy of pursuing an extensive detailed division of labor was seen as constituting not only a de-skilling of labor, but also a dehumanization, or "degradation," of labor--its objectification, and its alienation. In sociology, skilled work is seen as involving both complex operations and "autonomy" for the worker. Post-Taylorist and post-Fordist work organizations are viewed positively in this light, insofar as they herald greater worker involvement in their production activities at various levels, including having more influence and discretion over daily work tasks. This perspective on what counts as quality in a job is also taken by non-Marxist sociological schools, which, though they tend to envisage more optimistic trends in job quality within capitalism, place similar emphases on initiative and influence, skill acquisition, and participation.
Complementing this sociological tradition of concern with autonomy and involvement, occupational psychology has developed since the 1950s an impressive body of empirical knowledge about the correlates of job satisfaction and other measures of affective well-being. Underlying this work have been implicit or explicit theories of human need. The characteristics of high-quality work settings are derived using Maslow's theory of needs or other similar frameworks. This empirical work repeatedly confirmed the high importance to workers of being allowed discretion and trust in their jobs. Notable contributions of evidence in Britain came from the Tavistock Institute of Human Relations. The psychological tradition also affirmed the importance of the workplace as a social arena, and hence the relevance to the quality of work life of having good social relations among workers. In recent decades, one of the objectives of (academic) research has been to refine and improve the design of quality of work life instruments for survey usage in selected settings.
Drawing on the earlier work, the 1960s and 1970s saw the beginnings of a Quality of Work Life movement as a driver of social policy. This movement was most advanced, and appeared to have most effect, in the Scandinavian countries (Gallie, 2003). Such programs involved the development of policies designed to improve a range of areas of work life, including health and safety; the humanization of work (including limiting the division of labor); the promotion of worker involvement in decision-making; and security.
These factors are also emphasized in most modern accounts of work quality outside economics.12 Surprisingly, such accounts usually do not include wages, which is often the only factor on the economist's list. This omission could be a matter of semantics, where the phrase "job quality" is taken to refer to factors other than the wage. Nevertheless, the economist's near-exclusive emphasis on wages, and the sociologist's focus on factors other than wages, create a potential obstacle to the development of understanding about job quality. The economist stands accused of not dirtying hands with hard-to-measure subjective aspects of work, the sociologist of ignoring the vulgar reality of workers' material rewards.
An Interdisciplinary Perspective on Job Quality
To counter this tendency, it is useful to observe that there is a broad convergence of the sociological position on work quality with an application of the ideas of Amartya Sen who, among prominent economists, has perhaps done most in recent years to question the scientific and moral case for utilitarianism and individualism in economics. Out of his analysis of the problems facing developing economies, Sen proposed that valuation of human experiences can be conceived in different ways, relevant to the purpose of valuation. Assessment of the achievement of personal well-being is one main method of valuation. Another is through the achievement of "agency goals," that is, the pursuit of goals that may extend beyond the advancement of personal well-being (Sen, 1987, 1993). And, for each of these means of valuation, it is possible also to delineate and evaluate a person's freedom to achieve the objects of valuation. To assess the quality of life, Sen argues that it is important to know a person's "capabilities," which signifies the ability to carry out a range of "functionings" (either doing something or being something) (Sen, 1993). Thus, the ability to perform a range of activities, even if only one activity is chosen, is itself valuable. And, "doing activity x" is seen as inferior to "choosing and doing activity x."
The notion of capability can be applied in the domain of work life.13 An individual whose job involves choosing a set of tasks t from a wider set T can be taken to have a higher quality of work life than one whose job precisely prescribes that tasks t will be performed. The latter has but one choice, whether or not to partake in the exchange: take the job (and do tasks t) or quit. This person's quality of work life is higher than that of slaves who must do tasks t by virtue of their enslavement, having no choice to quit; but less than that of workers who, through participation in the conception as well as the execution of the work, elect to perform the tasks.
Using Sen's approach, the way to evaluate the quality of a job is through the capabilities that are afforded to workers in the job to achieve well-being and to achieve agency goals. The capability to achieve well-being can be thought of as depending on the set of wages and other reward conditions, including future prospects (pensions, security, and so on), and on the range of tasks to be chosen from and actually carried out in the job. The capability to achieve agency goals can be thought of as depending on the extent to which the job enables the individual to pursue personal goals. For example, jobs that permit the satisfaction of providing a service to others could be seen as high quality in this respect, on the assumption that most people do possess such a goal. Alternatively, jobs which afford the possibility of developing fulfilling relations with other people could be considered high quality on the assumption that participating in cooperative action is a valued functioning (at least for many workers, if not all).
The value of the job will depend both on the range of functionings and on the valuation attached to each functioning. For most jobs, the range of types of material rewards is not that large: it chiefly embraces wages, bonuses, any fringe benefits, pensions, and holiday privileges. Though an ability to choose between these types of rewards might be valued in itself, the chief element of job quality is normally the magnitude of the rewards, primarily the wages on offer. In contrast, the range of possible work activities is potentially high. Through the freedom to choose between such activities, sometimes in different ways from day to day, a higher quality of work life can be achieved. The complexity and uncertainty surrounding most modern work processes ensures that it is normally impossible to write very precise job descriptions. The valuation of the work activities is thus likely to be significantly affected both by what those activities are and by the extent of genuine remaining choices among activities within the job. With regard to both the work activities and the wages, the range of capabilities need not be constrained by what the job offers immediately: a high-quality job might be one which offers the prospect of acquiring skills to be promoted or to move to other jobs with more highly valued tasks and higher wages.
The quality of jobs is one of the elements that contributes to a person's quality of life. Having access to a job is a positive indicator of living standards. Conditional on having work, a high-quality job is one which affords the worker a certain capability--the ability and the flexibility to perform a range of tasks (including the necessary sense of personal control), to draw on the comradeship of others working in cooperation, to choose from and pursue a range of agency goals, and to command an income that delivers high capability for consumption. This conception of high-quality work, derived by a simple application of Sen's valuation framework, is similar to the emphasis traditionally given in sociological accounts of job quality, with their focus on autonomy, skill, and social relations, but it includes also the present and prospective material rewards of the job.
This valuation framework admittedly does not provide a decisive means of relative valuation of the different aspects of work quality. Sen rejects a universal law of market valuation and suggests that the choice of items that receive high valuation should be appropriate to circumstances. The choice will be different for poor countries, from what it is in the industrialized world. Relatively good wages might, for example, figure more strongly in what makes for a good job in Bangladesh than it does in the United States. But what makes for an appropriate valuation of the various ingredients of job quality remains undecided. The choice is related to "underlying concerns and values," and there is no escape from declaring one's choice. This incompleteness is of course unsettling for anyone in search of a universally applicable approach to work quality. Sen's defense is that his framework of capabilities and functionings fares in this respect no worse than other evaluation frameworks (Sen, 1993-32).
This book will focus on skill, effort, personal discretion, wages, and risk as key indicators of job quality. There are, of course, many other aspects of jobs, but here the choice of job-quality aspects to be examined is driven both by the interdisciplinary conceptual and theoretical approach just outlined and by the constraints of available representative survey evidence. An additional motivating factor is that each of these aspects of job quality has potential links with major institutional and politico-economic changes that have been features of the modern era, including ongoing economic growth of service industries at the expense of manufacturing, globalization, technological change, and the liberalization of markets.
This approach to job quality is based ultimately on need and not premised exclusively upon individualism or upon a market-driven valuation. To justify including a job characteristic among the ingredients of job quality, it is not necessary that a compensating wage differential should be paid for it. Considerable weight is given to features of jobs emphasized in the psychological and sociological literatures, in addition to those more traditionally analyzed in economics literatures. Individuals are, however, still an important and indispensable part of the analysis. Indeed, the socio-psychological literature draws on analyses of individuals in arriving at the conclusion that intrinsic job features are very important in the determination of work quality; and the evidence on compensating wage differentials contributes important information on job quality. Individuals are legitimate informants of value but not its sole arbiters.
The following summarizes the aspects of job quality to be examined:
1. The skill involved in a job is considered because the utilization of skill is an end in itself, with intrinsic value. Engagement in complex production processes, requiring both conception and execution of tasks in various measures, is the hallmark of distinctively human production activity, and is the means by which people have the potential for self-fulfilment. Changes in skill requirements figure prominently and widely in debates about work life. In recent years an optimistic perspective has emerged through claims about the emergence of a "knowledge economy" in modern industrial nations, in which better knowledge and information are said to have become the keys to competitive success. The belief exists that the new economy requires jobs to become higher skilled throughout the economy.
In chapter 2 the discussion asks whether the average skill levels needed in jobs have really risen as asserted in this "knowledge economy" paradigm. Though this proposition is empirically confirmed for a number of countries, a more mixed picture arises in answer to the question of how far-reaching is the spread of the knowledge economy. In some countries there is suggestive evidence of a skill polarization taking place. Though the highly skilled jobs are the ones that have increased the most, in these countries there is also a disproportionate growth of low-skilled jobs. Though the upskilling is plainly associated, to some degree, with technological advances of the modern era--principally the computerization of jobs, there remains a demand for relatively low skilled jobs where the work has not yet been automated. The probable reason is that this low-skilled work is nevertheless nonroutine and not easily programmed for replacement by a machine.
Another worrying trend unearthed by social survey analysis is the growing tendency for workers and their qualifications to be poorly matched to their jobs. The supplies of qualified workers have expanded everywhere; in Britain this expansion of supply has proceeded at a faster pace than the increased demand for qualified workers.
2. Work effort is considered a key element of job quality, because it is an argument in the standard economists' utility function originating in 1871 with Jevons's labor supply theory, because it is central to a major stream of modern economic theory (efficiency wages theory), because of widespread perceptions in popular discourse that effort is being intensified in the modern era, and because there is psychological evidence linking high work effort to stress, which is the major growing occupational health hazard in industrialized countries.
Chapters 3 and 4 explore evidence about the trends in work effort, both in Britain and in a number of other countries in Europe and elsewhere. The picture is one of fairly widespread, though not universal, effort intensification in the workplaces of the advanced industrialized nations. Here, too, the evidence is in line with much popular comment. It is not so clear exactly why people are working harder. There are multiple causes, including declining union power, increased competitive forces, and in some cases greater managerial pressures to raise productivity. However, the evidence also robustly supports the hypothesis that changing technology and work organization are strongly associated with effort intensification.
3. The personal discretion of workers over their job tasks, and forms of participation in workplace decisions, are covered in chapter 5, with emphasis on the former. These aspects are implied by the capabilities approach, which leads to an attribution of high value to the process of determining and choosing one's activities, as is confirmed by a considerable body of empirical psychology.
The social survey evidence here, confined to just two countries (Britain and Finland), confounds the hopes of management gurus like Tom Peters and Rosabeth Kanter, as well as a strand of discourse among management theorists. Rather than the proposed steady rise in job quality as newly "empowered" workers are allocated more responsibilities and more autonomy, in Britain the trend is unambiguously toward a reduction in the scope for workers' personal discretion over the tasks that they must perform. It is hypothesized that new technologies may be part of the reason for workers experiencing less discretion; but the evidence is sketchy, and it is equally likely that managers have considerable leeway to determine the way that jobs are designed, and hence that their strategic decisions are an independent factor determining the extent of personal discretion. In Finland, the evidence points largely toward stability, though with some increases in personal discretion being observed. Two lessons can be drawn from this. On the one hand, it should teach skepticism of speculations about workplace trends that are borne of writers' imaginations rather than proper evidence; while, on the other hand, it is a reminder that institutional and cultural factors can induce quite different trends in different countries.
4. Pay is, for most workers, the prime source of income, as well as a key aspect of job quality. Chapter 6 presents a cross-national analysis of how pay has changed over recent decades. As expected, in most countries workers have been able to share in the fruits of economic growth, receiving raises in their real average pay. But there are a few exceptions, of which the most significant is the United States. Though American workers received the highest wages during the 1970s, this was no longer true by the turn of the century; for decades, the average pay had stagnated. Job quality is also affected by the fairness of wages--it is inhibited by discrimination and by unjustified inequalities. There are also big differences across nations in the extent to which pay has become more unequal in recent decades: inequality has increased more in the "liberal market" economies, such as Britain or the United States. The considerable increases in American pay inequality, coupled with the failure of average pay levels to keep pace with economic growth, has meant decreases in pay in the lower half of the pay distribution for Americans, while those further up the pay scales have had their job rewards improved a great deal. Though technology has frequently been the assigned prime cause of the rises in inequality, and sometimes increased international trade with poorer industrializing countries such as China has been blamed, much of the evidence points to the significance of institutional and political factors, in particular the decline in union power and the political decision to reduce the real value of minimum wages during the 1980s.
5. Low risks and security are also key aspects of job quality. This book identifies health and safety as the most important aspects of intrinsic security. In this respect, job quality has been improving unambiguously in most countries: the chances of having a serious or fatal accident at work have declined over recent decades. Partly due to the decline in the proportions of workers involved in heavy industry, the improvements also reflect collective learning and improved regulation of workplaces. By contrast, insecurity of the job itself has been claimed in a good deal of popular discourse to be a distinctive feature of modern workplaces, reflecting a supposedly increased pace of organizational and technical change. The evidence, however, does not support the view that there is some secular decline in job security. There have been no dramatic declines in the stability of jobs (as measured by average job tenure). Perceptions of the risk of job loss are found (in Britain, the United States, and Germany) for the most part to track the movements of the aggregate unemployment rate. An exception is the mid-1990s, when perceptions of insecurity in both Britain and the United States were somewhat on the high side, especially among the higher educated sections of the workforce.
6. Changes in job satisfaction and indicators of affective well-being at work are covered in chapter 8. Their inclusion might be justified as proxies for changing utility and hence well-being, within the neoclassical tradition. For those unwilling to accept that assumption, trends in affective well-being should nevertheless be regarded as a significant component of a wider concept of well-being, and demanding of an explanation. It will not now do to dismiss trends in subjective data as lacking informational content or as irretrievably biased. The social survey evidence presents a picture of declining job satisfaction in Germany and in Britain; in the United States there is a remarkable stability in job satisfaction over a long period. The decline in job satisfaction of workers in Britain is found to be associated with the increases in average workloads and with the decline in personal discretion over job tasks.
There could be a temptation to regard the last set of features, the subjective measures of well-being, as the outputs of work-based happiness, while the inputs are the wages, effort, and other job characteristics. One could then estimate the determinants of job satisfaction, analogously to estimating a production function. Thus, one might in this view, trusting in the individual data, deduce what matters in a job from the estimated co-efficients. Using this method, Clark (1998) finds that the content of the job and having good social relationships at work are the most highly valued aspects of job quality. An alternative to a direct well-being measure is to use as a dependent variable individuals' own evaluations of their job relative to an average job. Jencks et al. (1988) regress individuals' job evaluations on job characteristics, and use the resultant coefficients to project an index of job desirability for all jobs. However, to use either of these methods one has to endorse fully the subjective conception of work quality. The valuations are likely to change and to vary across societies. A practical drawback is that surveys of nationally representative samples carrying both subjective job quality indicators and rich details on the characteristics of jobs are scarce. It is preferable, at least for now, to regard the trends in wages, effort, autonomy, security, and environmental health of jobs as evidence in themselves of changes in the quality of work life.
From Quality of Work Life to "Quality in Work"?
Clear thinking about what factors constitute for a good job is a natural prerequisite for understanding the changes affecting many millions of working lives in the present day. From that understanding can come more enlightened policy-making and strategies for all governments and agents seeking to make a difference. There is increasing interest in improving job quality, so it is not surprising that some governments and supranational organizations have also stepped in with their explicit or implicit conceptions. It is useful to compare two official conceptions of job quality with each other and with the approach used in this book.
The International Labor Office (ILO), which attempts to influence work through the agreement of standards and basic rights for workers across the world, has its own laudable concept of "decent work":
The goal of decent work is best expressed through the eyes of people. It is about your job and future prospects; about your working conditions; about balancing work and family life, putting your kids through school or getting them out of child labor. It is about gender equality, equal recognition, and enabling women to make choices and take control of their lives. It is about personal abilities to compete in the marketplace, keep up with new technological skills and remain healthy. It is about developing your entrepreneurial skills, about receiving a fair share of the wealth that you have helped to create and not being discriminated against; it is about having a voice in your work-place and your community . . . For everybody, decent work is about securing human dignity." (ILO, 2001: 7-8)
These criteria of decent work overlap with those to be used in this book, but are broader. Though some aspects would be hard to measure, at least the overall concept is unambiguous because it is, as it should be, worker-centered. The ILO's problems start not with its vision but with the difficulties of addressing the "decent work deficit" around the world (Bruton and Fairris, 1999), especially in developing countries.
Another supranational organization, the European Union (EU), also expresses the desire to influence job quality. It may have more power than the ILO to put policies into effect, but this is within the EU's own member countries, whose economies for the most part are very much richer than many of the economies under the remit of the ILO. Unfortunately, the EU's conceptualization of job quality is clouded by being subsumed under the objective of raising the competitiveness of the European economy. This confusion limits the extent to which EU-driven policy initiatives will be implemented in member countries to improve the quality of work life. Neverthelessthe EU's explicit attention to job quality has highlighted the need for improvement in several aspects that are in the interests of both employers and workers, including skills and workplace safety.
Improving job quality became an explicit objective of the European Union at the Lisbon Summit of the European Commission in March 2000, and the agenda was developed further at subsequent European Council meetings in Stockholm and Nice (European Commission, 2001a, 2001b, 2002). Concern with "quality" was said to be at the heart of the European Social Model, and, while it had already driven the development of social policymaking, employment policies had hitherto largely been shaped by the objective of reducing social exclusion, which largely meant maximizing employment opportunities. After the Lisbon summit, however, policy evaluation, at least at the European level, was to be driven by the idea that what was needed was not only more jobs, but better jobs. The question, however, was how to define "better jobs." The Commission has therefore developed its own concept of "quality in work," which can be contrasted with the approach in this book.
The Commission sees quality as "taking into account" objective characteristics of the job, subjective views of workers, worker characteristics, and the match between the worker and the job. This approach naturally required more focus. With no guidelines as to what could be an appropriate single index of work quality, the Commission drew up a list of job characteristics, classified according to whether they were intrinsic features of the job or features of the work environment. This classification, it is claimed, gave the list of proposed indicators a "coherent framework and approach." It is stressed also that the associated policy framework should form a coherent whole. Considerable emphasis is placed on the claim that policies to raise the quality of jobs are consistent with the aim of increasing the employability of workers. In addition, the chosen features are said to provide a coherent assessment of the progress toward policy conformity with the goals of the Lisbon summit (European Commission, 2001a: 4). "Quality in work" is set to be monitored regularly, and the results used as part of the evaluation of the development of employment policies in member countries and at the EU level. This objective placed a significant constraint on the choice of quality indicators, in that the measures have to be available for all member countries.
The general features to be monitored are listed in table 1.1. Many of the features in table 1.1 have an obvious place in social and employment policies. In fact, the list appears to be largely driven by pre-existing policy objectives of the European Commission, rather than any explicit reference to social or economic theories, and the drive for better jobs thus has the appearance of a repackaging exercise. Moreover, this official interpretation of the concept of job quality shows an ambiguity that is nicely captured in the phrase "quality in work" which appears to merge in one concept the interests of workers and their employers. The tension in the concept shows itself both in what is left out and in what is included. Even though one of the sources of evidence on work intensification in European countries is the European Foundation for Living and Working Conditions (an EU research institution), indicators of work effort are not on the list. This exclusion cannot be because of a distaste for subjective measures, since job satisfaction is proposed as a key indicator of intrinsic job quality. One suspects that work effort is less easy to include because of the built-in ambiguity in the Commission's concept of "quality in work": effort is conflictual and contested, and indicators of effort could imply opposite values for employers and employees. Similarly, one wonders why average wages should be excluded from the list: the reason, again, may be the embarrassment surrounding whether high wages are to be applauded or deplored. Pay is only included indirectly as an aspect of gender equality and as contributing to an indicator on mobility (which is itself part of "intrinsic job quality").
So much for omissions. The list also includes measures that are business oriented, and not directly part of the quality of workers' lives. As is typical among official analyses the world over, potential conflicts of interest between workers and employers are underplayed. Thus, it is asserted that "quality can, and must, go hand in hand with improving efficiency, especially as far as public finances and labour market incentives are concerned" (European Commission, 2001a: 5). De-emphasizing the underlying potential for conflict may be essential, if "quality in work" objectives are to be made compatible with the objective of raising competitiveness in the "knowledge economy." In the Commission's analysis, high quality in work is indicated by good performance, for example through productivity indicators (see item (x) in table 1.1). However, this hardly rings true for the worker who is stressed and exhausted as a result of performing so well; and the flexible labor market that is deemed profitable for employers may not be conducive to security and peace of mind on the part of employees obliged to take temporary and otherwise insecure employment.14 Such tensions are apt to test the neutrality of many analyses by national or supranational governments.
How to Measure a Good Job: Surveys of the Quality of Work Life
What are the opportunities and constraints of a survey-based approach to work quality measurement? The quality of work life is not easily measured. The concept is complex and lacks an obvious unitary index. Essential components of work quality are in practice measurable primarily through the reports of workers. Any comprehensive measure is thus affected by the potential limitations (e.g., social esteem bias), as well as the advantages (e.g., first-hand knowledge) of subjective data. Nevertheless, the raison d'être for this book is that acceptable quantitative measures are now available in a number of countries. In the last thirty years, surveys have evolved into a widely used instrument of social enquiry. Hitherto, representative large-scale survey methods have been used less frequently for quality of work life issues than for other areas of social life. Survey methods can be inferior to, say, ethnographic methods, for the purpose of uncovering processes, ideologies, and conflicts at the workplace. But, for their generalizability to large populations of interest, their ability to uncover structural changes, and their transparency, survey methods are indispensable.15
The quality of the survey, however, is crucial. Many are compromised by poor design, insufficient finance, undue selectivity in their choice of sample population, or overdependence on survey sponsors' nonresearch objectives. Unfortunately, surveys of work issues are no exception, perhaps because interests are especially prone to intrude in this most delicate and central of areas. Also common is the loose use of statistics on work in popular discourse. The general principles that drive the choice of surveys of work used in this book are:
- Sampling methods need to be of sufficiently high quality. This criterion normally entails the use of some form of random sampling procedure.
- The achieved sample should be representative of a nation's workforce, or at least a subgroup forming the large majority.
- There needs to be a good response rate--if possible, explicitly reported.
- The questionnaire needs to contain well-designed instruments for measuring aspects of work quality.
- The surveys need to be able to address the issue of change, preferably through comparison within a series of surveys, or if not, through valid change questions within a single cross-section.
The surveys drawn upon directly in this book are described in the Data Set Appendix. Inevitably, their scope is selective, in terms of both geography and subject matter. This is not a comprehensive account of all aspects of work life across the industrialized world, and does not touch on work life in poor countries. Perhaps such a picture could be drawn up in the future, if the means and methods of social and economic analysis improve as much in the next thirty years as they have in the past. The more modest aim here is to arrive at a picture and an understanding of changes in work life from the workers' perspective, across a reasonable range of countries where the data allow.
Return to Book Description
File created: 8/7/2007
Questions and comments to: email@example.com
Princeton University Press