Details and requirements are attached below.

According to the economic definition of well-being, higher levels of income are associated with higher levels of well-being through greater levels of material consumption. Taking into consideration this definition, it is worth asking just how important income is for an individual’s happiness? Existing studies in psychology have found a positive correlation between economic well-being (socioeconomic status) and subjective well-being (happiness). However, this positive relationship is weak and a large percentage of human happiness remains unexplained.

Although the mentioned studies make a good approximation of the existing relationship between income and happiness, the characteristics or specification of the function that generates the relationship between these two variables is often assumed to be linear and positive. The main objective of this paper is to investigate further the relationship between subjective and economic well-being. In this study different specifications and approaches are used to approximate the relationship between these variables.

An empirical analysis is made from the results of a survey conducted in two Mexican cities.

The investigation studies the impact of demographic, social, and economic variables on subjective well-being in Mexico. Several hypotheses are tested to identify the relationship between household income and individual well-being. It is found that income does not have a strong influence on neither well-being nor on the probability of happiness. However, people tend to overstress the impact that additional income would have on their subjective well-being.

This fact could explain the importance that people place on increasing their income level, and it could possibly explain the relative sense of dissatisfaction once a higher income level is achieved.

The relationship between income and the sense of basic need satisfaction is also explored.

A main assumption in economic theory suggests the existence of a direct relationship between these variables; however, empirical results show this relationship to be extremely weak. Results indicate that subjective well-being is positively related to the sense of basic need satisfaction but not to income.

Keywords: Subjective well-being, happiness, economics, Mexico.

1. Introduction Several studies have been dedicated to the topic of well-being. Most of them have been conducted in industrialized countries where income is distributed relatively more 2 equitatively and the population tends to be more homogeneous. This paper studies the relationship between subjective and economic well-being in Mexico, a country where the economic differences among the population are more marked.

According to the economic definition of well-being, higher levels of income are associated with higher levels of well-being through greater consumption levels. Taking into consideration this definition, it is worth asking how important income is to an individual’s happiness? Existing studies in psychology have found a positive correlation between economic well-being (socioeconomic status) and subjective well- being (happiness). However, this positive relationship is weak and a large percentage of human happiness remains unexplained.

Although the mentioned studies make a good approximation of the existing relationship between income and happiness, the characteristics or specification of the function that generates the relation between these two variables is often assumed to be linear and positive. The main objective of this paper is to investigate further the study of the relationship between subjective and economic well-being. In this study different specifications and approaches are used to approximate the relationship between these variables.

An empirical analysis is made from the results of a survey taken in two Mexican cities. The investigation studies the impact of demographic, social, and economic variables on subjective well-being in Mexico. Several hypotheses are tested to identify the relationship between household income and individual well-being. It was discovered that income does not have a strong influence on neither well-being nor in the probability of happiness. However, people tend to overstress the impact that additional income would have on their subjective well-being. This fact could explain 3 the importance that people place on increasing their income level, and it could possibly explain the relative sense of dissatisfaction once a higher income level is achieved.

The relationship between income and the sense of basic need satisfaction is also explored. A main assumption in economic theory suggests the existence of a direct relationship between these variables; however, empirical results show this relationship to be extremely weak. Results indicate that subjective well-being is positively related to the sense of basic need satisfaction but not to income.

2. Relevant Literature 2.1 Economic theory and well-being According to the economic definition of well-being, higher levels of income are associated with higher levels of well-being. As income increases a greater number of needs are satisfied (due to an increase in consumption) and a higher standard of well- being is attained. The preoccupation of economists to increase percapita income arises from the economic definition of well-being. According to the theory of utility, individuals are able to express preferences over commodity bundles. With the assumption of several axioms, these bundles can be ranked (the higher the bundle is in the ranking, the more preferred it is by the individual) and a utility function can be constructed to reflect the ordering of the bundles. Therefore, an increase in the utility level of the individual is associated with attaining more preferred bundles. Thus, a rational individual pursues the maximization of his/her utility within his/her constraints, the budget constraint being the most important factor. From utility theory it follows that an individual’s main objective is to achieve higher utility levels and his/her main tool is to expand the budget constraint. (Needs are unlimited whereas 4 resources to satisfy them are limited). Microeconomic theory derives the conclusion that human well-being increases with income. Following this conclusion, economic growth becomes an important objective of economic policy in any country. Economists associate economic growth directly with well-being, and the success of a country and of its government is generally measured in terms of increments in percapita income.

2.2 Psychology and well-being The approach of psychology to the study of well-being is different from that of the economic approach. Psychologists have studied the physiological and emotional nature of human beings to explain and understand the sources that cause happiness or the lack of it. Psychologists have tried to identify what makes an individual happy: is happiness a trait or a state? Do people stay happy? What characteristics do happy people have?

What influences people’s happiness? Is happiness related to factors such as gender, race, and age? Are education, income, type of employment and civil status related to happiness? Information from surveys and statistical techniques are combined to provide answers to these questions.

Psychologists often use the term subjective well-being to refer to happiness. 1 There are many different definitions of happiness. For instance, Mullis (1990) states that subjective well-being is related to personal goals, life expectations and the means to attain them. Diener and Diener (1996) define happiness as the cognoscitive and affective evaluation by the individual of his/her life. The cognoscitive evaluation refers to long-run life objectives satisfaction, while the affective evaluation is associated with daily emotions experienced by the individual (Venhoveen, 1994).

2.3 Theories about subjective well-being and economic variables The relationship between subjective well-being (happiness) and economic status 5 (income levels) has been explored theoretically and empirically in the field of psychology. The following are some of the theories that discuss this relationship.

Relative Theory. Easterlin (1974) sustains that the impact of income on subjective well-being depends on standards that change over time according to the individual’s expectations and social comparisons. Thus, factors such as the relationship between the present and former economic situation and the individual’s wealth in relation to that of reference individuals (Meadow, 1992) could influence a person’s happiness regardless of his/her income level (Diener and Diener, 1996; Parducci, 1968, 1982).

Absolute Theory. Venhoveen (1988, 1991) assumes a relationship between basic needs satisfaction and subjective well-being. People with higher income levels can easily satisfy their basic needs (food, housing, health, etc.) and, therefore, attain a higher subjective well-being. Venhoveen’s approach suggests the existence of a threshold level where income’s impact on subjective well-being is not important.

Adaptation Theory. Brickman et al. (1978) focus on the emotional capabilities of individuals to adapt to positive and negative events. Thus, individuals with higher adaptation capabilities tend to be happier (even in situations of low-income levels).

Aspiration Theory. According to this theory the degree of satisfaction/dissatisfaction experienced by a person is related to the ratio of his/her satisfied desires to his/her total desires. Individuals who believe that their desires are fully satisfied tend to be happier than individuals who think they have unsatisfied desires, regardless of their income levels. This approach to the concept of happiness takes into consideration not only the degree of satisfied needs (which is presumably related to income) but also the individual’s total desires (which are also presumably related to income).

2.4 Measuring subjective well-being 6 In order to study well-being economists use indicators such as income and consumption, presuming that these are correlated to well-being. Psychologists, on the other hand, take into account the individual’s own manifestation of his/her subjective well-being. The study of happiness often uses information from surveys, turning the measuring of this variable into one of the main problems. Many scales have been used to measure subjective well-being. Most of these ask the individuals to report how happy or satisfied they feel about their lives in general. These scales offer a list of options, which are ranked according to the levels of happiness. The following are some of the scales found throughout the relevant literature:

Three-point happy scale. This scale asks the following question: Taking everything into consideration, how happy are you? And provides three answering options: very happy, more or less happy, unhappy (Palomar, 1997). Gurin, Veroff, and Feld (1960) use a variation of this scale.

Life as a Whole Index. Individuals are asked the question: How happy are you with your life in general? The individual chooses one out of seven options: terrible, unhappy, generally unsatisfied, neutral, generally satisfied, satisfied, and delighted.

This scale is also known as the “D-T Scale” (delighted-terrible) and is widely used in psychological studies. Andrews and Whitey (1976).

Self-fulfillment Index. It uses the D-T scale, but asks a different set of questions:

How do you feel about your sense of purpose in life? How do you feel about your achievements in life? and How do you feel about the degree of your success in life?

(Palomar, 1997).

Other scales such as the Self-Anchoring Ladder scale created by Cantril (1965) and Fordyce’s Scale (1977) offer the individual more than seven answering options. In 7 addition, Fordyce asks the individual to report what percentage of time he/she feels happy, unhappy or neutral. 2.5 Subjective well-being and economic theory: former studies Many studies have been conducted in the field of psychology to determine if a person’s economic well-being (socioeconomic status) and subjective well-being (happiness) are related. Psychologists not only have tried to identify the existing relationship between income and happiness but they have also tried to determine the nature of this relationship and the importance of economic well-being influence over subjective well-being. In most studies a positive relationship is found (Mullis, 1990; Venhoveen, 1988, 1991; Headey, 1991; Douthitt et al., 1992). However, the results show an extremely weak relationship. The explanatory capability of economic variables is so low (Diener et al., 1992; Mullis, 1990; Douthitt et al., 1991), making it possible to conclude that economic variables are not an important component of subjective well-being. 2 Demographic variables such as gender, age, and civil status have also been tested in an attempt to explain subjective well-being. However, they too do not explain happiness satisfactorily (Andrews and Whitey, 1976; Campbell et al., 1976).

Results from this study reinforce findings from the previous work here mentioned.

It is possible to state that income does not exert a strong influence on well-being; a large fraction of the variability in happiness is not explained either by economic nor by social and demographic factors.

Neither economic nor demographic variables seem to have an important impact on subjective well-being. So, what explains people’s happiness? Headey (1991) mentions four groups of explanatory variables: personality (traits); health; interpersonal 8 networks; and demographic, economic and related variables. Structural models using variables from the four groups reach explanatory levels between 70% to 80% of the total variability of subjective well-being (Abbey and Andrews, 1985; Andrews and Whitey, 1976). 3 Following the results discussed above it can be concluded that people’s happiness is better explained by innate characteristics and by the personality traits of each individual.

3. Methodology and Results 3.1 The survey A survey was conducted in Monterrey and Puebla, the third and fourth largest cities in Mexico. Monterrey is located relatively close to the border with the United States and has a large industrial base. Puebla is near Mexico City and has an important textile industry. 4 A questionnaire was designed to collect information from a sample of individuals mostly selected at random. 5 However, the process was to some extent controlled to balance the number of representatives in relevant categories, such as: household income, happiness and gender. 339 questionnaires were properly completed, the sample size can be considered acceptable for inference in urban Mexican cities. The survey gathered information regarding the following quantitative and qualitative variables:

Demographic and Social Variables: education, age, gender, civil status, religion and participation in church activities, health condition, working situation and nature of job.

These variables are categorical.

Economic Variables: current household income, former economic level, future 9 income expectations, perception of satisfied material needs, concern for the family’s economic situation, and house ownership.

Subjective Well-Being: a combination of Fordyce’s scale and of the D-T scale by Andrews and Whitey (1976) was used to measure well-being in this study. A nine- option scale was constructed with adjectives associated to each option in an attempt to homogenize the answers. 6 The following are the scale’s answering options: extremely happy, very happy, happy, satisfied, generally satisfied, neutral, generally unsatisfied, unhappy and very unhappy. The answer was given a value between 1 and 9, where 1 was assigned to the lowest level of happiness and 9 to the highest.

Additional question: Individuals were asked to answer the following hypothetical question: If you had twice as much income as you currently have, how happy do you think you would be? The same nine-option scale was given.

3.2 General overview The distribution of the 339 respondents in each of the categories used in the regressions is summarized in Tables I to IV. Table I shows the distribution of respondents along the categories associated to the happiness variable. Results indicate that, all things considered, most individuals feel happy. The categories happy, very happy and extremely happy comprise 65% of the respondents.

(INSERT TABLE I) Table II illustrates the distribution of respondents according to household income.

Of the sampled individuals, a total of 37% receive a monthly income of less than $160.00 US dollars; 19% under $80.00 US dollars and only 19% make more than $1000.00 US dollars.

(INSERT TABLE II) 10 Table III shows the distribution of respondents according to economic variables.

Only 14% of the polled individuals consider their income is enough to satisfy all material needs. Categories related to future income expectations reveal that 60% of the respondents are optimistic about their future economic situation; people under 30 years old mostly comprise this fraction.

(INSERT TABLE III) Table IV illustrates the distribution of respondents according to different social and demographic variables. Categories related to religion show that 88% of the people have a religion; however, only 38% attend church more than once a week and merely 22% take part in church activities.

(INSERT TABLE IV) 3.3 Explanatory power of each variable Simple regression analysis was used to test the influence of each independent (explanatory) variable on happiness (dependent variable). The statistic R-square reflects the percentage of the variability of happiness that is explained by each independent variable. Table V shows the results for variables of economic nature and Table VI shows the results for social and demographic variables.

(INSERT TABLE V) Among the economic variables considered it was found that Income per se has little explanatory power: less than 5% of the variability in subjective well-being. It is worth mentioning that Perception of Material Needs Satisfied has an explanatory power of around 11%, much greater than income. It is possible that the influence of the economic situation over happiness is not a matter of income, but a matter of expectations. 11 (INSERT TABLE VI) Even though social and demographic variables are introduced in the present analysis to correct for any mispecification problem, it is worth stating that health, working conditions and age are the variables with the greatest explanatory power.

3.4 Impact of income on subjective well-being The economic definition of well-being suggests a positive relationship between income and happiness. Higher income levels allow the satisfaction of a greater number of needs and therefore well-being is increased. A linear and positive relationship between income and happiness suggests that greater income levels always translate into greater well-being, regardless of wealth composition or the means by which a higher income is achieved. A multiple regression analysis was carried out to test the impact of income on subjective well-being. Table VII shows the results of this analysis. (INSERT TABLE VII) The t-test indicates that the null hypothesis (income has no impact on subjective well-being) cannot be rejected.

7 Hence, it is impossible to affirm that income is an important variable in people’s happiness. Even if the null hypothesis were accepted, the coefficient’s value of the variable is so small that an increase in income would translate into an insignificant increase in happiness. 8 Many social and demographic variables, such as health, seem to have a stronger impact than income on subjective well-being.

Venhoveen's Absolute Theory (1988, 1991) suggests a nonlinear relationship between income and subjective well-being. According to this theory, beyond a certain income level the influence of income on people’s happiness tends to diminish, because 12 the satisfaction of basic needs is no longer a problem. To test this theory, a multiple regression was run with the income variable expressed in logarithmic terms. The results from this analysis are shown in Table VIII. In the case of Mexico, income has no impact on subjective well-being, not even when allowing for a decreasing impact of income on happiness.

(INSERT TABLE VIII) 3.5 Perception of satisfied material needs and subjective well-being Following the Aspirations Theory and based on the results from Table V, which show that Perception of Satisfied Material Needs has more explanatory power than Income on subjective well-being, a multiple regression was run taking the former variable as explanatory. It was found that the perception of satisfied material needs does have an impact on subjective well-being. The movement from the insufficient to the sufficient category represents an increase in happiness of almost one step in the scale of subjective well-being.

Income per se (the monthly amount of money that a household receives) is not as important in people’s happiness as is the perception of materials needs they believe they are able to satisfy with that income. Accordingly, it is possible that an individual with a low-income level is happy just with the fact that he/she considers that his/her needs are satisfied, whereas an individual with a high-income level feels unhappy because he/she considers that his/her income is insufficient. This is probably a matter of expectations; however, it is worth stating that the perception of satisfied material needs does not improve significantly with income. It suggests that a higher income implies greater needs and, consequently, it does not reduce the perception of needs not being satisfied. Results from this analysis are shown in Table IX. 13 (INSERT TABLE IX) 3.6 Is there a threshold income level?

The Absolute Theory also suggests the existence of a threshold income. Subjective well-being is lower below that threshold level because basic needs are not being satisfied. To test this hypothesis a multiple regression was run taking happiness as the independent variable and the variables shown in Table VII as explanatory; however, this time income was treated as a dummy variable. Different threshold levels were tested. It was found that an income of about $3200 Mexican pesos per month (approximately US$320 dollars) constitutes a threshold level. People tend to have lower happiness below this threshold income. Hence, even though the impact of income on happiness is weak, it was found that very low income levels are associated with less happiness.

3.7 Does income affect the probability of happiness? Logistic specification A different approach to studying the relationship between income and happiness is to assume the existence of a latent probability of being happy or unhappy, and to relate this probability to economic variables. Thus, the dependent variable (subjective well- being) was transformed into a dichotomous variable 9 and a logistic specification was used to run a regression. The objective was to test whether income has an impact on the probability of happiness. According to the specifications used in this exercise, it was found that income does not have an impact on the probability of being happy. Table X shows the results of this analysis.

(INSERT TABLE X) 3.8 Is the impact of income on subjective well-being overrated? 14 According to microeconomic theory and the economic definition of well-being, it is assumed that a higher income will translate into a greater well-being. Thus, it is expected that in the scenario where an individual sees himself/herself having a higher income, he/she also sees himself/herself being happier.

Individuals were asked the hypothetical question: “If you had twice as much income as you currently have, how happy do you think you would be?” The same subjective well-being scale was used. Thus, the objective of this exercise was to make a comparison between the happiness expected from individuals with lower income levels considering a hypothetical higher income, and the happiness of those individuals who really earn that higher income. If there is no difference between these two scenarios, then individuals correctly appraise the impact of an increase on income on their subjective well-being. However, if there is a difference then individuals overrate or underrate the impact of higher income levels on their happiness.

Regression pooling on both the hypothetical and the real observations was run and a structural-change test was applied. Table XI shows the results of this analysis. It was discovered that people tend to overrate the impact of higher income levels on their happiness. This is an important result for human behavior theory. If people act on the basis of their expected well-being then they may tend to give income a higher importance than it really has; 10 and once the higher income level is achieved people feel disappointed with the associated well-being. (INSERT TABLE XI) 3.9 Do past and expected future economic situations have an impact on well-being?

So far this study has only considered the relationship between current income and 15 happiness. However, individuals make decisions by considering not only their present situation but also what they expect from the future, thus they make intertemporal decisions. An individual’s happiness in relation to income is determined not only by his/her current situation, but also by his/her past situation and future expectations. Two additional regressions were run to study the impact of former economic situation and future income expectations on current subjective well-being. The explanatory variables shown in Table VII were also included in this regression. Table XII shows the results of this analysis. It was found that those individuals who experienced a decline in their economic situation tend to declare lower levels of well-being with respect to those who experienced an increase in their situation. It was also found that future income expectations do not have an impact on current well-being.

(INSERT TABLE XII) 4. Conclusions The following are the main findings from this investigation:

In the case of Mexico, the relationship between income and subjective well-being is extremely weak or completely nonexistent. The explanatory power of income on subjective well-being is less than 5%. This conclusion questions the widely spread belief about income being a key factor in people’s happiness.

The perception of satisfied material needs has a greater influence on subjective well-being than household income per se. As the perception of unsatisfied needs increases, people tend to be less happy.

The conclusions above substantiate the Aspirations Theory. It is not directly from income that people obtain their well-being, but from the perception that their needs are being satisfied. The correlation between income and the perception of satisfied 16 material needs is weak. Thus, it is possible to conclude that an increase in income is associated with an increase in desires; therefore, income and well-being are not expected to be related. However, perception of satisfied material needs and well-being are related.

A threshold income level of $3200 Mexican pesos ($320 US dollars) could exist.

People with income levels below this threshold are unable to satisfy their basic needs and, consequently, their well-being is lower.

People tend to overrate the impact of additional income on their happiness. Two possible explanations can be advanced. First, people ponder the impact of higher income levels on their happiness assuming that their desires will remain constant; however, when income really goes up, people realize that their desires have also increased. Thus, their perception of unsatisfied material needs does not decline and their happiness remains the same. This argument contradicts traditional consumer theory, which assumes that desires remain constant while income increases and that happiness depends on satisfied needs rather than on perception of unsatisfied needs. A possible second explanation states that people are unable to figure out the higher responsibilities that generally accompany a higher income level. In this case, economic theory wrongly separates the study of consumer behavior from the study of the income-leisure choice.

According to the logistic specification, it was found that income does not affect the probability of being happy.

It was found that those individuals who experienced a decline in their economic situation tended to be less happy in the present. It was also determined that future income expectations do not have an impact on current well-being. 17 All multiple regressions show that the joint explanatory power of economic, social and demographic variables is very low, less than 30% in all cases. A high percentage of people’s happiness remains unexplained. Thus, some variables not considered in this study, such as personality traits or innate characteristics, might have a stronger influence on an individual’s happiness.

As a final consideration, the idea of happiness or subjective well-being has been approached from different perspectives. Several fields of science have made an attempt to understand the origins of an individual’s happiness. Psychology has studied human physiological and emotional nature, while economics has focused on people’s consumption behavior. Both disciplines share the concern to understand individual well-being. However, they make different assumptions and work in different directions. An interdisciplinary approach to the study of happiness is recommended.

Under this perspective psychology and economics should complement each other. A better understanding of human physiological and emotional nature could translate into a more effective application of economic policies. It is worth noting that the concept or perception of well-being varies across different locations. It is also possible that this concept changes over time. Societies have cultural characteristics that make them unique. These characteristics turn the idea of happiness into a relative concept that depends on the evaluation of individuals of a certain region or country. 5. Bibliography Abbey, A. and Andrews, F. M.: 1985, “Modeling the Psychological Determinants of Life Quality”, Social Indicators Research 16: pp. 1-34. Andrews, F. M. and Withey, S. B.: 1976, Social Indicators of Well-being: Americans Perceptions of Life Quality, N. Y.: Plenum Press. 18 Brickman, P., Coates, D., and Janoff-Bulman, R.: 1978, “Lottery Winners and Accident Victims: Is Happiness Relative?” Journal of Personality and Social Psychology 36, pp.

917-27.

Campbell, A., P. E. Converse and W. L. Rodgers: 1976, The Quality of American Life (Russell Sage Foundation, New York).

Cantril, H.: 1965, The Pattern of Human Concerns (Rutgers University Press, New Brunswick, NJ).

Diener, Ed and Diener, Carol: 1996, “Most People are Happy”, Journal of Psychological Science, May Vol. 7 (3), pp. 181-185.

Diener, Ed, Sandvik, Ed, Seidlitz, Larry and Diener, Marissa: 1993, “The Relationship Between Income and Subjective Well-Being: Relative or Absolute?” Social Indicators Research 28: pp. 195-223. Douthitt, Robin A, MacDonald, Maurice and Mullis, Randolph: 1992, “The Relationship Between measures of Subjective and Economic Well-Being: A New Look”, Social Indicators Research 26: pp. 407-422.

Easterlin, R. A.: 1974, “Does Economic Growth Improve the Human Lot? Some Empirical Evidence”, en P. A. David y M. W. Reder (Eds.), Nations and Households in Economic Growth (Academic Press, New York), pp. 89-125.

Fordyce, M. W.: (1977), “The Happiness Measures: A Sixty-second Index of Emotional Well-Being and Mental Health”, unpublished manuscript.

Gurin, G., Veroff, J., y Feld: 1960, Americans View Their Mental Health, Basic Books, New York).

Heady, Bruce: 1991, “An Economic Model of Subjective Well-Being: Integrating Economic and Psychological Theories”, Social Indicators Research 28: pp. 97-116, 1993. 19 Meadow, H. L., Metzer, J. Rahtz, D. R. and Sirgy, M. J.: 1992, “A Life Satisfaction Measure Based on Judgement Theory”, Social Indicators Research 26: pp. 23-59.

Mullis, Randolph J.: 1990, “Measures of Economic Well-Being as Predictors of Psychological Well-being”, Social Indicators Research 26: pp. 119-135, 1992. Myers, David G. y Diener, Ed: 1995, “Who is Happy?” Journal of Psychological Science 6:

pp. 10-18.

Palomar, Joaquina: 1997, “La Pobreza y el Bienestar Subjetivo”, unpublished manuscript.

Parducci, A.: 1968, “The Relativism of Absolute Judgments”, Scientific American 219, pp.

84-90.

Parducci, A.: 1982, “Toward a Relational Theory of Happiness”, Paper presented at the 90 th Annual Convention of the American Psychological Association (August), Washington, DC.

Veenhoven, Ruth: 1988, "The Utility of Happiness?” Social Indicators Research 20: pp.

334-354.

Veenhoven, Ruth: 1991, "Is Happiness Relative?” Social Indicators Research 24: pp. 1-34.

Veenhoven, Ruth: 1994, “Is Happiness a Trait?” Social Indicators Research 32: pp. 101- 160.

Nicole Fuentes is a Graduate Student in Quantitative Methods in the Social Sciences at Columbia University. Mariano Rojas is professor of economics at Universidad de las Américas-Puebla, Mexico. 20 Table I Distribution of Respondents: Happiness Categories Percentage Extremely happy 9% Very happy 25% Happy 31% Satisfied 15% Generally satisfied 11% Neutral 6% Generally unsatisfied 2% Unhappy 0% Very unhappy 1% N = 339 100% 21 Table II Distribution of Respondents:

Monthly Household Income in US Dollars Categories Percentage No income 1% 0 to 40 8% 45 to 80 11% 85 to 160 17% 165 to 240 10% 245 to 395 12% 400 to 785 15% 790 to 1175 6% 1180 to 1565 5% 1570 to 1950 4% 2000 and more 10% N = 339 100% Approximate exchange rate: 10 pesos of 1998 = 1 US dollar.

22 Table III Distribution of Respondents: Economic Variables Categories Percentage Income and Satisfied Material Needs Income is enough to satisfy all needs 14% Income is enough to satisfy most needs 33% Income is enough to satisfy certain needs 45% Income is not enough to satisfy needs 9% House Ownership Owns a house 77% Rents a house 15% Borrows a house 7% Former Economic Situation Higher 23% Equal 32% Lower 44% Concern for Current Economic Situation Not concerned 8% Somewhat concerned 60% Always concerned 32% Future Income Expectations Higher 60% Equal 28% Lower 12% 23 Table IV Distribution of Respondents: Social and Demographic Variables Categories Percentage Categories Percentage Status Health Employed 61% Excellent 17% Unemployed 1% Good 60% Student 17% More or less 21% Housewife 20% Bad 1% Retired/Handicapped 1% Gender Nature of Job Male 50% White collar 36% Female 50% Blue collar/farmer 22% Civil Status Others 42% Single 47% Exercise Married 50% Does exercise 55% Divorced 2% Does not exercise 45% Widowed 1% Exercise days per week Age Everyday 16% 20 years old and under 17% Three days a week 24% 21 to 30 45% Once a week 14% 31 to 40 18% Never 45% 41 to 60 14% Religion 61 years old and over 4% Has a religion 88% Education Does not have a religion 12% Uneducated 2% Church Attendance per week Elementary School 12% More than once 6% Junior High 18% Once a week 32% High School 17% From time to time 47% Bachelor 33% Never 15% Graduate Degree 5% Participation in Church Activities Technical Degree 14% Does participate 20% Does not participate 80% 24 Table V Explanatory Power of Economic Variables on Subjective Well-Being Simple Regression Analysis Variable Category Coefficient Significance Prob. >t R-square Income Income .0474 Yes .0001 .0423 Perception of Sufficient 1.3744 Yes .0000 Satisfied Material Needs Sometimes .7225 Yes .0000 .1143 Insufficient 6.3646 Former Higher -.5230 Yes .0109 Economic Equal -.0309 No .8680 .0212 Situation Lower 6.9281 Concern Never 1.2182 Yes .0000 For Economic Sometimes .5389 Yes .0000 .0536 Situation Always 6.3679 Future Higher .5458 Yes .0335 Income expectations Equal .2458 No .3774 .0173 Lower 6.4 Note: variables in Italics indicate the reference category.

25 Table VI Explanatory Power of Social and Demographic Variables on Subjective Well-being Simple Regression Analysis Variable Category Coefficient Significance Prob. > t R-square Working Student 1.130 Yes .0000 Situation Worker .5229 Yes .0080 .0556 Housewife 6.2837 Nature of White .4586 Yes .0000 Job Blue 6.5 .0132 Other .3169 No .1331 Working Hours Less than 8 6.6595 Per Day 8 to 12 .2495 No .3367 .0033 More than 12 .1041 No .6691 Health Excellent 1.1857 Yes .0000 Condition Good .9987 Yes .0000 .0866 Regular 5.9866 Education Primary 6.1875 Junior High .4125 No .1456 High School .5711 Yes .0458 .00491 University .8482 Yes .0008 Graduate 1.4125 Yes .0011 Technical .7038 Yes .0201 Religion .4778 Yes .0561 .0107 Church More than 1 7.3636 Attendance Once -.5380 Indecision .1086 .0106 Per Week Sometimes -.6315 Yes .0582 Participation in church activities Participatio n .0917 No .6478 .0006 Exercise Three .2702 No .1174 Days per One -.5579 Yes .0219 .0329 Week None 6.7662 Gender Male .1604 No .3209 .0029 Civil Single 6.9625 Status Married -.2898 Indecision .0773 .0133 Divorced -.6897 No .1360 Age Less than 20 7.0655 20 to 30 -.0262 No .9048 31 to 40 -.7639 Yes .0037 .0526 41 to 60 -.5239 Indecision .0629 More than 60 -.9941 Yes .0217 Note: variables in Italics indicate the reference category. 26 Table VII Subjective Well-being as Independent Variable Linear Specification Multiple Regression Analysis Variable Category Coefficient Significance Prob. > t Intercept C1 6.4324 Working Student .9482 Yes .0014 Situation Worker 1.2003 Yes .0011 Nature of Job Blue -.9046 Yes .0104 White -.5501 No .1232 Health Excellent .8049 Yes .0016 Condition Good .7353 Yes .0001 Church Attendance Sometimes -.7417 Yes .0137 Per Week Once -.5426 Indecision .0857 Concern for Never .9272 Yes .0012 Economic Situation Sometimes .4334 Yes .0076 Exercise days Three .0008 No .9960 Per week One -.6001 Yes .0085 Civil Status Single/Divorced -.5200 Yes .0060 Age 31 to 40 -.7063 Yes .0010 More than 40 -.3358 No .1441 Income .0188 No .1485 Goodness of Fit: R-square = 0.2437 27 Table VIII Subjective well-being as Independent Variable Logarithmic Specification For Income Multiple Regression Analysis Variable Category Coefficient Significance Prob. > t Intercept C1 6.4804 Working Student .9329 Yes .0016 Situation Worker 1.1497 Yes .0018 Nature of Job Blue -.9040 Yes .0106 White -.5343 No .1339 Health Excellent .8121 Yes .0015 Condition Good .7440 Yes .0001 Church Attendance Sometimes -.7395 Yes .0142 per Week Once -.5129 Indecision .1043 Concern for Never .9199 Yes .0014 Economic Situation Sometimes .4106 Yes .0121 Civil Status Single/Divorced -.5180 Yes .0066 Age 31 to 40 -.7254 Yes .0007 More than 40 -.3440 No .1367 Exercise days Three .0081 No .9603 Per week One -.5619 Yes .0154 Income Logarithmic .0827 No .1956 Goodness of Fit: R-square = 0.2375 28 Table IX Subjective well-being as Independent Variable Focus on Perception of Satisfied Material Needs Multiple Regression Analysis Variable Category Coefficient Significance Prob. > t Intercept 5.1793 Working Student .8433 Yes .0042 Situation Worker 1.1489 Yes .0017 Nature of Job White -.6178 Indecision .0800 Blue -.8312 Yes .0180 Health Excellent .7195 Yes .0047 Condition Good .6756 Yes .0006 Education Primary/Secondary .3221 No .1779 University .3577 No .2210 Graduate .6370 No .1405 Religion .4369 Yes .0547 Perception of Satisfied Material Sufficient .8298 Yes .0008 Needs Sometimes .3862 Yes .0266 Concern Current Never .5345 Indecision .0759 Economic Situation Sometimes .2855 Indecision .0908 Exercise days Three .0174 No .9146 Per week One -.6102 Yes .0064 Civil Single/Divorced -.5627 Yes .0033 Age 31 to 40 -.5481 Yes .0118 More than 40 -.2029 No .4008 Goodness of Fit: R-square = 0.2671 29 Table X Dichotomous Subjective Well-Being as Independent Variable Logistic Specification: Latent Probability of being Happy Multiple Regression Analysis Variable Category Coefficient Significance Prob. > t Intercept C1 1.310780 Working Student .6732 No .1950 Situation Worker 1.2912 Yes .0474 Nature of Job Blue -1.1952 Yes .0551 White -.7414 No .2422 Health Excellent 1.0745 Yes .0161 Condition Good .6869 Yes .0327 Church Attendance Sometimes -1.5107 Yes .0240 Per Week Once -1.0729 No .1194 Concern for Never 1.6028 Yes .0085 Economic Situation Sometimes .6311 Yes .0195 Exercise days Three -.0728 No .7993 Per week One -1.1649 Yes .0024 Civil Status Single/Divorced -.8081 Yes .0175 Age 31 to 40 -.7254 Yes .0495 More than 40 -.5669 No .1552 Income .0140 No .5465 30 Table XI Appraisal of the Impact of an Increase in Income on Subjective Well-Being Multiple Regression Analysis. Structural-Change Test Variable Category Coefficient Significance Prob. > t Intercept C1 6.7075 Working Student .7270 Yes .0002 Situation Worker .9152 Yes .0002 Nature of Job Blue -.7901 Yes .0008 White -.4411 Indecision .0655 Health Excellent .6202 Yes .0002 Condition Good .5317 Yes .0000 Church Attendance Sometimes -.5516 Yes .0064 per Week Once -.3882 Indecision .0675 Concern for Never .7384 Yes .0001 Economic Situation Sometimes .3768 Yes .0005 Exercise days Three -.0440 No .6907 Per week One -.6420 Yes .0000 Civil Status Single/Divorced -.5480 Yes .0000 Age 31 to 40 -.7809 Yes .0000 More than 40 -.3309 Yes .0325 Income .0118 Yes .0300 Structural- Change Dummy .3561 Yes .0005 Goodness of Fit: R-square = 0.2216 31 Table XII Impact of Former Economic Situation and Expectation of Future Income on Subjective Well-Being Multiple Regression Analysis Variable Category Coefficient Significance Prob. > t Former Economic Higher -.4303 Yes .0218 Situation Equal -.1621 No .3395 Future Income Expectations Higher .3059 No .2118 Equal .2120 No .4213 R-square Former Economic Situation = 0.2792 R-square Expectation about Future Income = 0.2709 1 In this paper the terms subjective well being and happiness are used as synonyms. 2 Mullis (1990) run a regression between subjective well-being and economic variables and got an R 2 of around 0.11, while Douthitt et al (1992) got an R 2 of around 0.10. 3 Myers and Diener (1995) state that the following personality traits are linked to happy people: self-esteem, self-control, and optimism. Veenhoven (1994) discusses whether these personality traits are innate or learned.

4 Monterrey and Puebla represent two important urban zones in Mexico; therefore, results from this study can be generalized to apply only to urban areas in the country. A study of this nature is strongly recommended for rural regions in Mexico, given that these comprise the poorest people of the country.

5 Comparable questionnaires have been applied mostly in the United States and European countries (Andrews and Whitey, 1976; Fordyce, 1977; Campell et al., 1976; Gerdtham and Johannesson, 1997; Veenhoven, 1993.) The questionnaire used in this study is a new contribution for Mexico.

6 Economists are reluctant to compare well-being across individuals. Utility theory emphasizes the subjectivity of well-being and states that any comparison across individuals is senseless.

7 The tables presented in this paper directly use the Prob>t criterion instead of the t-test. The Prob>t criterion indicates the probability of rejecting the null hypothesis (no relationship between the independent and the dependent variable) under the case of the null hypothesis being true. A value for this criterion of less than 0.05 is considered to show the existence of a statistically significant relationship between the variables.

8 A reference individual with the lowest income has an original level of happiness of 6.4324 (satisfied), while a reference individual with the highest income has a level of happiness of 6.8121 (satisfied).

9 The dichotomous variable was constructed with people who answered being extremely happy, very happy, and happy in one category, and the rest in the other category.

10 Two possible explanations can be advanced: First, desires might be positively related to income, and people ponder the impact of higher income levels on their happiness without considering that their desires will also increase. However, when income really goes up, people find that their perception of material needs not satisfied does not decline (consumer theory assumes that desires remain constant while income increases and that happiness depends on needs satisfied rather than on perception of needs not satisfied). It could also happen that people are unable to figure out the higher responsibilities that generally accompany a higher income level (economic theory makes the same mistake when it separates the study of consumer behavior from the study of the income-leisure decision). View publication statsView publication stats