Introduction

With the purpose of explaining and predicting the behavior of individuals, economists established the theoretical concept of utility, which originally represented the level of satisfaction, happiness or wellbeing of the individual. However, the use of this term was modified until expressing individual preferences. This change also transformed the measurement of wellbeing. Currently, wellbeing is measured based on the preferences that it can satisfy (^{Ferrer-i-Carbonell, 2002}). It is assumed that individuals, given their restrictions, make rational decisions that maximize their wellbeing. This implies that decisions of consumption generate the maximum level of wellbeing that individuals can reach within the options available to them. Since the number of options available increases as the income increases, a higher income should represent a higher level of wellbeing (^{Diener, 2009a}). Therefore, income has become the measurement of wellbeing most frequently used. However, in recent years it has been questioned whether income is the most appropriate measurement of wellbeing because it offers a limited vision of the individual's wellbeing. Thus, other approaches have been developed to present the individual's wellbeing as a multidimensional and subjective phenomenon.

The objective of this research was to compare two types of indicators of wellbeing through a case study in Mexican rural communities. In particular, the study has centered in evaluating the wellbeing of individuals under conditions of poverty. For this purpose, the approaches of income poverty and subjective wellbeing were considered.

Traditionally, poverty has been measured through the individual's income³, thus assuming that his/her wellbeing is highly correlated to his/her income. This way of measuring poverty has the advantage that obtaining the information necessary to generate an index is relatively easy, in addition to the fact that comparisons between countries are simplified with this measurement. However, if poverty is considered to be a situation of low wellbeing, then the approach of measuring income implicitly equates wellbeing with the ability to acquire a set of goods that a group of experts judges to be valuable for individuals. Likewise, the approximation to poverty through income ignores that some basic necessities cannot be purchased, because they are not available in the market or are provided by the State. Lastly, the approach of income poverty assumes that individuals use their income to maximize their wellbeing. This contrasts with the empirical evidence that shows that people are frequently wrong when predicting the effect that their decisions will have on their happiness (^{Dunn et al., 2011}).

In recent years the unidimensional measurement of poverty has been substituted or complemented by the multidimensional measurement. Multidimensional poverty is an approximation that recognizes that poverty entails different dimensions or scopes that cannot be grasped exclusively by income. The multidimensional perspective considers poverty as a group of shortages in a group of predefined dominions. Since 2009, the National Council for Evaluation of Social Development Policy (Consejo Nacional de Evaluación de la Política de Desarrollo Social, CONEVAL) adopted this approach for the official measurement of poverty in México. In the methodology developed by this organization, it is considered that a person is poor if his/her income is insufficient to acquire the goods and services that cover his/her basic needs and if he/ she does not exercise at least one of the rights for social development (education, food, housing, health and social security) (^{CONEVAL, 2009b}). Although a multidimensional approximation represents an advance with regard to measuring by income, under this approach it is a group of experts and not the individual who determines the relevant dimensions for measuring poverty. In addition, building a poverty index implies assigning weight to each dimension, which may not coincide with individuals' priorities.

An alternative to complement the income and multidimensional poverty indicators, as well as other indicators considered objective, is the approach of subjective wellbeing. The term subjective wellbeing is used to refer to the evaluation that an individual makes of the quality of his/her life. It is measured by asking the individual to evaluate the quality of his/ her daily experiences (emotional wellbeing), or else to perform a global assessment of satisfaction with his/ her life in general (life satisfaction). This method can also be used to determine the individual's satisfaction with specific areas, such as satisfaction with his/her health, family life or work.

Psychology, for more than forty years, has used these self-reports for satisfaction and happiness as a valid method to evaluate individuals' wellbeing (^{Diener et al., 1999}). However, it has not been until recent years that Economy has incorporated elements from the literature of subjective wellbeing to the study and development of measurements of wellbeing. According to ^{Rojas (2007)}, one of the main methodological advantages of the approach of subjective wellbeing is that it recognizes that a person's wellbeing is essentially subjective, and therefore, the individual is the best authority to evaluate it. Thus, since it is declared by the person, it constitutes an integral evaluation of it, instead of an evaluation of some aspects of the individual's life. Additionally, this approach admits that it is he/she who determines what are the relevant factors for his/her wellbeing and, stemming from an inference analysis, allows identifying the variables related to wellbeing.

Through the analysis of wellbeing reports, economists and psychologists have deciphered some of the determinants of subjective wellbeing. It has been found that most of the variation in these reports is explained by differences in personality (^{Diener, 2000}; ^{Kahneman and Krueger, 2006}). Although the literature suggests that the wellbeing reports are stable throughout time, it has been found that exogenous events in the life of the individual (positive and negative) have a significant effect on subjective wellbeing. Although this effect tends to disappear after three to six months since the event takes place (^{Suh et al., 1996}), it tends to be greater for negative events (^{Gomez et al., 2009}).

The sociodemographic factors and life circumstances of the individual (including his/her income) have a weak correlation with subjective wellbeing. In transversal studies, the correlation between income and life satisfaction has been found to be positive and statistically significant, although relatively small. This correlation tends to be higher in poor countries, as well as among poor people in rich countries (^{Veenhoven, 1996}; ^{Argyle, 1999}; ^{Diener et al., 1999}; ^{Diener, 2000}; ^{Ferrer-i-Carbonell, 2002}; ^{Rojas, 2009}).

Different hypotheses have been proposed to explain the weak effect of income on the subjective wellbeing. First, it is argued that relative income (the individual's income compared to his/her reference group) is more important to explain life satisfaction than the income level. The empirical evidence is consistent with this hypothesis: the correlation between relative income and life satisfaction is higher than the correlation between income level and life satisfaction (^{Argyle, 1999}; ^{Ferrer-i-Carbonell, 2002}; ^{Rojas, 2009}).

A second hypothesis proposes that individuals adapt to material goods. After some time of having increased their income, they adjust to this new level of life. This provokes for their life satisfaction to return to its original level. In the literature, this phenomenon is known as hedonic treadmill. Aspirations also adapt; with an increase in possessions, the number of things desired increases, gradually. This new gap between what is owned and what is desired prevents the level of satisfaction from increasing (^{Easterlin, 2006}), and therefore, for the correlation between income and satisfaction to be small.

Finally, a third hypothesis connects subjective wellbeing to activities on which people use their time. As income increases, the use of time is destined mostly to activities associated to greater tension and stress (for example, longer workdays or an increase in time commuting to the workplace) and the time devoted to activities associated with greater happiness is reduced (e.g., leisure activities) (^{Kahneman et al., 2006}). This could provoke for an increase in income not to be reflected in an increase in life satisfaction.

This study stems from the hypothesis that measurements based exclusively on income are not the best approximation to measure the wellbeing that people experience. To evaluate this hypothesis, information from a set of households interviewed in four rural communities located in the states of Jalisco and Michoacán is used. The results for the case study confirm the hypothesis and show that, as several authors have suggested (^{Schwarz and Strack, 1999}; ^{Veenhoven, 2002}; ^{Kahneman and Krueger, 2006}; ^{Diener, 2009b}), the approach of subjective wellbeing can be a good complement to the traditional measurements of welfare.

Materials and Methods

The data used in this research come from a survey that was applied in four rural communities, Yerbabuena and Ayotitlán in Jalisco, and San Pedro Tarímbaro and La Salud in Michoacán. To select the communities, first they were chosen randomly, based on the marginalization index from 2005 (^{CONAPO, 2006}), one municipality with a very high or high degree of marginalization, and one with medium, low or very low level of marginalization in each state. The municipality had, in addition, to be adjacent to the priority terrestrial regions where the biosphere reserves of Sierra de Manantlán and Sierra de Chincua are located. For the community sample selection, all the communities with less than 2500 residents inside each of the four municipalities selected were taken as the sample framework, and community was selected per municipality.

In each one of the four rural communities, 3 households were surveyed, giving a total of 14 households. To carry out the survey, the households were selected randomly, guaranteeing a complete spatial coverage of the community. Two biannual surveys were done: July 2008 and January 2009. With the first survey, information about the fir semester of 2008 was obtained, and with the second, information about the second semester of the same year was obtained. Of the 140 households surveyed in the first semester of 2008, 134 were able to \ contacted again during the second semester. The results of the survey provide detailed information about sociodemographic characteristics, production assets and income by sources for each one of the households.

Measurement of income poverty

The first step to identify the population i situation of poverty is to fix a poverty line that separates the population that is poor from that which is not. For this purpose, a minimum level of wellbeing is established, which is generally defined as the income that allows the household to cover its basic needs, such as food and elemental housing Once the poverty line is defined, the next step is to obtain a measurement of aggregation. In this study the measurement used is the poverty index known as FGT, which was proposed by Foster, Greer and Thorbecke (1984) and which is calculated in the following manner:

where z represents the poverty line, y _i is the individual's income, n is the size of the population and q is the number of poor households (for which y _i ≤ z).

In this index, α is a measurement of aversion to poverty; that is, with a higher α the FGT _α index gives greater weight to individuals that are the poorest among the poor. Commonly, the FGT index is reported as a measurement of poverty for α=0, 1 and 2. When α=0, FGT ₀ calculates the proportion of people under the poverty line; until today this is the unidimensional measurement of poverty most commonly used at the international level. When α=1 , FGT ₁ calculates the normalized poverty gap, which reflects how far the average income of the poor is from the poverty line. In this way, FGT ₁ is sensitive to changes in the income of the poor. Finally, when α=2, FGT ₂ is a measurement of the severity of poverty, since individuals whose income is farther away from the poverty line have a higher weight in the index.

Measurement of subjective wellbeing

The questionnaire that was used includes a question about satisfaction with life in general. This allows us to compare the results of the traditional measurement of poverty with the one obtained by following the subjective wellbeing approach. In addition, given the low relation that has been found in the literature between socioeconomic variables and the levels of satisfaction, a series of questions about satisfaction with specific areas of life were included in the questionnaire. This is in agreement with the approach of areas or dominions of life, which establishes that the life of a person can be seen as the construction of specific scopes and life satisfaction depends on the satisfaction in these areas. Likewise, it establishes that satisfaction in areas of life can be explained through socioeconomic variables.

In principle, there are countless activities and spheres that can be seen as partitions in the life of the individual. ^{Cummins (1996)} finds that the dominions that are consistently relevant when explaining life satisfaction are: material wellbeing (economic and financial situation, income), health, work, intimacy (relationship with family, friends and spouse), security, community and emotional wellbeing (leisure and free time). Table 1 describes the questions that are included in the questionnaire to measure these dominions. The possible answers for the questions of subjective wellbeing are: extremely dissatisfied, very dissatisfied, dissatisfied, neither satisfied nor dissatisfied, satisfied, very satisfied, and extremely satisfied.

Table 1 Description of questions about subjective welbeing. 

Ámbito	Pregunta
	¿Qué tan satisfecho está usted con...
Vida en general	su vida en general?
Salud	su salud actual?
Económico	la situación económica en su hogar?
Ocupación	la ocupación/empleo que tiene (o la actividad doméstica que realiza)?
Familia	la relación con el resto de su familia?
Amistad	la relación con sus amigos y vecinos?
Tiempo	el tiempo de que dispone para hacer lo que a usted le gusta (pasatiempos, actividades personales)?
Comunidad	los servicios públicos en el lugar donde vive? (recolección de basura, seguridad, salubridad, transporte, alumbrado, calles, etcétera).

The first thing we did in this study to evaluate the relationship between income and subjective wellbeing is to contrast the indicators of wellbeing from income poverty (which is associated with not being capable to acquire a basic food basket) and subjective poverty (which represents a situation in which the individual experiments a low level of life satisfaction).

Then we analyzed with econometrics the importance of economic satisfaction in the life satisfaction of a person. This is done through the estimation of ordered probit models and following the methodology by ^{Rojas (2008)}. For this, the following relation is proposed:

(1)

where S _i is life satisfaction in general for the individual i and Sa _i (health), E _i (economy), Tr _i (work), F _i (family), A _i (friendship), T _i (time), C _i (community) refer to the satisfaction in these areas for the same individual. If in this estimation the economic dominion does not explain life satisfaction significantly, there will be evidence to support the hypothesis that income is not the best approach to measure the wellbeing that individuals experience.

Lastly, we studied the correlation between income and satisfaction in the areas. If income is highly related to satisfaction in the areas mentioned, then the measurements based on income can reflect, in part, the subjective wellbeing of the individual, for they would affect the variables that explain it. Thus, the following relation is used:

(2)

where: S _i ^k is the satisfaction on dominion k of the individual i, and Y _i is the natural logarithm of income per capita for the household.

Results and Discussion

Conclusions

In this study, two approaches to measure welfare were evaluated. We departed from the following hypothesis: the approach that is used most frequently (income poverty), presents an incomplete image of the wellbeing that the individuals experience. To complement it, the approach of subjective wellbeing was proposed.

Income poverty, although relatively easy to measure, represents a limited vision of the individual's wellbeing. Income can hardly capture the diverse factors that determine an individual's wellbeing. The approach of subjective wellbeing offers a broader vision of welfare, since it considers that wellbeing is an essentially subjective concept (with which the individual is the best authority to evaluate his/her wellbeing).

The approach of income poverty showed that the rural households in the study case have a very low level of wellbeing, because the poverty levels in these are higher than the national average. In contrast, when the reports on satisfaction were used as a measure of wellbeing, it was found that individuals in these communities enjoy a relatively high wellbeing. This suggests that the wellbeing that an individual experiences is associated with different factors that both the income and other objective indicators are incapable of capturing.

Likewise, the approach of subjective wellbeing, through an analysis of inference, allows identifying the sources of wellbeing of a person. Although empirically some systematic correlations have been found between reports of satisfaction and objective and personality variables, further research is still necessary to establish the factors that affect the subjective wellbeing and its determinants. Therefore, this study additionally sought to contribute to the understanding of these determinants, in particular the relationship there is between income and life satisfaction.

The literature suggests that income, at least in absolute terms, has little relevance in explaining the subjective wellbeing. The results from this study confirm this, since it was found that the economic scope is not significant when explaining the individual's satisfaction with his/her life in general. Likewise, income was also not related to the principal areas of life, so it provides little information about wellbeing, at least in terms of the perception that individuals have of it. If income were to be used as the sole guide for public policy, important information about the wellbeing that individuals experience would be omitted. Under this context, the approach of subjective wellbeing is a good complement to the traditional measurements of welfare.