Introduction / Summay

In previous contributions, there should have been wide discussions to draw the limits of leisure. I would like to get assistance one that you have read those chapters, in order to write some introductory paragraph that lets the reader know if my contribution is particularly related to some other one. In this contribution, I will analyze it by using the subjective well-being approach. I will use the "leisure experience" dimension (as described bellow), and I will study the determinants of subjective well-being. There will be plenty of conceptual discussion, some regularities will be reported, an empirical exercise will be performed - results analyzed, and some insights for future research will be presented.

In this paper, we will study leisure and its beneficial aspects over individual welfare by using a quite new approach: the subjective well-being or happiness approach to measure individual welfare. Along the discussion, we will present two main points. First, we will discuss on the dimensions of leisure in order to justify that by using subjective well-being procedures, we can get a comprehensive approximation to the, somehow difficult to measure, leisure concept. Second, to determine which are the personal and environmental factors that are needed so an individual can produce and consume enjoyable leisure experiences.

In this chapter, we are not going to consider that leisure is just free time, i.e. time that is not dedicated to market work, nor to household maintenance activities. We are not even going to consider that leisure is discretionary time (Goodin, et al. 2005). What we state is that leisure is a universal human need that has to be fulfilled by the production in the household and the personal consumption of what we may call "leisure experiences". Each experience is a commodity that enters directly in the individual's utility function. This means that leisure is one of the arguments of the utility function of the individual, one of the instances from which she will get welfare. By doing this, we will adopt from the beginning a beckerian approach (Becker, 1965, and 1990). Other arguments are (Gronau and Hamermesh, 2006). From that list of commodities, we can agree that leisure is the most time intensive one.

Individuals have this particular basic need, leisure, to be fulfilled using the most suitable combination of personal resources. As always, we are living in a scarce world where every input has some alternative use, so people have to make allocation choices about the best way to fulfill this leisure need as well as others such as food, shelter, and so on. However, we will introduce into our analysis a basic feature of leisure: the presence of enjoyable others. Only recently has this aspect been introduced in the economic analysis of leisure (Osberg, 2009). In this paper, we address the question of how personal inputs are optimally combined to satisfy the leisure need in a social context. By means the analysis of the leisure domain satisfaction, we will be able to asses how personal free time transforms into leisure and how this outcome contributes to individual welfare.

Each person would define the boundaries of leisure on the basis of her tastes, on different resource availability to fulfill her needs, and may value the final outcome in many different ways depending on the social norms, her personal aspirations, social interactions and past experiences. Since using a personal definition of leisure would make any analysis impossible, we will present the main three different constructions of leisure, as proposed by Kelly (1982). The first approach of leisure is the most basic one that defines leisure as quantifiable leisure time, either residual or discretional, based on the freedom to choose. The second one defines leisure as the activity that is chosen at a given time and place so that it is the quality of the activity which defines it as leisure. The third one defines leisure as a subjective condition on the grounds of a freely chosen experience based on intrinsic motivation. The integrative approach proposed by Kelly is the one that we follow in this research, where "Leisure is an action that takes place at a given time, develops an identifiable activity and is perceived as a pleasant experience by the actor".

In what follows, we would refer to this last integrative approach either as leisure or leisure experience. Actually, it fits very well with the following definition of leisure satisfaction by Beard and Ragheb (1980). For them, leisure satisfaction is the "...positive perceptions or feelings that an individual forms, elicits, or gains as a result of engaging in leisure activities and choices. It is the degree to which one is presently content or pleased with her general leisure experiences and situations. This positive feeling of pleasure results from the satisfaction of felt or unfelt needs of the individual".

Traditional economic theory studies human behavior by means of individual’s observed choices. In such a spirit, observed time allocation can be an outcome of interest recorded on time-use surveys. Actually, as we will discuss in the concluding section, time-use registers are a very valuable source of information, and many of the questions that we are going to address could be complementarily studied by testing those hypotheses with that type of data. However, even if some authors consider that time is the ultimate source of utility, time by itself provides no utility to individuals, since the mere passing of time does not fulfill any human need (possibly except from sleeping time). Moreover, since we have no means of observing the final leisure output, we have to rely on the subjective assessment of how satisfied people feel with the leisure that they enjoy.

At the end of the day, the main challenge is to determine how an unobservable, such as leisure, can contribute to individual welfare. In this case, we are considering a double black-box. First, not everyone defines leisure in the same way and not everyone produces leisure experiences by using the same technology or the same inputs. For some people, the presence of others will be much more needed that for some other people. Some people could be much more materialistic than others. Some people could be much more efficient in the production of pleasurable experiences because of their higher education. Second, as indicated before, we know that leisure contributed to enhance the quality of life of people, but the valuation of those experiences is determined by societal norms and arrangements and by personal aspirations, past experiences and comparison effects.

Next section will present the happiness or subjective well-being approach. We will introduce a brief discussion of the rationale for using this approach for economic research and for leisure research. To do so, we will present the domain approach; in this setting, leisure satisfaction will be considered a mediator between individual leisure experience and overall satisfaction or happiness. In section 3, we will discuss the relationship between leisure time and well-being. Other crucial aspects will be discussed in section 4, where we review a series of social and economic factors that are said to influence leisure enjoyment, so leisure has a high quality and contributes to a better quality of life. Particularly, we will report previous findings on the social dimension of leisure, one of the attributes that determine high quality leisure experiences. In that same section, some determinants of overall satisfaction, or of particular domain satisfaction will be discussed. Last, sections 5 and 6 will present, respectively, some conclusions and a brief overview of needed research to better understand the contribution of leisure to a better quality of life.

Subjective well-being approach

Traditional research on quality of life relied heavily on objective and materialistic indicators of living conditions. Actually, Gross Domestic Product has been the “champion” indicator when studying the evolution of living standards and when comparing economies (Mankiw, 2007). Under the realm of objective indicators, nearly all non market activities and many aspects of human development, such as leisure, are neglected.

New studies have highlighted the superiority of including the subjective approach to the investigation of quality of life in developed and developing societies, and happiness research has become quite of a fashionable and popular topic (Layard, 2006). There is a growing interest on using the subjective well-being approach to analyze living conditions and there has been an emerging literature on social sciences. Among other reasons for that flourishing, we can highlight the following: (i) this approach offers richer insight about the quality of life, and considers other indicators of development apart from the traditional indicators; (ii) nowadays there is more information available about living conditions, opinions and perceptions of people and societies, and; (iii) with this approach it is possible to identify the major needs and problems of the population, which is useful for governments and policy makers (Frey and Stutzer, JEL 2002).

Economists and other social scientists broadly define `happiness' and `life satisfaction' as subjective well-being. Following Diener and Seligman (2004, pp. 4) life satisfaction is defined as a "global judgment of well-being based on information the person believes is relevant", while well-being "includes all of the evaluations, both cognitive and affective, that people make of their lives and components of their lives". While according to some authors, the terms "happiness," "subjective wellbeing", "well-being", "satisfaction" and "quality of life" are somewhat different and each have their own specific meaning, responses in different surveys are highly correlated (Fordyce, 1988; Frey and Stutzer, 2002b), and many analyses use them indiscriminately. In this current study these terms are used with the understanding that they have a similar connotation.

The present study will use a "bottom-up" approach to the analysis of subjective well-being. This approach considers that overall life satisfaction is determined by what is called "domain satisfaction"; the evaluation of own personal situation on different dimensions of life such as: financial situation, housing conditions, health, leisure, job or education, among other dimensions. Some authors signal the "mediator" role of those domain satisfactions to determine overall happiness (Cummins, 1996; van Praag et al., 2003; Easterlin and Sawangfa, 2007). In what follows, we will consider that leisure satisfaction has leisure experiences as the main input; higher leisure satisfaction will contribute, in turn, to higher overall satisfaction or happiness.

In order to assess the size of different influences upon happiness and satisfaction with life in general, psychologists have been using surveys since long ago, while only recently economists have recognized that there is useful information in a subjective well-being answer as an empirical approximation for the theoretical concept of utility. With the exception of the seminal work of Easterlin (1974), most research has taken place during the last two decades. The existing state of research suggests that, for many purposes, happiness or reported subjective well-being is a satisfactory empirical proxy of individual utility. From the information about the determinants of individual happiness, different situations of economic and social policies inside a country or a region can be analyzed .

Frey and Stutzer (2002b) give some important reasons for economists to consider happiness research. First, happiness research can help to evaluate net effects, in terms of individual utilities, for different economic policies. Understanding the determinants of subjective well-being can thus usefully inform economic policy decisions. Second, this research also has relevance to economists because of the effect of institutional conditions such as the quality of governance and the size of social capital on individual well-being. It may also help to solve empirical puzzles that conventional economic theories find difficult to explain. For instance, using this approach it is possible to understand why for several countries since World War ll although they have raised their real income drastically, the self-reported subjective well-being of the population has not increased or has even slightly fallen.

Data about happiness are collected through direct questioning via interviews or self-administered questionnaires in which individuals self-rate their happiness on a single item or on a multi-item scale. These scales offer a list of options, which are ranked according to the levels of happiness . Most studies of subjective well-being are based on some variation on the question "How satisfied (or happy) are you with your life?" The range of possible responses is defined over a scale that varies between datasets (one to four, one to seven, or one to ten), the lowest grades indicating a poor level of life satisfaction. The main use of happiness measures is not to compare levels in an absolute sense but rather to seek to identify the determinants of happiness.

The strategy is to use the answers that people give when asked questions about how happy they feel with life. Similar questions are posed with respect to job satisfaction, health satisfaction, housing satisfaction, satisfaction with marital relation, etc. …, and leisure satisfaction or satisfaction with leisure time. This study of the different aspects of life is called "domain" satisfaction.

Although this approach could have limitations, as was said by Oswald (1997, p. 1816) "if the aim is to learn about what makes people tick, listening to what they say seems likely to be a natural first step". The domains-of-life literature states that life can be approached as a general construct of many specific domains, and that life satisfaction can be understood as a result from satisfaction in these domains of life (Cummins, 1996; van Praag et.al, 2003; Easterlin and Sawangfa, 2007; Rojas, 2006a, 2006b).

It is evident that different domains may be distinguished. In many studies, the domains to be analyzed are determined by data availability. For instance, in the British Household Panel Survey leisure satisfaction is split up into two sub-dimensions; namely, the amount of leisure and use of the leisure time (Van Praag and Ferrer-i-Carbonell, 2007); the European Community Household Panel considers only satisfaction with leisure time , and the Latinobarómetro only includes satisfaction with the amount of leisure (Rojas, XXXX).

Rojas (2007) affirms that the enumeration and demarcation of the domains of life are arbitrary. In addition to this, there are many possible partitions of a human life, and the selected partition depends on the research's objectives and the available information. For example, Cummins (1996) has argued for a seven-domain partition: material well-being, health, productivity intimacy, safety, community and emotional well-being; van Praag et al. (2003) study the relationship of satisfaction in different domains of life (health, financial situation, job, housing, leisure and environment) and satisfaction with life as a whole.

Rojas (2006b and 2007), on the basis of factor analysis, identified seven domains of life: health, economic, job, family, friendship, personal and community. Using information from Mexico , he showed that satisfaction in the family domain is crucial for life satisfaction. Family satisfaction includes aspects of satisfaction with one's spouse, children and with the rest of the family. Rojas also showed that the satisfaction in the health, job and personal domains is also very important for a person's happiness. Satisfaction in areas such as housing and living conditions, financial solvency and income are relatively less important for life satisfaction. Rojas (2007) found that income is an explanatory variable of relevancy for economic and labor satisfaction, but not for family satisfaction or leisure satisfaction. For that reason, it is possible to find situations where a person is satisfied with his/her life while he/she is unsatisfied economically, or where a person is unsatisfied with his/her life and, at the same time, his/her economic satisfaction is high (Rojas, 2008b).

Empirical research has focused on different factors associated with subjective well-being and satisfaction. In agreement with psychological and sociological studies (Argyle, 1999), economic research has identified a set of personal and social characteristics associated with life satisfaction. Most studies using data from North America and European countries have found the level of reported life satisfaction to be high among those who are married (Blanchflower and Oswald, 2004b; Easterlin, 2003; Carroll, 2007; Clark et al., 2005; Ferrer-i-Carbonell and Frijters, 2004), women (Oswald, 1997; Clark, 1997), whites (Oswald, 1997; Alesina et al., 2004), the well-educated (Blanchflower and Oswald, 2004a; Frey and Stutzer, 2003; Borooah, 2005), the self-employed (Blanchflower, 2000; Blanchflower, 2004; Frey and Benz, 2003; Alesina et al., 2004), the retired (Di Tella et al., 2003), and those occupied with home duties (Di Tella et al., 2003; Borooah, 2005).

The relation between an individual's age and happiness seems to be a bit more complex. Many people believe that the quality of life deteriorates with age and that old people should be unhappier than young people since the old tend to have a worse health, less income, and few are married. Nevertheless, many studies have surprisingly thought that old people report levels of happiness comparatively higher than young people, though this effect tends to be small. Frey and Stutzer (2001) have indicated four reasons that can explain this positive relationship between age and happiness:

(i) the old have lower expectations and aspirations. For example, an elderly person waits to remain without work and possibly widower, so the effects of the loss will be lower on the old than on the young.

(ii) They have little disparity between goals and achievements, since the eldelrly's goals are fixed closer to what reasonably they can reach.

(iii) Older individuals have had more time to adjust to their life conditions, and

(iv) old people have learned how to reduce the negative events of the life and how to regulate the negative affects. Besides, economists have identified a U-shape in the relationship between age and happiness (e.g. Oswald, 1997; Blanchflower and Oswald, 2004a). This implies a convex shape in the relationship of life satisfaction with age. Life satisfaction decreases with age until it reaches a minimum, increasing afterwards. For North America and European countries this minimum typically occurs in the forties (43 in Frey and Stutzer (2001) and Ferrer-i-Carbonell (2005); 46 in Peiro (2007)).

Aspirations and comparisons effects also are important in relation with income and other factors affecting subjective well-being. The individual's reported subjective well-being in the present is based on a norm of what is `bad', `sufficient' or `good'. Such norms not only depend on the present situation, but also on what the individual has experienced in the past, on what he/she expects to experience in the future and on what other people think and do (van Praag and Ferrer-i-Carbonell, 2004). In relation with income, individual well-being does not only depend on income in absolute terms but also on the subjective perception of whether one's income is adequate to satisfy one's needs. In addition, individual income perception is subject to the individual's own situation, past and present, as well as to the income of other people. The latter reflects the importance of the relative position of individuals in society for their satisfaction with life. This is often referred to as the "comparison income" or "relative utility" effect.

It is often argued that individuals adapt to new situations by changing their expectations (Easterlin, 2005; Clark et al., 2008). This implies that higher incomes are accompanied by rising expectations that lead to what is known as "the hedonic treadmill" (Brickman and Campbell, 1971) or "hedonic adaptation" (Frederick and Loewenstein, 1999). Thus, individuals strive for high incomes even if these lead only to a temporary or small increase in well-being. This ability to adapt would appear to be a ubiquitous feature of the human condition, some recent examples of adaptation in nonmonetary spheres are Lucas et al. (2003) and Lucas (2005) with respect to marriage and divorce, Wu (2001) and Oswald and Powdthavee (2006) for adaptation to illness or disability, and Lucas et al. (2004) regarding unemployment.

The comparisons with different social reference groups are also an important factor that has been widely present in the analysis of two dimensions; namely, the analysis of the effect of relative income on financial satisfaction and/or satisfaction with life as a whole (McBride, 2001; Stutzer, 2004; Luttmer, 2005; Clark, Frijters and Shields, 2008) and the influence of unemployment on subjective well-being. A standard result in happiness literature is that the unemployed report significantly lower levels of subjective well-being than other labor force groups (Winkelman and Winkelman, 1998; Frey and Stutzer, 2002). Indeed, the pecuniary and the non-pecuniary costs of the unemployment are that high that adaptation is non-existent (Lucas et al., 2004) or only very moderate (Clark, 2002). Clark (2003) uses seven waves of the British Household Panel Survey to test for social norms in labor market status. In his analysis, he found that the well-being of the unemployed is the higher, the higher the unemployment rate in a reference group (at the regional, partner, or household level). It seems that, the more unemployment becomes the norm, the less individuals are affected by it (Winkelman, 2006). Lalive and Stutzer (2004), using a different strategy, obtain the same results for information from Sweden.

Social interactions could be either a negative or a positive factor. As previously mentioned, an individual's happiness depends on that individual's own relative (or positional) situation or status, and comparison with others, what would expose that individual to negative externalities in terms of "peer-effects" (Luttmer, 2005) in utility and/or consumption. Alpizar, Carlsson and Johansson-Stenman (2005) show that positionality matters far more for commodities as houses and cars than for vacation and insurance, but also that both absolute and relative consumption matter for each category, these are "positional goods". The positive influence of social interactions may come from social relationships and other "relational goods" or social capital factors.

For instance, Rojas (2007), Winkelman (2006), Argyle (1999), among other social scientists have found that social relationships are a major source of well-being. Although marriage is the relationship that has the most influence on happiness, there are other relationships that affect happiness, as well as health and mental health, by providing "social support". Argyle (1999 p. 361) refers some studies where it was found that if all kinds of social support are combined, a social support factor is found to have a strong correlation of 0.50 with happiness. Social scientists in many countries have observed that social support or social networks (and the associated norms of reciprocity and trust (Helliwell and Putnam, 2004)) have powerful effects on the level and efficiency of production and well-being, broadly defined, and they have used the term social capital to refer to these effects (Coleman, 1988; Putnam, 2000; Woolcock and Narayan, 2000).

Lately, some cross-sectional studies from both sociology and economics have shown the importance of key aspects of social capital -- such as trust, social contacts and membership in voluntary associations -- over individual well-being (Inglehart 1999; Putnam 2000; Helliwell 2003 and 2006b; Powdthavee, 2008). In Bowling Alone, Putnam (2000) suggested that people prosper in neighborhoods and societies where social capital is high, that is, where people trust one another and are mutually helpful. Putnam reviewed evidence showing that communities with high rates of volunteer activity, club membership, church membership, and social entertaining (all thought to be indirect manifestations of social capital) all had higher well-being than communities that were impoverish these characteristics. Many studies that use cross-sectional data have shown that individuals with rich networks of active social relationships, that do not include people living in the same household, tend to be happier with their lives (Phillips 1967; Burt 1987). Helliwell (2003) reported that well-being is high and suicide rates are low where trust in others is high, and he also found that well-being is high where memberships in organizations outside of work are at high levels. Thus, there is evidence that individuals are more likely to experience high well-being when they live in nations with high social capital than when they live in nations with low social capital, a finding that dovetails with the results of studies on individuals' social interactions. Helliwell and Putnam (2004) and Powdthavee (2008) are comprehensive reviews about the importance of social capital factor over subjective well-being.

Health status is a factor that can be expected to be an important determinant of life satisfaction. In the 1950s the use of concepts such as "welfare", "adjustment" and "mental health" had much in common with the traditional concept about "happiness" (Argyle, 1991). Research on the health-related quality of life was developed in the mid 1970s by health scientists and psychologists in order to track people's perception of their health status (Gough et al., 2007). This was mainly in response to the need for more sensitive measures to compare treatments for chronic illness and to identify the most cost-effective treatments . Good health is considered an important factor included in the capabilities and the necessary functionalities in order for an individual to face life (Deaton, 2007; Sen, 1999). Since the 1980s the state of health has been identified as an important determinant of life satisfaction, as happy people are healthier, both physically and mentally (Veenhoven, 1991; Argyle, 1999). Consequently, poor health, which limits an individual's ability to carry out their daily activities, reduces overall satisfaction.

The literature about subjective well-being in Latin American countries is few and very recent. Graham and Pettinato (2001) were some of the first to analyze Latin American countries. Using the Latinobarómetro 2000, they found that Latin America is not all that different from the advanced industrial economies in relation to some of the determinants of happiness. Similar to the OECD countries, happiness has a quadratic relationship with age, initially decreasing and then increasing monotonically after 49 years of age. As in the industrial countries, being married had positive and significant effects. In contrast to the advanced economies, a significant gender effect was no found in Latin America. Also, as in the industrial countries, the coefficients for level of wealth were strong, positive, and significant in happiness. When wealth was included in the regressions, the coefficient for education level became insignificant or weakly significant, depending on the regression used. Being self-employed or unemployed both had significant and negative effects on happiness. When they included country-fixed effects, the coefficient on self-employment became insignificant. While being unemployed also has negative effects on happiness in the advanced industrial economies, being self-employed has positive effects. The most credible explanation is intuitive and it was given by the authors: most self-employed people in the latter are self-employed by choice, while in developing economies, many are self-employed due to the absence of more secure employment opportunities and live a precarious existence in the informal sector.

Other analyses by countries have been conducted in Latin America. Among the most important, Rojas (2006b and 2007), using the domains-of-life approach in Mexico, found that people are on average, more satisfied in the family domain, while they are less satisfied in the consumption, personal and job domains. Rojas (2007) found that income is an explanatory variable of relevancy for the economic and labor satisfaction, but not for either family or leisure satisfaction. Due to that, he found a weak relationship between income and life satisfaction Gerstenbluth et al. (2007) studied the relationship between happiness and health in Argentina and Uruguay using the Latinobarómetro 2004. Cruz and Torres (2006), using the Encuesta de Calidad de Vida 2003, tested various happiness hypotheses among Colombians and Cid et al. (2008), using the survey called Salud, Bienestar y Envejecimiento en América Latina y el Caribe (SABE), explored the correlation between happiness and income in the elderly in Uruguay. To our knowledge, the previous studies conducted about Latin America have not included the effect of social capital on subjective well-being, and they have analyzed the self-employment as a homogeneous labor market status.

However, when considering the specificity of the leisure domain, we should take into account that while satisfaction with other realms of life may lie upon the valuation of objective situations (such as one’s financial situation, health or housing conditions), satisfaction with leisure brings in an additional challenge as individual’s boundaries of leisure are defined by her perception of what is pleasant (Ateca-Amestoy et al., 2008).

Conceptual discussion on the nature of leisure time in contemporary societies

Time allocation decisions within the family: economic approaches and models. We will attach to the economic approach to human behavior by Becker (moreover, bring arguments such as those contained in a theory of social interactions).

'Temporal autonomy' is a matter of having 'discretionary control' over your time.
Discretionary Time. A New Measure of Freedom (Goodin et al., 2005)
Other approaches: we have found these relevant arguments:

1.Veblen's theory
2.The omnivore
3.Bourdieu's distinction
4.Putnam's social capital

3.1. What is Social Capital?

There is a traditional consensus that there exists three distincs traditions that conceptualize and analyze social capital. All three would be relevant for our reasoning.

- Pierre Bourdieu <cite>bourdieu2</cite>:

who conceptualised social capital as the `actual or potential resources' that an individual has at his/her disposal as a result of `a durable network of more or less institutionalised relationships of mutual acquaintance and recognition', i.e. membership in a group. Some authors point out that this definition must be viewed as part of his broader concern with developing the different types of capital in order to explain the means by which the social stratification system is preserved and the dominant class-reproduction strategy is legitimised.

- J.S. Coleman <cite>coleman</cite>

defines it as the set of resources that inhere in family relations and in community social organisations and that are useful for the cognitive or social development of a child or young person. Social relations were viewed by Coleman to make up important `capital resources' for individuals by means of processes such as setting `obligations, expectations and trustworthiness, creating channels for information, and setting norms backed by efficient sanctions'. These resources may be influenced by factors such as generalised trustworthiness which ensures that obligations are met, the extent to which a person is in need of help and differences among cultures as to whether aid should be requested or given.

- Robert Putnam <cite>putnam1</cite>, <cite>putnam2</cite> and <cite>putnam3</cite>: Social capital refers to features of social organisation, such as trust, norms, and networks, that can improve the efficiency of society by facilitating co-ordinated action. Or rather,"Social capital" refers to features of social organisation such as networks, norms and social trust that facilitate coordination and cooperation for mutual benefit. Underpinning the concept is the notion that social networks encourage norms of reciprocity, ideally generalised rather than specific reciprocity. Whilst acknowledging that there are different forms of social capital (e.g. involving `multi-stranded networks', formal organisation or `public-regarding purposes'), he argues that forms of social capital vary (more or less) along two key dimensions: between bridging (or inclusive) and bonding (or exclusive). There is some disagreement in the literature as to the usefulness of this distinction. On the one hand, Warde and Tampubolon <cite>warde&tampubolon</cite> (2002: 158) argue that the distinction has proved difficult to operationalise.

This disagreement is rooted unsurprisingly in the absence of consensus not only as to how social capital should be conceptualised, but also as to its usefulness, and in the contentious debate concerning how social capital should be measured. The variety of measures of social capital that exist are evidence of the underdeveloped nature of the relationship between theoretical and empirical research on social capital. There are two principal ways in which academic scholars operationalise social capital for the purposes of empirical investigation. By including their own approach which emphasises the importance of social context for social capital, we can consider three approaches in all:

- First, the analysis of social capital by political scientists as well as some economists and psychologists, following Putnam tends to focus on trust, norms and values. His measures of social capital are typically based on a composite index containing the following elements:

(i) intensity of involvement in community or organisational life,

(ii) Public engagement (e.g. voting),

(iii) Community and volunteering,

(iv) Informal sociability (e.g. visiting friends), and

(v) Reported levels of interpersonal trust.

The use of aggregate measures of social capital derived from survey research (e.g., `generalised social trust', membership in organisations and norms such as reciprocity, cooperation and tolerance) has been criticised for overlooking how the norms and attitudes of individuals may be affected by the different social contexts in which they find themselves.

- Second, by contrast, Foley and Edwards argue that sociologists in particular tend to conceive of social capital as present in the structure of social relations between individuals and among individuals, which is operationalised as (both formal and informal) social networks, organisations or connections between individuals and/or organisations. The presence of social capital is viewed to be connected to local social structures (e.g. community social organisations), which in the tradition of Coleman has benefits for particular individuals or groups.

- Finally, and in a related way, Foley and Edwards (1999: 146) argue that it is important to consider how social context affects the production of social capital: The specific social context [e.g., community, organisation or network] in which social capital is embedded not only influences its `use value' [e.g. the potential for building links across different social groups]; it also shapes the means by which access to specific social resources [e.g., informal and formal and social organisation] is distributed and managed.

As will become clear in the next section, given how Putnam is associated with the "popularisation" of the concept of social capital, the influence of his approach to the analysis of social capital is evident in the literature which focuses on the cultural sector. However, an underlying trend is also emerging, in the vein of Foley and Edwards whereby scholars seek to explicate the importance of the access of individuals, groups and communities etc. have to social capital as well as their ability to translate this access into positive benefits.

Descriptive findings (start by most general reference: Gershuny).

We distinguish in the GSOEP data set between three kinds of time use, i.e. working time, household work, and leisure. Not unexpectedly, the number of working hours has a strong negative effect on leisure satisfaction, while the number of hours spent on leisure has a small positive effect (Table 6).

Table 6. Leisure satisfaction

GLS with individual random effect and fixed time effects.

The age effect is again U-shaped with a minimum at about 35 for workers and 31 for non-workers. Household income is not a strong factor for leisure satisfaction, but the level effects are always positive. More education leads to less satisfaction with leisure. It seems that there is a tendency for people to enjoy their leisure time most when they live alone. Both, the presence of adults and that of children have a negative effect on leisure satisfaction, and living together has also a negative effect, although only significant for Eastern non-workers. Males enjoy their leisure more than females.<cite>vanpraag2</cite> "The anatomy..." JEBO

The data are from the Danish Time-Use Survey 2001, merged with register data. Results show that the substitution of money for time is more prominent for women than for men, because they have a larger income share of time-intensive value of housework, while men have the larger share of disposable income. Furthermore, when the spouses share income resources the women give up more value of housework than they get disposable income in return.<cite>bonke1</cite>

Time and money are basic commodities in the utility function and are substitutes in real terms. To a certain extent, having time and money is a matter of either/or, depending on individual preferences and budget constraints. However, satisfaction with time and satisfaction with money are typically complements, i.e., individuals tend to be equally satisfied with both domains. In this paper, we provide an explanation for this apparent paradox through the analysis of the simultaneous determination of economic satisfaction and leisure satisfaction. We test some hypotheses, including the hypothesis that leisure satisfaction depends on both the quantity and quality of leisure-where quality is proxied by good intensiveness and social intensiveness. Our results show that both the quantity and the quality of leisure are important determinants of leisure satisfaction, and, since having money contributes to the quality of leisure, this explains the empirical findings of the satisfactions being complementary at the same time as the domains are substitutes. Interestingly, gender matters. Intra-household effects and especially individual characteristics are more pronounced for women than for men for both domain satisfactions. Additionally, good intensiveness is more important for men (e.g., housing conditions), whereas social intensiveness is more important for women (e.g., the presence of children and participation in leisure-time activities).<cite>bonke2</cite>

The Temporal Welfare State: A Cross-national Comparison

Welfare states contribute to people's well-being in many different ways. Bringing all these contributions under a common metric is tricky. Here we propose doing so through the notion of "temporal autonomy": the freedom to spend one's time as one pleases, outside the necessities of everyday life. Using surveys from five countries (the United States, Australia, Germany, France, and Sweden) that represent the principal types of welfare and gender regimes, we propose ways of operationalizing the time that is strictly necessary for people to spend in paid labor, unpaid household labor, and personal care. The time people have at their disposal after taking into account what is strictly necessary in these three arenas --- which we call "discretionary time" --- represents people's temporal autonomy. We measure the impact on this of government taxes, transfers, and childcare subsidies in these five countries. In so doing, we calibrate the contributions of the different welfare and gender regimes that exist in these countries, in ways that correspond to the lived reality of people's daily lives.<cite>rice, goodin, parpo</cite>

Discretionary

People's welfare is a function of both time and money. People can -- and, it is said, increasingly do -- suffer time-poverty as well as money-poverty. It is undeniably true that people feel increasingly time pressured, particularly in dual-earner households. But much of the time devoted to paid and unpaid tasks is over and above that which is strictly necessary. In that sense, much of the time pressure that people feel is discretionary and of their own making. Using data from the 1992 Australian Time Use Survey, this paper demonstrates that the magnitude of this `time-pressure illusion' varies across population groups, being least among lone parents and greatest among the childless and two-earner couples.<cite>goodin, rice, bittman, sanders</cite>

Time pressure is a familiar phenomenon. The quantity of spare time people have clearly effects their satisfaction with their leisure and with their life as a whole. But so too, we show, does how much control people have over how much spare time they have. We measure this through an indicator of "discretionary time", which proves to be equally or more important than spare time itself in these connections. Lina Eriksson, James Mahmud Rice, and Robert E Goodin, "Temporal Aspects Of Life Satisfaction", Social Indicators Research, in press.

Nazio and McIness study the satisfaction with the amount of leisure time as an indicator of time stress and determine the influence of childbering into the stress of their parents. Usign the data contained in the ECHP for all waves and 11 countries <cite>nazio&macinnes</cite>.

They also indicate that Our second substantive finding, that job satisfaction is a very substantial prophylactic against time stress is also, at first sight, surprising, but is nevertheless corroborated by evidence of quite a different kind for the UK (MacInnes 2005 MacInnes, John. 2005 `Work Life Balance and the Demand for Reduction in Working Hours: Evidence from the British Social Attitudes Survey 2002.' British Journal of Industrial Relations, 42(3):273 -- 295.). Both men and women who say they are very satisfied with their jobs, or with their `main activity' if they are not employed, rarely suffer time stress even if they work long hours or have young children. This finding demonstrates vividly the essentially social nature not only of time perception, but the social construction of work and leisure, effort and reward, necessity and freedom or alienation and expression. Those who spend their time doing what they value or enjoy, however intensively or extensively they may do so, rarely feel bereft of `free' time'.

Model and empirical especification

In a model of economic behaviour, individual utility depends on commodities that are produced (using both consumption goods and time), and consumed within the household. In our approach, we follow Becker's <cite>becker</cite> idea of the household as a factory combining market goods and time to produce the utility-maximizing set of commodities. Whether one purchases groceries to combine them with time spent shopping, cooking or enjoying a nice meal is up to the analyst, who must decide into what consistent set of commodities to classify these goods and time uses. Thus, any definition of commodities requires reasonable choices about categories and the classification of inputs of goods and time (for further details on the topic see Gronau and Hamermesh <cite>gronauham5</cite>).

In what follows, we will present a simple model of utility maximization and will concentrate in the role of time - in particular what variables will determine the optimal allocation of time devoted to leisure fullfilment - as a productive factor that enters into the commodity production function and as a resource that is equally distributed a priori among individuals.

To simplify the discussion, since we focus on satisfaction with leisure time, our model assumes individuals derive utility over two specific commodities, namely leisure experience and household maintenance. The latter is treated as a composite commodity aimed at the fulfillment of any other household manteinence need. Both commodities are produced by each individual in the household by means of the optimal allocation of personal resources (time and market goods, as well as other non-market resources such as social relations).

There have been several proposed utility functions / valuation functions / satisfaction models with a given domain function. In all these cases,
For instance, Blanchflower and Oswald <cite>blanchfloweroswald</cite> postulate the existence of a reported well-being function of the type

r=h(u(y,z))+e

with r some self-reported number from a given satisfaction scale, u the person true well-being or utility, h(⋅) a continuous non-differentiable function relating actual to reported well-being, y is the argument or life-domain being valued -either could it be real income, number of leisure hours, health status...-, z is a set of sociodemographic and personal characteristics, and e is the error term. They assume u(⋅) to be a function only observable to the individual, and the error term e to subsume -among other factors- the inability of human beings to communicate accurately their happiness level. The structe of such a function is suitable for an ordered probit or logit model, considering that utility is a lantent and unobserved variable that is related to the elicited satisfaction level.
An alternative way is that presented in Ateca-Amestoy et el. <cite>ateca</cite> that will be followed in order to keep track of the individual decision making and valuation process. Assuming additivity of the arguments (commodities) in the utility function, we could define the indirect utility function associated to the leisure experience VO^{i}(Z₁^{∗}) in terms of the solution to a utility mazimization problem.
When evaluating the outcome Z₁^{∗}, we should expect individuals to evaluate it in relative terms, that is in comparison with some specific personal reference level Z_{i1}, so their valuation of this particular domain would be of the form

OLS^{i}=f^{i}(VO^{i}(Z_{i1},Z₁^{∗}))

where f^{i}(⋅) should be increasing on the level of Z₁^{∗}, and decreasing on the gap between Z_{i1} and Z₁^{∗}.
Given that the survey asks individuals about the degree of satisfaction withthe amount of leisure time and they answer by giving a number between a range from a minimun level of 1 (not satisfied) to 6 (fully satisfied), there is a correspondence between this indirect utility and the stated satisfaction. The empirical exercise that we perform will focus on the determinants of utility derived from leisure time and its correspondence with the declared satisfaction.
In what follows, we will develop a model aimed to derive testable propositions on the determinants of satisfaction with leisure time.

Theoretical Model

Suppose that each commodity is produced by using a private market goods and time vector of the following form

U^{i}(Z₁,Z₂)=UO^{i}(Z₁)+UH^{i}(Z₂) #3.1

where U^{i}(⋅) is utility of agent i (i=1,...,N), UO^{i}(⋅) is the leisure utility function and UH^{i}(⋅) is the household manteinance utility function. Z₁ stands for the amount of the so-called "leisure experience" commodity of individual i, and Z₂ includes all other "activities" that are needed for individual survival (this commodity, which will be our composite good, is called "household manteinance" from now on).

<footnote>El hecho de que un individuo adquiera comida para combinarla con tiempo empleado en la compra, en cocinarla o en disfrutar de una comida agradable es algo que ha de ser modelizado por el investigador, que tiene que decidir cómo va a clasificar estos bienes y usos de tiempo dentro de un conjunto consistente de mercancías. Cualquier clasificación de mercancías va a requerir elegir de forma arbitraria aunque razonable entre las diferentes categorías y clasificación de factores de bienes y tiempo (para más detalles sobre el tema, ver <cite>gronauham5</cite>).

</footnote>
Both ((∂UO^{i})/(∂Z₁)) and ((∂UH^{i})/(∂Z₂)) are positive.
Each commodity is produced using a vector of private market goods and time as follows:

Z₁=f(x₁,...,x_{n},t_{l};X_{S},X_{e}) #3.2

where x=(x₁,...,x_{n}) is a vector of (private) market goods that are used to produce leisure experience, and t_{l} is individual's discretionary time understood as that moment when individuals perceive themselves to be free of external demands and in control of their own situations (Kelvin, 1979). Further, X_{S} is a vector of personal characteristics that parameterise production of leisure experience (i.e., how skilled and productive the individual is), and X_{e} is the vector of environmental conditions that shape the production function --no matter if they work as "extra" non personal public goods or as constraints--. These "environmental variables" reflect the state of the art of production, or the level of technology of the production process.

Equally, the household maintenance commodity is produced according to the following production function:

Z₂=f(y₁,...,y_{m},t_{h};X_{S},X_{e}) #3.3

where y=(y₁,...,y_{m}) is a vector of (private) market goods that are used to produce household maintenance, and is individual's time devoted to household activities (production and consumption).

The utility function is maximized subject to the production functions constraints (equations (3.2) and (3.3)) and the usual budget and time constraints:

wt_{w}+A = ∑_{i=1}ⁿp_{i}x_{i}+∑_{j=1}^{m}q_{j}y_{j} #3.4
T = t_{w}+t_{l}+t_{h}

where w is the wage rate; A is the individual's non-wage income; and (p_{i})_{i=1}^{m} , (q_{j})_{j=1}^{m} are the prices of market goods used in producing ; t_{w} and is the time spent in the labour market. For the development of this model, the way individuals understand their leisure time is especially relevant: while time devoted to household maintenance activities can be exchanged in the market, the subjective nature of leisure makes it essential to devote personal time to the production and consumption of leisure.

<footnote>In this analysis it is specially relevant the so called first person criterion for leisure time. While time devoted to household maintenance can be purchased in the market, time available for leisure experience production and consumption has to be a personal resource, given the subjective condition of leisure commodity.
</footnote>
The budget and time constraints can be collapsed into a single resource constraint on the individual's "full income"

wT+A=∑_{i=1}ⁿp_{i}x_{i}+∑_{j=1}^{m}q_{j}y_{j}+wt_{h}+wt_{l} #3.5

The utility function (3.1) is maximized subject to the constraints of the production functions (3.2) and (3.3) and full income (3.6). The solution to such a problem leads to optimal allocation of resources and optimal prodution of commodities, Z₁and Z₂, that characterizes consumer behaviour.

Dado el carácter aditivo de la función de utilidad, podemos definir para el individuo i la utilidad proporcionada por la experiencia de ocio mediante la sustitución de su cantidad óptima en la función de utilidad

<footnote>El problema del consumidor está escrito en términos deliberadamente ambiguos en cuanto a qué variables temporales son variables de elección. De esta forma, el modelo permite acomodar distintas situaciones que van desde que el tiempo dedicado al ocio tenga carácter residual hasta el caso en el que se elige en pie de igualdad con respecto al tiempo dedicado al trabajo y al mantenimientos del hogar. En todo caso, independientemente de la jerarquía otorgada a las distintas variables temporales, la función indirecta de utilidad tomará la expresión incluída en el texto.
</footnote>
The formulation of optimal decision-making can be rewritten in terms of the individual indirect utility function

VO^{i}=UO^{i}((p_{i})_{i=1}ⁿ,(q_{j})_{j=1}^{m},w,A;X_{s},X_{e}) #3.6

This last indirect utility function will allow the model to be solved for utility given a value of leisure experience and household maintenance, in terms of all exogenous parameters.

Understanding leisure satisfaction as a "mediator" between leisure experience and happiness, we focus our analysis on the measurement of individual leisure satisfaction, as a specific domain of general satisfaction with life, and the identification of its determinants. We believe that life satisfaction is influenced by many factors other than leisure experience, while leisure satisfaction has leisure experience as a major input.

We cannot observe the objective leisure experience satisfaction (OLS_{i}) that a particular individual has reached under her surveyed conditions which would depend on a level of leisure experience and on personal characteristics capturing objective and subjective heterogeneity (some of which are observable and some of which are completely unobservable, such as ambitions or aspiration levels). However we can get a measure of her subjective leisure satisfaction (LS_{i}). This is done by asking individuals how they feel about their current leisure experience. The answer to this question takes discrete values from 1 (totally unhappy) to 7 (totally happy), and we assume that such an answer is meaningful and comparable between individuals (<cite>clarkoswald</cite>, <cite>clark</cite>, <cite>ferrer</cite>, <cite>ferrerfrijters</cite>), providing interesting and plausible results.

On our empirical specification, the decision on which variables to include is ultimately based on exploratory analysis and data availability. Thus, explanatory variables will include objective ( ) and subjective ( ) personal variables, socio-economic ( ) and household composition variables ( ), and social capital ( ) and environmental ( ) variables as follows,
(9)Table 1 reports the definition of the specific variables used for this research. The hypotheses and empirical regularities of these available variables will be discussed in Section 3.

Since LSi is an ordered categorical variable, we estimate the usual Ordered Probit model (Greene, 1990). In line with mainstream Subjective Well-being literature, we consider that the ordered variable is linked to an underlying latent variable. The real axis which is the domain for our latent variable is divided in intervals , such that the latent variable OLS if LS = k ; (k={1,…,7 }).
non-negativity constraints and an upper limit to the time that can be devoted to work in the market, will introduce new insights into the most simple model.
However, we feel that we can go one step further by introducing no continuous, but discrete time decision variables.
Our individual decision making model differs form Jenkin's <cite>jenkins&osberg</cite>. When using commodities - particularly a commodity called leisure - as the argument of individual utility functions, we cannot apply directly the argument that the marginal utility of each individual's leisure depends on the choices made by other people. Instead, the presence of enjoyable / companiable others enter as an argument into the commodity production function.

Hypothesis presentation, data description and empirical specification -- estimation methods


Hypothesis; variables others than price and income

Van-Praag & Ferrer-i-Carbonell (2004) report different Satisfaction questions that are measured in Social Surveys across Europe. In our Survey, there is just this "outcome" evaluation question, since there is no reference to the question of satisfaction with leisure time.

We now focus on the identification of determinants of individual's leisure experience satisfaction. Although current economic research provides little guidance on testable hypotheses for individual's leisure experience satisfaction, we will infer some results from our theoretical model and bring some others reported on the literature on time allocation in an attempt to provide a reasonable framework for testing. In this sense, our theoretical model helps us keep track of different level of leisure experience production based on individual resource availability since it captures the trade-off ruling among competing uses of time and goods. Our empirical specification cannot say anything on how aspirations are individually formed. Since aspirations will introduce additional individual heterogeneity in our model, we control for other socio-demographic variables that may have a relevant effect on the unobserved aspiration and goals regarding leisure.

Gronau and Hamermesh <cite>gronauham5</cite> point out that leisure is one of the most time-intensive commodities. Therefore, "time" is perhaps one of the most influential factors on individual's satisfaction with her leisure experience. However, we assume that time is not valuable by itself. The passing of time does not provide any pleasure to a given agent. Instead, it is an input that is combined with other productive factors in the household production function. We can state further that while other productive factors can exhibit some degree of substitutability, the presence of time resources will be critical for the production of leisure.

Since unfortunately, our Survey does not contain any direct question on the amount of time devoted to market or household activities, nor does it contain any reference to leisure time, we must proceed indirectly. Hence, we begin by considering the relevance of variables that will model the time availability that a given individual may enjoy. In doing so, individual occupational status may be one of the most important features affecting the level of one's satisfaction with her leisure experience. Market work is a time consuming activity, so it is supposed that working activities that require more in the market working hours will reduce time availability for leisure (Z₁), as well as for other household maintenance activities (Z₂). This restriction on the use of time will eventually have a potential negative effect on the level of individual leisure satisfaction. Empirical evidence supports this idea as Van Praag and Ferrer-i-Carbonell (2004) find how the number of working hours has a strong negative effect on the amount of leisure satisfaction. This result is further supported by the empirical research of Ahn, et al. (2003). In contrast, researchers have investigated the relevant significance between work and leisure as a quality of life measurement (Campbell, Converse & Rodgers <cite>campbell</cite>, 1976; Haavio-Mannila, 1971 <cite>haavio</cite>; London, Crandall, & Seals, 1977). Results indicate that market work can also deeply influence the enhancement of one's leisure satisfaction.

Time availability can further be constrained with the number of household responsibilities. Thus, household composition and, in particular, the presence of children and/or of handicapped persons or elderly with continuous aid requirements may impose a higher demand of both time and resources devoted to Z₂ (household maintenance). It is expected then, and empirically tested, that individuals with household responsibilities will enjoy their leisure experience to a lesser extent (Van Praag and Ferrer-i-Carbonell, 2004; Gronau and Hamermesh, 2003).

Besides time, personal resources understood both as private goods and as personal conditions are productive factors needed to produce and consume individual leisure experience. Accordingly, reported household income is an individual's resource likely to be positively correlated with leisure satisfaction, as more income means more expenditure capacity in market goods and services to produce leisure experience. Nevertheless, there may also exist a negative relationship between available income and leisure satisfaction (Bonke, Deding and Lausten, 2004), especially when the bulk of household income comes from work (as it limits the availability of time for other non-work commodities -- Z₁ and Z₂). To overcome this problem we specifically introduce individual's leisure expenditure capacity as the amount of money that potentially goes to produce leisure experience. We expect a positive relationship between this leisure expenditure capacity and the level of leisure satisfaction. In line with this argument, we further control for the amount of durables within the household as the amount of market goods, services or amenities others than basic ones (e.g., private swimming pool or green areas, garage, dishwasher, pay-TV, PC, second house) that also enter, as private resources, into the leisure experience production function.
Equally, individual's (reported or self assessed) health status, understood as a personal condition that enables the individual to display more physical effort activities with a smaller level of mental stress, may additionally have a significant positive effect on both leisure time and leisure satisfaction. Since healthier people are more likely to invest fewer resources on health -physical care-, being this one of the components of our composite commodity, they are supposed to choose larger amounts of leisure. Moreover, they may also be more efficient when enjoying their leisure (Chang, W., Oh, Sae-Sook, Oh, Sei-Yi. 2001).

Although we attempt to model individual's satisfaction with one's leisure experience, utility derived from leisure time undoubtedly benefits from the presence of companionable others. Many of the things people do in their non-work time involve other people, and are distinctly more pleasurable if done with other; indeed many things are impossible without others (for an empirical approach on the implications of Leisure Coordination see Jenkins and Osberg <cite>jenkins&osberg</cite>).

We hypothesize that an individual's time use choices are contingent on the time use choices of others because the utility derived from leisure time often benefits from the presence of companionable others inside and outside the household. We develop a model of time use, and demonstrate that its consistency with the behaviour of British working couples in the 1990s. We present evidence of the synchronisation of working hours by spouses and report estimates indicating that propensities to engage in associative activity depend on the availability of Suitable Leisure Companions outside the household. Our results indicate the importance of externalities in the working time decisions of individuals.

On the role of social capital . Warde and Tampubolon <cite>warde&tampubolon</cite>, point out the relevance of social capital on leisure consumption.

This paper reflects on the way in which personal ties affect the nature and content of consumption. While it is banal to observe that friends, kin, colleagues and neighbours influence anyone's pattern of consumption, comparatively little work exists on how this process operates. The paper will be illustrated by some secondary analysis of the British Household Panel Survey. The BHPS has a panel of approximately 10,000 people who have been interviewed annually since 1991. It analyses aspects of individual consumption in relation to people's associational involvement and friendship ties. The data is explored in the context of debates about social capital, attempting to apply the concept in order to analyse recreational practices. It is argued that social capital is a flawed concept and that greater appreciation of the complexity and diversity of network ties is required to understand how personal connections influence consumption.

Taking into account this social dimension of leisure, contacts with known people and participation in associations (understood as two dimensions of social capital, namely: informal and formal sociability), are resources that can be used to develop social leisure. Furthermore, variables such as household type and marital status should also be relevant in the sense that individuals may prefer to enjoy leisure with their closest relatives. However, since many of those activities may lie in the boundaries between leisure experience and household maintenance (depending on a personal appreciation which might also vary with time), we cannot hypothesize a clear effect. Empirical evidence with German panel data (GSOEP) reports a tendency for people to enjoy their leisure time most when they live alone (Van Praag and Ferrer-i-Carbonell, 2004).

The level of satisfaction with one's leisure experience is lastly affected by other personal variables, which mainly capture individual's heterogeneity. In this sense, it is important to take into account the relevance of both tastes (as each agent define the boundaries of her own "leisure experience") and skills (since she implements an optimal allocation of resources given a technology to produce and consume that leisure experience). Individual's age and gender are some of the factors, which are likely to affect one's evaluation of her own leisure experience. Empirical evidence suggests a u-shaped behaviour of the age regressor and a greater level of leisure satisfaction among men (Van Praag, and Ferrer-i-Carbonell, 2004). When controlling for age, we are allowing for distinct satisfaction statements to depend on different life moments. In this way, our estimation can accommodate the extent to which past experiences and expectations for the future operates in the comparison between effectively enjoyed leisure and desired leisure. Further, we include subjective social class as a proxy for people's status (being this a socio-economic concept with multiple dimensions). We believe, that the level of individual leisure satisfaction and the reported definition of one's social class are not independent areas as they may capture personal self-perceptions and habits (which potentially brings into the model additional unobserved individual heterogeneity).

Finally, although individual socio-demographic and socio-economic characteristics seem clear determinants of individual's leisure experience satisfaction, we have to take a look at environmental factors. Some environmental conditions will affect both tastes and skills (technology) since the environment will supply a bundle of non-personal physical capital: green areas, public or commercial areas, cultural and social equipment, and other amenities and services. In order to control for the potential effect on leisure satisfaction of these environmental factors, type of habitat is included in our analysis. Residence in small sized towns (less than 20,000 inhabitants) is likely to decrease the level of leisure satisfaction as the services delivered in these places are sometimes far from those claimed to be fair by their residents.

Data description:

How satisfied are you with your present situation in the following areas? Using the scales 1 to 6 [position "1" meaning that you are not satisfied at all, and "6" that you are fully satisfied], please indicate your degree of satisfaction with each case: (...) The amount of leisure that you have.
We can usually find that satisfaction questions are addressed to the evaluation of different domains of life. For instance, the ECHP: satisfaction with job.... BSP / GSOEP ... However, we are using a question regarding the satisfaction that a given productive resource (time, and more precisely, leisure time) provides people.

Results

Some preliminary results:

The amount of leisure time and the satisfaction with leisure time There is not any available direct measure of the number of leisure hours that the individual enjoys. Indirectly, we can infer the quantity of leisure time that an individual can enjoy in terms as the complement of working hours and hours devoted to household maintenance activities. As expected, there is a negative relationship among non-leisure time and satisfaction with leisure time. We have found evidence that indicates that the effect is relevant for both the average level (inter-person) and the change effect (intra-individual).

Role of occupation However, after controlling for these time-related variables, there is a significant effect of occupation. In this sense, employed-workers are significantly more satisfied with their leisure time than people in any other category. For the comparison between self-employees and employed, this is absolutely intuitive: we will expect employers to have less uncertainty on in which moments they would be able to use their leisure time resource; for unemployed there is additional evidence to their lower levels of satisfaction in each domain of life; whereas for the other groups it would be more difficult to get to the conclusion that freedom of disposal of time is a problem (which is what may be suggested by this argument to a limit). We can also think on a technological argument: even if time is a physically continuous variable, there could be difficulties in accumulating it when disposing at irregular intervals. In this sense, those occupations that declare to be less dissatisfied with their leisure time turn out to be those that have a greater uncertainty on the distribution of their leisure time along the day.

Caring activities Another result is achieved for those people that are engaged in caring activities such that is impedes them get a paid work: they report smaller satisfaction with their leisure time. This result brings evidence on the characteristic of paid-work in the market as an economic good -being a desirable good- (recall that we have already controlled for the effect of working hours, for the effect of the intensity of this care and, more importantly, for the effect of income).

Household income Income has a significant effect on the satisfaction with leisure time. However, this effect is only significant for the level effect, whereas changes in income do not have a significant effect.

Durables and material goods For the ownership of durables, we do not find any significant effect on the satisfaction with leisure time (this happens with every empirical specification that we have tried). From a theoretical point of view, there is a complementarity relationship among durables to be combined with leisure time for the production of enjoyable experiences. However, no significant effect is found for the satisfaction statement.

The presence of companionable others We find out the relevance of social capital variables, which turn out to be one of the most relevant results.
General characteristics Overall results indicate that general regularities rule: age effect being no-monotone, and for the overall European sample, females being less satisfied (although for the Spanish sub-sample this is not applicable).

Final discussions and comparison with previous work on leisure satisfaction.

In this paper we have discussed how satisfaction with leisure time is determined, taking into account differences in time allocation and personal divergences for the availability of other personal and social resources that might enter into the leisure commodity production.
For our empirical analysis, we have taken into account in a direct way the number of hours spent on labour in the market, as well as time devoted to house production activities. This last issue entered in an indirect way, given that there is not a reported measure on the total number of hours devoted to home production issues, such as childcare. As expected, we found a negative relationship between the number of working hours, the .Furthermore

One overall conclusion:

A neutral definition of leisure time will be a final-oriented one: it is the time that individuals can freely decide to assign to enjoyable and intrinsically motivating activities (which is one of the definitions of leisure -- the 2nd one -- leisure as activity). Thus,

Individuals in contemporary societies are engaged in activities regulated by contracts: the main example is work, but this would also apply to many non-market activities (though remember that the first-person-criterion does not apply for the production of those commodities, just for the consumption of them).

This introduces a more complicated time restriction in the individual maximization problem. It is not only to optimally allocate time to alternative activities (3 such activities in the most simple theoretical setting); but these alternatives would now be bounded between an upper and a lower bound. This fact may also be related to threshold effects in the production function of the main commodities: if less that a given quantity of time is available for the production of the commodity leisure at a given moment of time, then it is useless.

Further research

On of the main pitfalls of our empirical analysis, as we have pointed out, is the lack of a measurement of the amount od time that the individual dedicates to leisure activities.
New insights can be gained if information such as the one contained in Time Use Surveys is introduced in living conditions studies that include questions that allow a subjective well-being study. Hammermesh, 2009 points out a series of research areas

Ateca-Amestoy, V., Vera-Toscano, E. and R. Serrano-del-Rosal (2004), The Leisure Experience, IESA WP Series 18-04. Accepted for publication in The Journal of Socio Economics.
beard : Beard, J.G. and M.G. Ragheb (1980), "Measuring Leisure Satisfaction", Journal of Leisure Research 12 (1), 20--33.
becker : Becker, G. (1965), "A Theory of the Allocation of Time", Economic Journal 75, 493--517.
becker1990 : Becker, G.S. (1990), The Economic Approach to Human Behavior. The University of Chicago Press, Chicago.
becker 1996 : Becker, G.S. (1996), Accounting for Tastes. Harvard University Press, Harvard.
blanchfloweroswald : Blanchflower, D. G. and Oswald, A. J. (2004), "Well-being Over Time in Britain and the USA", Journal of Public Economics 88, 1359--1386.
bonke1 : Bonke, J., Deding M. and M. Lausten (2004), Time and Money --Are they Substitutes?, Globalization, Society and Welfare WP 11 2004, The Danish National Institute of Social Research.
bonke2 : Bonke, J., Deding M. and M. Lausten (2006), Time and Money. Subtitutes in Real Terms and Complements in Satisfactions, Journal of Happiness Studies 10(2), 113-131.
bourdieu : Bourdieu, P. (1988), La Distinción. Criterios y Bases Sociales del Gusto. Translated by Ruiz de Elvira, M.C. Taurus, Madrid.
bourdieu2 : Bourdieu P (1984) Distinction: a social critique of the judgement of taste London: Routledge & Kegan Paul.
campbell : Campbell, A., Converse, P.E., and W.L. Rodgers (1976), The Quality of American Life. Russell Sage Foundation, New York.
clark : Clark, A. E. (1997), "Job satisfaction and gender: Why are women so happy at work?", Labour Economics 4 (4), 341--372.
clarkoswald : Clark, A.E. and A. J. Oswald (1994), "Unhappiness and Unemployment", Economic Journal 104, 648--659.
coleman : Coleman, J. C. (1988), "Social capital in the creation of human capital", American Journal of Sociology 94, 95--120.
donovan : Donovan, N., Halpern, D., and R. Sargeant (2002), Life Satisfaction: the state of knowledge and implications for government. Strategy Unit, London.
espingandersen2002 : Esping-Andersen, G. (2002), Why We Need a New Welfare State. Oxford University Press, Oxford.
ferrer : Ferrer-i-Carbonell, A. (2002), Subjective Questions to Measure Welfare and Well-Being: A survey. Tinbergen Institute Discussion Paper 2002-020/3, Amsterdam.
ferrerfrijters : Ferrer-i-Carbonell, A. and P. Frijters (2004), "How important is methodology for the estimates of the determinants of happiness?", Economic Journal 114, 641--659.
Gershuny2005#7 : Gershuny, J. (2005), What do we do in Post-industrial Society? The nature of work and leisure time in the 21st Century, Working Papers of the Institute for Social and Economic Research, paper 2005-7. Colchester: University of Essex.
Gershuny2005#8 : Gershuny, J. (2005), Time Allocation and the Comprehensive Accounting of Economic Activity, Working Papers of the Institute for Social and Economic Research, paper 2005-8. Colchester: University of Essex.
Gershuny2003#3 : Gershuny, J. (2003), Time, trough the Lifecourse, in the Family, Working Papers of the Institute for Social and Economic Research, paper 2003-3. Colchester: University of Essex, and in the Blackwell Companion to the Sociology of Families,edited by Jacqueline Scott, Judith K. Treas and Martin P. Richards. Oxford: Blackwell Publishing (2003).
gershuny2000 : Gershuny, J. (2000), Changing Times. Work and Leisure in Postindustrial Society. Oxford University Press.
gronauham5 : Gronau, R. and D.S. Hamermesh (2006), "Time vs. Goods: The Value of Measuring Household Production Technologies", Review of Income and Wealth 52(1), 1-16.
greene : Greene, W. (1990), Econometric Analysis. MacMillan, New York.
goodin, rice, bittman, sanders : Goodin, R.E., Rice, J.M., Bittman, M., and P. Sanders (2005), "The Time Pressure Illusion: Discretionary Time versus Free Time", Social Indicators Research 73: 43 - 70.
hamermesh : Hamermesh, D. (2009), "It's Time to "Do Economics" with Time-Use Data", Social Indicators Research 93: 65-68.
haavio : Haavio-Mannila, E. (1971), "Satisfaction with family, work, leisure and life among men and women", Human Relations 24, 585 - 601.
jenkins&osberg : Jenkins, S.P., and L. Osberg (2003), Nobody to Play With? The Implications of Leisure Coordination, IZA DP. No 850, IZA, Bonn.
kelly : Kelly, J.R. (1982), Leisure. Prentice-Hall, Englewood Cliffs, NJ.
Layard, R. (2006), Happiness: Lessons from a New Science, Penguin, London.
nazio&macinnes : Nazio, T. and J. MacInnes (forthc), "Time Stress, Well-being and the Double Burden" in G. Esping-Andersen (ed.), Family Formation and Family Dilemmas in Contemporary Europe, FBBVA.
Osberg, L. (2008), “Leisure” en The New Palgrave Dictionary of Economic, MacMillan.

putnam1 : Putnam, R. D. (1993), Making democracy work. Civic traditions in modern Italy. Princeton NJ, Princeton University Press.
putnam2 : Putnam, R. D. (1995), "Bowling alone: America's declining social capital", The Journal of Democracy 6 (1): 65-78.
putnam3 : Putnam, R. D. (2000), Bowling alone. The collapse and revival of American community. New York, Simon and Schuster.
rice, goodin, parpo : Rice, J.M., Goodin, R.E., and A. Parpo (2006), The Temporal Welfare State: a Cross-national Comparison, The Levy Economics Institute WP 449, The Levy Economics Institute of Bard College, NY.
vanpraag1 : Van Praag, B.M.S. and A. Ferrer-I-Carbonell (2004), Happiness Quantified. A Satisfaction Calculus Approach. Oxford University Press. Oxford.
vanpraag2 : Van Praag, B.M.S., Frijters P. and A. Ferrer-i-Carbonell (2003), "The anatomy of well-being", Journal of Economic Behavior and Organization 51, 29--49.
veblen ocio : Veblen, T. (1965 - reed.), The Theory of the Leisure Class. Augustus M. Kelley, New York.
warde&tampubolon : Warde, A. and G. Tampubolon (2002), "Social Capital, networks and leisure consumption", The Sociological Review, 50 (2), 155-180
winkelmann2004 : Winkelmann, R. (2004), Subjective Well-Being and the Family: Results from an Ordered Probit Model with Multiple Random Effects, IZA DP. No 1016, IZA, Bonn.
zeckhauser : Zeckhauser, R. (1973) "Time as the Ultimate Source of Utility", Quarterly Journal of Economics, 87 (4), pp.668-75.