Job-satisfaction in the Broader Framework of the Capability Approach **

Being one of the leading paradigms for analyzing human well-being, the Capability Approach (CA) may serve as a frame of reference for job-satisfaction. The CA mainly argues for (1) complementing information on job-satisfaction by the more objective conception of “valuable work” to avoid biases due to adaptive preferences and (2) taking a holistic perspective of the multidimensional work-situation while (3) emphasizing the importance of opportunities and individual choice and agency. Thereby the CA provides a new perspective on classical insights on job-satisfaction by putting them into a broader framework. This paper is but a first tentative appreciation of the CA in this context.


Introduction
The Capability Approach (CA), developed by Amartya Sen and Martha Nussbaum, is one of the leading paradigms for the analysis of a person's well-being.It defines a person's well-being in terms of the beings and doings (the functionings) a person achieves and her capability to choose among different combinations of such functionings.Thus, the CA is not specifically concerned with job-satisfaction, but provides a broader frame of reference in which job-satisfaction can be considered.Mainly, this results in a change of the perspective on job-satisfaction.The CA can help solving the puzzles put forward by empirical findings on job-satisfaction (discrepancy between objective and subjective conditions etc.) and offers a new basis for theoretical considerations in this undertheorized but empirically well explored field (Büssing, 2003, p. 462).
This paper shows how the CA sheds new light on some insights on jobsatisfaction gained so far and how it may be used for systematizing them.It does so in five steps: First, the CA is briefly introduced and the change in scope of analysis concerning job-satisfaction is sketched.Secondly, because of adaptive preferences, the CA is critical of subjective measures of well-being including job-satisfaction.The CA does not discard subjective measures altogether, but rather argues for interpreting the data carefully.Therefore, job-satisfaction should be, thirdly, complemented by a concept of valuable work.From a capability perspective the quality of work should not be seen in isolation, but as the "capability for work" in an integrated manner that takes personal and social factors into account.This changes the interpretation of existing indicators of quality of work.Fourthly, the issue of multidimensionality is discussed.This involves choosing dimensions and exploring their interdependency.It is impossible to provide a standard definition of the dimensions of "capability for work", but the CA argues for viewing it as an irreducible pluralistic conception that includes subjective and objective dimensions.Fifthly, in emphasizing freedom of choice, the CA highlights the importance of participation and process freedom for well-being in general and for "capability for work" in particular.If the diversity of people is taken seriously, their views on what constitutes well-being have to be heard.Thus, as a normative consequence of the CA it is essential that people get the opportunity to develop their views and voice them.The final section concludes that due to the complex relationship between work and well-being the article cannot provide but a first tentative appreciation of the CA's potential in this context.

A brief introduction in the Capability Approach
The basic concepts of the CA are functionings and capability, the distinction between achievements and freedoms and the distinction between well-being and agency which is related to that between the opportunity aspect of freedom and the process aspect.These concepts are introduced in the following.
Functionings are the "various 'doings' and 'beings' a person achieves" (Sen, 1987, p. 29).They are constitutive of the person's well-being and can be used to describe her life situation.They range from elementary ones like "being nourished" to rather complex ones like being able to appear in public without shame (Sen, 1999, p. 75).The availability of resources, e.g.bread, is a necessary prerequisite to achieve functionings, e.g.being nourished, but personal features such as sex, height, weight and metabolic rate as well as social, cultural and environmental conditions such as eating habits and food conservation methods influence a person's ability to convert the resource bread into the functioning of being nourished.These conditions are summarized as personal and social conversion factors (see figure 1).The list of functionings that are regarded as relevant in a particular context 1 "reflects a view of what is valuable" (Sen 1987, p. 29) or in short of what constitutes a good life.The selected functionings then form the evaluative space for analyzing well-being in this context and any combination of them represents a way of life.Thus, the CA proposes functionings as the focal variables for evaluation without, however, fixing a weighting scheme or attaching otherwise a value to each combination of functionings such as a "utility" value.
The various combinations of functionings which are feasible for a person can be described as the various ways of life open to her.The set of all feasible options is called "capability set".From this set the person can choose but one combination.Thus, the capabilities of a person indicate her freedom of choice to lead a life she values.To put it differently: There are several combinations of functionings feasible for the person which constitute her well-being freedom, the one option she chooses is her well-being achievement.
Sen concentrates in his presentation of the CA on the opportunity aspect of freedom by modeling freedom of choice as an opportunity set (the capability set).This is an important aspect of freedom which has often been neglected by those who mainly think of the process aspect of freedom.The latter refers to the procedures of achieving a certain state.Sen (1982, pp. 216-217) gives the example of a person who likes to read on a Sunday morning and enjoys this activity (situation A) when someone enters her home and coerces her to continue reading (situation B).Sen contrasts this situation with yet another one when the person is coerced to go into the sewage system (situation C).The process aspect stays the same in B and C when the person is coerced to do something.What matters is the fact of coercion while it doesn't matter that in B the person is coerced to do what she likes to do anyway and in C she is coerced to do a very unpleasant task.Focusing on the opportunity aspect allows distinguishing between B and C: While in both cases the opportunity set shrunk from several options in A to only one option because of coercion, the value of the option(s) available is higher in B than in C.
Though the capability set puts more emphasis on the opportunity aspect of freedom it captures the process aspect as well by distinguishing between having a choice and having no choice at all.Choice is seen as an important ingredient of freedom and therefore the person is always seen as an agent and not a patient.Yet, the distinction between agency and well-being that Sen draws refers to something else: Sen (1985b, p. 186) argues that people pursue goals other than well-being and sometimes their activities do not contribute to their well-being.Examples of such goals are "the independence of her country, or the prosperity of her community" (Sen, 1992, p. 56).In gen-eral, the very idea of commitment catches the essence of goals not primarily directed towards one's own well-being.Agency goals refer to social behavior such as fulfilling obligations, behaving in line with an identity shared with others or following social norms and rules.Work motivation is an example as Sen (1977, p. 97) notices: "Every economic system has ... tended to rely on the existence of attitudes towards work which supersede the calculation of net gain from each unit of exertion."In reverse, Sen reserves the notion of well-being to the direct pursuit of personal welfare even if his notion is not restricted to mere self-interest like in utilitarian welfare economics.As long as a person aims at improving her own situation or that of people she cares for, her goals fall in the category of well-being goals.Thus Sen (1985b) arrives at four categories: (1) well-being achievements, (2) well-being freedom, (3) agency achievements and (4) agency freedom.As noted above, agency is closely connected to our morals, to commitments to certain (political) goals.Agency freedom includes political liberties.Freedom is valuable intrinsically as well as instrumentally.The instrumental value derives from being able to achieve one's goals.The intrinsic value of freedom is linked to autonomy and is the reason why in Sen's view both well-being and agency freedom have a positive impact on well-being: "For an integrated person it is likelypossibly even inevitable -that the person's well-being will be influenced by his or her agency."(Sen 1985c, p.187).This is true even though agency and well-being goals sometimes conflict.
Using the CA for analyzing job-satisfaction broadens the scope of analysis in several ways.First of all, aspects of life beyond the characteristics of the workplace enter the consideration.As Sen (1987, p. 12) points out it is important "to take an integrated view of the pattern of activities outside and inside the home that together make up the production processes."Therefore, labor force data must include questions on the household and its members in order to analyze employment from a capability perspective (Lugo, 2007).Additionally, knowledge on prevailing institutional arrangements concerning cared-for family members such as children and elder people is crucial for understanding the job-satisfaction of those who have to care for them.Of course, recent research on quality of life and job-satisfaction has already dealt with these issues (Cassar, 2010 provides an overview), but the CA goes beyond this research in some ways (see next section).
Secondly, employment in the informal sector is often neglected in standard approaches to job-satisfaction (Cassar, 2010).Though it may seem that this mainly raises difficulties for investigating job-satisfaction in developing countries where the share of employment in the informal sector is larger than in developed countries, the argument holds for developed countries as well since much of the "reproductive" work such as care and meals is provided by the informal sector or at home.It is difficult to ask people about their job-satisfaction if they do not have proper jobs, but it is possible to analyze their well-being in terms of capabilities even then.As explained above the capability derives from personal and social conversion factors.Thus, the CA distinguishes between factors that the person can affect and those she has to take as given by society.
Thirdly, the CA views the person as an agent who has goals beyond her own wellbeing.When looking at work motivation Sen (1977, pp. 97-98) discovered that the standard economic model of self-interested individuals cannot fully explain the commitment to work.Partly, Sen (1975, p. 5) has already covered this idea by distinguishing three aspects of employment: the income aspect, the production aspect and the recognition aspect which he described as following: "employment gives the person the recognition of being engaged in something worth his while".The CA captures this by allowing for goals other than personal well-being and emphasizing the importance of freedom of choice.It thereby avoids the traditional economic divide in production and consumption where the positive impact of work on well-being cannot be analyzed but with the help of additional assumptions.

Job-satisfaction from a Capability Perspective
Classical accounts of job-satisfaction rely on the declared level of workers' satisfaction.In this perspective, well-being is measured as the satisfaction with working conditions, timetables, wages, health and safety at work, opportunities for participation, etc. for instance on a scale from 0 to 10 as in the European Social Survey, where 0 represents "extremely dissatisfied" and 10 represents "extremely satisfied".Hence, objective job characteristics and objective measures of well-being are superseded by subjective appreciations considered to be representative of these objective conditions.This has certainly many advantages, such as using a single and simple index to measure a complex and multidimensional reality.It allows taking into account tastes or desires instead of imposing a uniform and unique view of what is a valuable job, thus making space for the diversity of human beings' wishes and expectations.For instance, working at weekends might be very convenient for students looking for some extra money, while it certainly makes the reconciliation of work and family life for many others more difficult.
However, such a view is flawed in many respects as is illustrated by the fact that data from the International Social Survey Program of 2005, relying on such subjective methodologies, indicate that job-satisfaction is highest in Mexico, much higher than in Scandinavian countries, West Germany or France (Munoz de Bastillo, 2009).As a matter of fact, the data on satisfaction or subjective well-being do not necessarily correlate with objective indicators of job quality such as wage, workload or job stability and are therefore difficult to interpret (Clark, 2009).Thus, they do not reliably tell us how well people are: The empirical work on job-satisfaction shows that it is difficult to distinguish between "true" satisfaction (subjective well-being) and resignation where satisfaction is derived from adapted preferences (Matiaske & Mellewigt, 2001), i.e. people adapt their preferences to the opportunities available in the societal context in which they are embedded.Hence, Mexican workers are, at least to some extent, more satisfied due to their capacity to adjust their preferences to the available opportunities rather than to the objective quality of these opportunities.As a matter of fact, the complex and often ambivalent link between subjective assessments and objective conditions demands to look for other ways of measuring job quality than the declared level of workers' satisfaction.This does not imply that subjective measures should be discarded altogether, but that they ought to be completed by objective measurements and integrated with them.
The necessity to combine subjective and objective measures is widely recognized in the literature.Such is also the perspective adopted by the CA, but the CA questions the role of information on happiness or satisfaction in general.Sen (1985a, p. 12, p. 21) warns against confusing valuing a life and measuring the happiness it generates.He holds that both the "physical conditions neglect" and the "valuation neglect" are important here.The first refers to the fact that people living in very poor circumstances are often found to say that they are happy.This can be explained by adaptive preferences: People adjust their expectations to their circumstances and learn "to take pleasure in small mercies" (Sen 1985a, p. 21).The second, the valuation neglect, can best be understood by distinguishing between these two sentences "(I) I value x and so I desire it and (II) I desire x and so I value it."While Sen (1985a, p. 32) agrees with the first and thus takes a desire for x or the happiness about x as evidence for the value of x, he opposes the second claim and points out that desire, happiness and satisfaction describe an emotional reaction whereas valuing denotes an exercise in practical reasoning (for a similar distinction with regard to job-satisfaction see Martin, 2003, p. 12).Hence, the CA is not opposed to using information on satisfaction in general, but careful about the interpretation and tends to see it as one dimension among many.The CA also emphasizes that preferences need to be taken into account when defining what a valuable job is.

"Valuable work" as a complement to job-satisfaction
The conception presented here is based on the key notion of "capability for work", that can be defined as the "real freedom to have the job one has reason to value".This is used as a yardstick to assess the quality of jobs and the "relevance" of declared job-satisfaction.When operationalizing this notion, it is important to capture "possible worlds" (De Munck & Orianne, 2008) rather than presenting only data about actual facts.
Two main challenges need to be faced in this perspective.First, identify the main components of "capability for work" and the way they are articulated together.Second, frame these components in terms of capabilities and not only functionings.A scheme, adapted from Ingrid Robeyns (2005), (fig. 1) is helpful in meeting these two conditions.Following Robeyns' scheme, three main conditions are needed to achieve a high level of capabilities, namely resources, personal and social conversion factors.How these can be interpreted in terms of capabilities and not only in terms of outcomes (i.e.actual beings and doings) is illustrated in the next paragraphs.
When investigating job quality, the issue of resources mainly relates to wages and transfer incomes provided via cash benefits.Not only should their amount be taken into account, but also the conditions to which they are submitted, as these significantly impact on job quality and the degree of "capability for work" enjoyed by workers.For instance, if remunerations depend to a large extent on individual performance (e.g.combining some modest degree of statutory wage with a large extent of performance pay), then this might translate into more work intensification and more subordination at work which would imply a restricted "capability for work".Hence, higher wages do not necessarily coincide with a higher degree of "capability for work".This complexity of the issue of resources is illustrated when lower paid workers display a higher degree of satisfaction than their better remunerated colleagues.The "quality" of remuneration matters as much as its "quantity".Also, the capability to refuse a badly remunerated job depends on the level of cash benefits provided by the social security system: if these are very low, then low remunerations can be more easily imposed on workers.An adequate view of resources in the capability perspective also requires taking additional parameters into account such as: the impact of other statuses than wage work (independent or informal activities) on the amount and regularity of resources and the reference base for the calculation of their amount: individual or household level.A low income is more acceptable in a dual earner household than in the conventional male breadwinner family.The calculation of resources at the aggregated level of the household conceals the negative impact of an unequal distribution of resources within the household on the lower paid members of the household.These issues are of course not exhaustive but illustrate the intricacies that need to be considered when envisaging resources in a capability perspective.
Capability for work also depends on adequate personal and social conversion factors.Among the relevant personal conversion factors, there is the issue of skills and competencies (e.g.diplomas, know-how, technical competencies, but also social competencies such as ability to work in teams, capacity to communicate, or ability to balance demands from private and professional life, etc.).When viewed from a capability perspective, this question requires the design of indicators that allow tackling issues such as the following: a) Does the development of competencies promoted at firm level contribute mainly to the competitiveness of the company?Or does it serve to enhance the workers' capabilities or self-fulfillment 2 ?Indeed, the promotion of their adaptability to the company requirements does not necessarily equate to the development of their capabilities.b) To what extent are these competencies transferable from one work environment to another one?What tools are developed to promote such transferability?If no such tools exist, then it means that the employee's capability for work is strictly linked to the firm employing him/her.In contrast, transferable competencies increase to a large extent the workers' capability for work.c) To what extent are all categories of workers (low-and high-qualified, men and women, etc.) entitled to the development of their competencies?Since such a development does not necessarily coincide with the enhancement of "capability for work", indicators are needed that capture issues such as which objective is pursued by training programs (firm's competitiveness and/or workers' self-fulfillment?), the degree of their transferability in 2 The EU research projects Eurocap and Capright have conducted many empirical investigations in this direction (see e.g.Corteel & Zimmermann, 2007).
other work settings (that matters when determining the number of alternative feasible options), the degree of their availability to all categories of workers or, more widely, also to people involved into informal activities.
Social conversion factors also play a key role with regard to capability for work.They comprise such issues as: a) the quantity of available jobs (are there enough jobs for every man or woman wishing to be employed?),b) the accessibility of these jobs to all potential applicants: figures about the employment rates of various categories (men, women, foreigners, etc.) are helpful in this respect, but they need to be completed by data about discriminatory practices on the labor market, the existence of legislative protective provisions and the degree of their enforcement etc.; c) job quality which encompasses two categories of objective dimensions relating to two different views of work: work as a disutility that requires compensation and work as a component of human flourishing that generates utility.When work is envisaged as a disutility it needs to be strictly delimited and adequately compensated for.Indicators such as wages, work duration, stability, impact on health, degree of work intensification, etc. can be used to capture this aspect of job quality, as they allow to measure the level of compensation provided to workers as well as the extent to which work interferes with other life spheres.Workers are here identified as self-interested individuals rather than committed or involved people.A second category of indicators envisages work as a utility or a necessary component of human flourishing.These look at the content of work in terms of self-fulfillment: indicators such as skills, competencies, degree of autonomy and responsibility, hourly productivity, etc. are mobilized.With respect to these, legislative and collective regulations require specific attention.If existing provisions guarantee that minimal conditions are respected in all firms, then looking for another job will not be as risky in terms of working conditions.By contrast, if working conditions depend on corporate actors' goodwill, such a move might be more penalizing for the workers' well-being.Hence, the state of the labor market, the scope of the existing regulations in terms of job quality or anti-discrimination practices, as well as the predominant values (e.g.work ethics, gender division of labor, etc.) are examples of social conversion factors that strongly impact on the degree of "capability for work" enjoyed by workers.
Hence, the CA asks for looking at the various parties that shape individual capabilities: first, the social environment including welfare state policies and established practices, second, the firm which exhibits certain working conditions and offers e.g.training programs, and third, the individual who chooses from her capability set.
The available databases at EU level, such as the European Labor Force Survey, the European Working Conditions Survey and the European Survey of Income and Living Conditions, provide detailed information on most of these parameters, as well as on the jobsatisfaction declared by workers.Existing indicators of job quality (e.g. the Laeken Indicators of Job Quality or the European Job Quality Index) combine objective and subjective data aiming at a comprehensive picture of the issue of job quality (for an extensive review, see Munoz de Bustillo, 2009).However, they tend to focus on functioningsthe actual situation of workers in terms of wages, timetables, skills, working conditions, health, degree of autonomy, etc.By contrast, limited information is provided on capabilities, i.e. on the set of other feasible options open to workers given their resources and their personal and social conversion factors.
However, such information on the capability set is crucial for meaningfully combining objective and subjective indicators.Indeed, for being able to adequately assess the quality of subjective appreciations it is necessary to know something about the available opportunities or the alternative feasible options (as March/Simon 1958 already noted).Empirical studies based on the CA explore two methods for generating knowledge on available opportunities (Lessmann forthcoming): Either information on resources, personal and social conversion factors is used for estimating the capability set (e.g.Burchardt/Le Grand, 2002, Strotmann/Volkert, 2008) or indicators that capture the opportunity aspect are introduced and carefully interpreted (Poggi, 2008).
When it is viewed via the lens of "capability for work", job quality requires taking into account a plurality of dimensions, as well as the ways these are intermingled.This issue is tackled in the next section.

Selection and interdependence of dimensions
As the discussion of resources and conversion factors in the previous section shows, they are interlaced in many ways and it is often not possible to classify the conditions in a uniform way as either resources or conversion factors.By conceiving the capability set as resulting from the interaction of both resources and conversion factors, the CA reminds us that neither material conditions -including commodities or goodsnor institutions and skills are sufficient for ensuring a good life.It takes the concurrence of both.
In this paper we will not provide a list of functionings relevant for evaluating the capability for work or the more specific issue of job-satisfaction in all contexts.Rather we agree with Sen that relevant dimensions should be chosen according to the specific context at hand.What should be clear, however, is that apart from rather specific functionings such as doing well what one does (Aristotle's perfectionism, see Nussbaum, 1995, p. 112), the whole set of basic functionings like being nourished and having shelter as well as other complex functionings such as taking part in the life of the community, having self-respect, and living together in a family play an important role in the evaluation of work.This is difficult to implement.Schokkaert and van Ootegem (1990) therefore select dimensions by using factor analysis in their study on the living standard of Belgian unemployed.They interpreted the latent variables they found as the following six dimensions: social functioning, psychological functioning, physical functioning, microsocial contact, activity and financial functioning.Similarly, Poggi (2008, p. 8) arrived at her list of 13 dimensions of working conditions by using factor analysis.She investigates how these working conditions affect job-satisfaction via expectations.
In these examples the selection of dimensions depends on the aim of the study as well as on data availability.OPHI (Oxford Poverty and Human development Initiative) works on improving the latter (especially with regard to analysing poverty).OPHI views job quality as a missing dimension in data bases.Lugo (2007) identifies information on the following topics as relevant for filling this gap: protection (informal employment), income, safety (occupational hazard), time (under-and overem-ployment, multiple activities), discouraged employment.She developed a proposal of indicators for this missing dimension.In the case of income, for example, she suggests questions to determine income from self-employment; questions on informal employment concern the legal status of the firm as well as having a contract and being entitled to paid holidays, paid sick leave, social security benefits, medical care and maternal leave.However, work-related aspects of life are but a subset of all functionings and sometimes other features of the available capability set may be crucial for the perception of one's employment and the resulting job-satisfaction.
Apart from choosing dimensions, any multidimensional approach must also define how they relate to each other.Sen's CA does this by looking at all feasible combinations of functionings. 3The person then chooses one of these multidimensional options -one combination of functionings or way of life.This reflects the kind of choice we face in our everyday life: We cannot choose any level of income we want, but a job that exhibits a specific combination of income and working hours and distance of the workplace from home and so on.If the highest income can be achieved only at the expense of long working hours far away from home, the job might be best in terms of income and yet another job that offers a combination of moderate income, moderate working hours and a decent level of commute might be preferred to it.
Any evaluation that takes combinations of dimensions into account will often exhibit incomparability of options.A job with a high level of income in combination with long working hours and a long commute is hard to compare to a job with a lower level of income, shorter working hours and a shorter distance to home.Sen (1992Sen ( , p. 134, 1999, p. 77) , p. 77) has pointed out several times that the incompleteness resulting from this incomparability need not be seen as embarrassing since it reflects substantial ambiguities of the underlying concept of well-being.However, the difficulties of selecting relevant functionings and evaluating combinations of functionings constitute a huge challenge to any operationalisation of the CA.For example, Cassar (2010) has shown that the influence of autonomy on job-satisfaction in Chile4 is balanced by the preference for protection.Studies that do not take both -the preference for protection and for autonomy -into account will only tell half of the story.They may be able to provide a more complete ordering of various situations, but only on the expense of missing an important argument.Thus, because achievements in one dimension cannot be evaluated independently from achievements in others, the overall evaluation hinges upon the selection of relevant functionings.
As argued above, the CA questions the role of information on happiness or satisfaction in general.With respect to the more specific information on job-satisfaction the CA sees it along the same lines as satisfaction in general as giving evidence about how much people value their job.The issue about interdependency with other dimensions in this case is, however, even more pressing since asking about job-satisfaction implicitly restricts the focus to the job and asks the respondent to separate between the job and other spheres of life.In fact, comprehensive information on subjective well-being is seen as another missing dimension by OPHI.Samman (2007) has therefore developed a proposal of indicators for subjective well-being composed of various questions suggested and tested by others (e.g.Cummins, 1996;Ryan & Deci, 2000).Cassar (2010) uses Chilean data based on both the indicators of job quality proposed by Lugo and the differentiated questions on subjective well-being proposed by Samman for investigating the role of autonomy for job-satisfaction as a part of general well-being in terms of capabilities.Her study thus touches upon a peculiar interdependence, namely the interdependence of agency freedom (autonomy) and well-being.As already pointed out in the introduction Sen reserves the latter term for the pursuit of personal welfare in a narrow sense and introduces the term "agency" to allow for the pursuit of other goals "closely connected with one's morals" (Sen 1977, p. 93).The relation is, however, not so simple.Agency goals, such as fulfilling obligations or acting in line with social norms and rules can conflict with personal well-being goals.For instance, taking a sick leave is often frowned upon for social reasons e.g. because this will add to the pressure on the colleagues in the team.Obeying the social expectations means to disregard one's own needs in terms of health.Thus, though agency freedom affects personal well-being positively it also necessitates making hard choices and occasionally diminishes personal well-being.
Interdependency of dimensions originates from the interaction of resources and conversion factors.Hence, evaluating the work opportunities open to a person demands considering which among all these dimensions are relevant and how they relate to each other.This discussion should not be left to researchers alone.The CA attaches importance to the opportunity to live a life one values.Hence, people should be enabled to think about their values and speak for them as the next section explicates.

Participation and process freedom as a prerequisite for "valuable work"
Process freedom designates people's real freedom to participate in public debates, i.e. express their viewpoints, expectations, wishes, etc. and make them count when it comes to making collective decisions.This does not imply that their points of view will prevail in the end, but that they must be taken into account during the decisionmaking process.The prerequisites for process freedom, what we can also label as "capability for voice" (Bonvin, 2008), include: a) the availability of adequate cognitive and political resources to all stakeholders involved in the process of collective decisionmaking, i.e. the ability to clearly express their viewpoint and efficiently defend it; b) an environment where such viewpoints can legitimately be expressed and pushed on the collective agenda, i.e. a context in which existing rules and procedures are framed in such a way to allow the recognition of a diversity of relevant informational bases of judgment in justice and opinions about them; c) the real freedom to take part in collective decision-making, which implies the ability not to actively participate in such public debates without having to incur unbearable penalties (without this, process freedom would coincide with a duty to take part in collective decision-making processes).Such a possibility to take part in public debates is a key component of the capability set in Sen's perspective (Sen, 1999).The concern for democratic delibera-tion and public debate is indeed crucial in the perspective of the CA, and many scholars have focused their attention on this issue (e.g.Bohman, 1996;Sen, 2009).
With respect to "valuable work", taking process freedom seriously requires to investigate the following dimensions.Firstly, the cognitive resources that people can mobilize to defend their point of view about "valuable work", e.g.their negotiating skills, their ability to communicate or to argue, are key in this respect.If workers are not competent enough to efficiently push their views about "valuable work", then they should resort to representatives.This implies that adequate attention be paid to the issues connected to representation.Indeed, if trade unionists defend the ideas of their bureaucracy and do not pay enough attention to the wishes expressed by workers, then the informational basis of judgment in justice will be biased, and process freedom in terms of participation to the collective decision-making process will be restricted.
Also political resources matter, i.e. the extent of influence workers and other stakeholders can have on issues related to labor market regulation and processes of definition of "valuable work".This can relate, for example, to the recognition of their legitimacy during the collective bargaining process.If they are excluded from the table of negotiations, this implies that their process freedom, if any, is merely formal, as it is not allowed to flourish when and where collective decisions are actually made.This is e.g. the case of excluded trade unions in so-called "closed shop" or "union shop" systems (as promoted by the Taft-Hartley Act in the United States), where minority trade unionists are not allowed to take part in collective bargaining.More generally, this relates to the issue of the representation of non-members of trade unions, which particularly matters in countries where trade unions are very close to political ruling bodies.But it also encompasses the issue of the tools that the various actors involved can legally mobilize while bargaining or deliberating about the most appropriate regulation of the labor market or the most adequate definition of "valuable work".So-called procedural rights, that do not aim to give any substantial rights to workers (e.g. in terms of level of wages or other working conditions) but to allow them to participate or somehow be involved into decision-making processes, are key in this respect.For instance, the right to be consulted or informed in case of collective dismissals, or the right of co-decision that can be found e.g. in Germany, certainly extends the process freedom of workers' representatives.In the same perspective, one could argue that the right to strike is an important political resource that increases workers' bargaining power and, as a consequence, process freedom.
This issue of political resources is tightly linked with the institutional ability or willingness to take into account other people's viewpoints when making collective decisions.Indeed, the selection of the adequate informational basis of judgment in justice when it comes to defining "valuable work" strongly depends on the actors that are allowed to participate in this selection process.If only employers are considered as relevant partners, then the relevant information will be selected along their concerns, and the process freedom of all other actors will be reduced as a consequence.If on the contrary only workers are entitled to have their say about what is "valuable work", then some information will not be taken into account.In Sen's perspective, process freedom requires as many stakeholders and informational bases of judgment in justice as possible to be involved in the collective decision-making process.In other words, collective bargaining or any process concerning the definition of "valuable work" should integrate as many actors and informational bases as possible.Therefore, the rules and procedures that govern collective decision-making processes about "valuable work" should be assessed against their ability to integrate such a plurality of actors, information and viewpoints, that is interpreted as the prerequisite for the flourishing of public reasoning.
All these conditions do not imply a specific substantial content of "valuable work", rather it insists on fair and capability-friendly collective decision-making processes.Hence, process freedom features as a necessary complement to opportunity freedom, as there is no one-size-fit-all definition of "valuable work", but only contextualized or situated notions of "valuable work".
Beside the collective level (i.e.participatory rights at collective decision-making processes about "valuable work"), a second level matters with regard to process freedom, namely the individual level.Indeed, the ultimate choice about "valuable work" lies in the individuals' hands: the collective framework determines the general conditions, but this still leaves (or at least should leave) some space for individual decisionmaking about the definition of "valuable work".This in particular matters for all decisions connecting work with other spheres of life, as it would be impossible to determine once and for all the significance of work in one's life; it also relates to the connection between well-being and agency at work, insofar as one's expectations with regard to work might considerably differ, and legitimately so.Therefore, a key issue in this respect is that the collective framework agreed upon (possibly along the lines identified above) should not infringe upon personal preferences, but allow them to flourish as much as possible along the perspective developed in Section 3 about "Choosing and Interdependence".
Hence, process freedom requires to take into account these two aspects: a) to what extent are collective definitions about "valuable work" made in accordance with the notion of process freedom; b) do these collective decisions leave enough space for individual choice?Or do they, on the contrary, constrain these individual choices, thus imposing the burden of "valuable work" as it has been collectively defined on possibly reluctant individuals?Indeed, process freedom in the capability perspective implies an adequate combination between collective rationality (along processes of democratic deliberation and bargaining) and individual rationalities allowed to flourish as much as possible.
For applying the CA to the issue of process freedom, quantitative analysis often reaches its limits and it is necessary to combine it with qualitative analysis (Zimmermann, 2006).Hence, field investigations at sector or at firm level ought to complement the available quantitative data about objective working conditions and their subjective appreciations by workers.In our view, the combination of these methods allows to come as close as possible to the identification of the workers' capability set and, therefore, to test the relevance of their subjective appreciations of job quality.

Conclusion
The CA views work as an important part of life and asks for its contribution to wellbeing.Valuable work is a constituent of living a life one values.It entails jobsatisfaction but extends beyond it.First of all, the CA objects to relying exclusively on the subjective evaluation of well-being because of adaptive preferences.If people display satisfaction due to adapted preferences, this is a sign of resignation.Therefore the CA argues, secondly, for complementing information on job-satisfaction by objective data on the life-situation as a whole, i.e. on resources, personal and social conversion factors, to arrive at an assessment of valuable work.The CA is, thirdly, a multidimensional approach that takes the various doings and beings a person can achieve into account and warns against restricting the perspective on too few dimensions.The relevance of dimensions depends on the context at hand.Hence, selecting dimensions is a challenge in operationalizing the CA, but many empirical studies show how this challenge can be met.The cogency of its evaluative space derives from the interest in people's well-being -their being and doing well.
In fact, well-being in terms of capabilities even goes beyond this: It encompasses the opportunity to choose among several ways of life.Freedom of choice contributes positively to well-being since human beings are agents not patients.However, people can exercise their agency only if processes are designed that enable their agency.The link is especially obvious in the case of labor relations because valuable work is the result of collective efforts to create conditions for human flourishing.In consequence the CA advocates participation and democratic procedures for arriving at a life one values.
In this article we do not provide a full-fledged conception of valuable work or a complete reassessment of job-satisfaction from a capability perspective.Our aim is to introduce the Capability Approach as a frame of reference for analyzing jobsatisfaction and to hint at the change in perspective and the resulting potential for new insights that could be gained by looking at issues such as informal employment, reproductive work and the consequences of legal regulations.In order to tap the full potential there is, however, a lot still to be done.Applying the CA empirically is a challenge for several reasons: Multidimensionality is challenging in itself and selecting dimensions is the first task for operationalizing the approach.The capability set is unobservable.Several routes have been followed to capture the opportunity aspect of capabilities, but there is no golden rule for estimating the capability set.Often existing data do not provide enough information.Further, the process aspect and the level of participation cannot be assessed by relying exclusively on quantitative methods.Therefore, the application of the CA to the labor market demands combining quantitative and qualitative issues.
In contrast to many other approaches to job-satisfaction, the CA does not aim at improving the productivity of the worker.The CA is interested in job-satisfaction as far as it is an indicator for well-being and human flourishing in general and for the functioning of working in particular.Yet, there may be some lessons to be learned from that perspective for firms and for the welfare state as well.

Figure 1 :
Figure 1: From resources and commodities to achieved functionings