Expectations and Happiness: Evidence from Spain

Abstract

What is the role of people's expectations in explaining their happiness? Most research in this area focuses on the impact of the realization of expectations on happiness. With micro-level data from Spain, this paper studies the impact of current expectations on current happiness. The paper uses an instrumental-variables approach to control for potential endogeneity problems in order to find out what role expectations play in explaining current happiness. Results show that having positive expectations contributes to people's current satisfaction with life. It was also found that people's expectations are related to their personality and the trend in their economic situation. The paper contributes to the literature about expectations and happiness and it shows that recommending people to keep their expectations low may not necessarily be good advice in terms of their current happiness.

Headnote

Abstract

What is the role of people's expectations in explaining their happiness? Most research in this area focuses on the impact of the realization of expectations on happiness. With micro-level data from Spain, this paper studies the impact of current expectations on current happiness. The paper uses an instrumental-variables approach to control for potential endogeneity problems in order to find out what role expectations play in explaining current happiness. Results show that having positive expectations contributes to people's current satisfaction with life. It was also found that people's expectations are related to their personality and the trend in their economic situation. The paper contributes to the literature about expectations and happiness and it shows that recommending people to keep their expectations low may not necessarily be good advice in terms of their current happiness.

Keywords: Life Expectations; Life Satisfaction; Optimism; Happiness; Spain.

"Blessed is he who expects nothing, for he shall never be disappointed". Alexander Pope

"The secret to happiness is low expectations". Barry Schwartz

(ProQuest: ... denotes formulae omitted.)

Introduction

The importance of expectations on happiness is a relatively new area of research (Frijters, Liu & Meng, 2008; Senik 2008), while the study of anticipation has a relatively longer history (Loewenstein 1987; Berns et al., 2006). Some authors, such as Schwartz (2003), have stressed the important role that holding low expectations may have in generating future happiness; according to this argument, people who hold high expectations are more likely of facing future disappointment. Although Schwartz's argument seems reasonable, it is important to state that it neglects the role that expectations play in present happiness.

A relative new approach postulates that expectations have consumption value in the present. For example, in order to study optimal savings decisions, Brunnermeier & Parker (2004) assume that expectations of future consumption have present consumption value. Frijters, Liu & Meng (2008) also assume that expectations are consumption goods, when examining the effect of income expectations on life satisfaction among the Chinese population; they present evidence showing that expectations play an important role in explaining current life satisfaction in urban areas. Senik (2008) also examines the causal effect of expectations on happiness using a panel data from Russia; she finds a strong effect of expectations on life satisfaction.

It is possible to think about three potential channels in the relationship between happiness and expectations: First, current expectations people have about the future may influence their current happiness. Second, current happiness may influence current expectations about the future. Third, the realization of current expectations - an event which will take place in the future - affects future happiness. The first two channels deal with a bidirectional relationship taking place in the present between expectations and happiness, while the third channel links present expectations to future happiness. The third channel has received greater attention by the 'keepingexpectations-low' literature, which emphasizes disappointment as a source of unhappiness. This paper focuses on the first two channels, with a particular interest on the impact of current expectations on current happiness, while recognizing the existence of endogeneity; this is: current happiness may also influence current expectations. An instrumental-variables technique is used to address the endogeneity problem in order to find out what the impact of expectations on happiness is. People's past economic situation and their optimist/pessimist personality are used as instrument variables, on the basis of the available information. This research project cannot study the third channel due to the cross-section nature of the database; panel data would be required to fully grasp all channels taking place in the relationship between happiness and expectations.

The paper relies on a micro-level database from Spain, which contains information about expectations, happiness, and many other relevant variables. Due to the cross-sectional nature of the database this paper uses an instrumental-variables approach to deal with potential endogeneity problems between current expectations and current happiness.

Literature Review: Happiness and Expectations

Expectations: The present and the future

Expectations comprise both the present and the future; expectations are made in the present and they make reference to what people foresee in the future. People may have expectations about many things; for example, their income, their marriage, and their job. It is important not to confuse expectations with their realization. Expectations take place in the present, while their realization takes place in the future.

The ability of thinking about the future is in human condition (Gilbert, 2007). Human beings are able of gathering and selecting information, recalling events, analyzing patterns, identifying trends, attributing possibilities to the occurrence of potential events, pondering the impact of their actions as well as of other persons' actions, and evaluating interactions among agents in order to envisage what the future will bring; most of these factors are studied by what nowadays is called Prospect Theory (Kahneman, 2011; Kahneman & Tversky, 1979). People may have expectations about general events -for example, how joyful a trip will be-, as well as about specific events -for example, the value of the exchange rate by the end of the year-.

The ability of people to accurately foresee events is a matter of discussion in some disciplines. The rational-expectations hypothesis in economics postulates that people do not make systematic errors in predicting the future value of economic variables (Muth, 1961); many economic models have incorporated this hypothesis which leads to particular public-policy recommendations (Lucas, 1976; Sargent, 1987). On the contrary, prospect theory has emphasized the study of the biases and heuristics involved in predicting future events (Kahneman & Tversky, 1979); this theory postulates that systematic mistakes in foreseeing events are common (Loewenstein & Schkade, 1999; Camerer, Loewenstein & Rabin, 2003). There is some literature studying the formation and accuracy of expectations (Dominitz & Manski, 1997a and 1979b; Pesaran & Weale, 2006).

Expectations and aspirations

It is important to distinguish between expectations and aspirations; while aspirations make reference to a desired state in the future, expectations make reference to a predicted state in the future. Aspirations may motivate people to undertake some actions to attaining what they aspire for. Expectations are events people believe will occur in the future; expectations influence people's current behavior because people anticipate events and take decisions on the basis of what they expect will happen. Aspirations are not necessarily expected to take place, while expectations relate to predictions. Not reaching aspirations may be considered as a failure, while unrealized expectations may be considered as a predicting error. There is a relatively large literature on aspirations and well-being (Stutzer, 2004; Frey & Stutzer, 2002), but not so on expectations and well-being.

Expectations and personality

People may differ in their expectations, even when living within a similar context. It is reasonable to assume that people intend for their expectations to be correct; however, the formation of expectations constitutes a complex exercise in forecasting where many factors are involved, such as personality traits and life trajectories. People with different personality traits and with different life trajectories may differ in the set of information they focus their attention on, as well as in the way they process this information (Katona, 1975). In addition, people may follow different paths in learning from mistakes and in adjusting their expectations to the evidence of errors in prediction (Evans & Honkapohja, 2001; McCrae & Costa, 1986).

Several decades ago Keynes (1936) coined the termed 'animal spirits' to make reference to difficult-to-understand processes involved in the prediction of future events; in addition, it is well-known that some transient factors may play an important role in the formation of expectations (Loewenstein & Lerner, 2003)

Personality traits do play an important role in the formation of expectations. Some persons are more optimistic than others; as a matter of fact, optimists are identified on the basis of their positive expectations about the future. Optimists may work within the same circumstance than pessimists, but by focusing on some information and by processing it in a different way they come out with a more positive expectation (Brissette, Scheier & Carver, 2002).

In addition, some persons may be in a better position than others to take advantage of the contextual trends; thus, different expectations may not necessarily reflect different personality traits but also different positions people hold. In this case, expectations may reflect people's position and capacity to take advantage of forthcoming changes. Recent events may provide useful information for people to assessing what their position is; if things have gone well in the recent past then people may reasonably assume they are in a better position to take advantage of the forthcoming events and, in consequence, they may end up having more positive expectations.

Expectations and current behavior

The expectations people have influence their behavior. People adjust their behavior on the basis of what they expect the future will bring. The future will reveal whether expectations were correct or not; however, independently of whether expectations will turn out to be false or true in the future, they have an impact in the present by influencing people's behavior, strategies and planning. In this way, expectations have an impact both, in present and in future results (Camerer, Loewenstein & Rabin, 2003; Elster & Loewenstein, 1992). For example, negative expectations about the future of the country may incline people to postpone their marriage or to look for jobs in other regions of the world. Different expectations motivate different actions, leading to different results. Thus, both current and future happiness are sensitive to current expectations.

Expectations and current happiness

It can be hypothesized that people with high expectations tend to be happier. Many factors could play a role in this relationship, for example: better expectations may: reflect the anticipation of good conditions that will take place in the future (Loewenstein, 1987; Knight & Gunatilaka, 2014), induce people to have a positive focus which helps in the selection of information and in its using (Brown & Brown, 2011), and increase people's motivation to undertake those actions that lead to greater happiness (Bengtsson, Lau & Passingham, 2009). It is also important to recall that expectations themselves are accompanied by essential well-being experiences, such as anxiety, fear, and so on; in consequence, expectations are part of the informational substrate people use to assess their current well-being situation. Hence, expectations at time t may be associated to happiness at time t through many channels.

It is also important to recognize that personality traits do also play an important role in the association between happiness and expectations. For example, optimism is correlated to both happiness and expectations, as such, the relationship between expectations and happiness may emerge from the influence that a third variable - personality - has on them. Optimists do tend to hold better expectations, and an optimistic attitude - and positive thoughts - are important factors in people's happiness (Sharot, 2012; Oishi, Diener & Lucas, 2007; Frijters, Liu & Meng, 2008; Foster, Frijters & Johnston, 2012).

The realization of expectations and future happiness

Through their realization, current expectations may be linked to people's future well-being (Davis, 1981; Michalos, 1980; Schwartz, 2003; Thomas, 1981; Vermunt, Spaans & Zorge, 1989). The predicted event may turn out to be not as good as expected - or worse than expected -; this makes people upset and contributes to reduce their well-being. The event may also turn out to be better than expected - or not as bad as expected -, which makes people content and contributes to their well-being. It is important to note that the realization of expectations links current expectations to future happiness, and not to present happiness.

It is reasonable to think that people with higher expectations are more likely to face a negative realization gap; this is: they are more likely to be disappointment in the future and, as a consequence, more exposed to unhappiness in the future (Knight & Gunatilaka, 2010). This has been the reasoning used by Barry Schwartz to come out with his statement that "The secret to happiness is low expectations"; it is clear that this view emphasizes the relationship between the realization of expectations and happiness, but it completely neglects the relationship between expectations themselves and current happiness. In addition, this view also neglects the impact that current expectations may have on people's decisions and behavior, which may end up influencing the realization of expectations and, in consequence, the realization gap itself.

Some authors have argued that people may act strategically when the realization of expectations takes place, by adjusting their recalled expectations in order to minimize the predicting negative gap so that their happiness - in time t+n - is not affected by the emergence of a large negative gap (Kopalle & Lehmann, 2001; Schwartz, 2003; Ojasalo, 2001; Walker & Baker, 2000). Of course, it is always possible for some people to strategically adjust their current expectations - in time t - so as to increase the probability of positive gaps in the future (Helson, 1947); however, own-credibility issues arise in this strategy because it is not certain at all whether people can truly deceive themselves by holding apparent low expectations while thinking otherwise.

Main hypotheses

Due to limitations with the database, this paper does not deal with the impact of the realization of expectations on people's happiness. The purpose of this paper is to study the relationship between expectations - not their realization - and happiness in the present, while controlling for potential endogeneity problems. Two main hypotheses are tested:

First, expectations are associated to people's personality and to recent events in their lives. Second, current expectations do positively influence current happiness.

The Database

The survey

This research uses the Barómetro, which is a database built by the Center for Sociological Research in Spain. The Barómetro surveys have been conducted on a monthly basis since 1979 for the purpose of studying the opinions of the Spanish population regarding the political and socio-economic situation in the country. This research uses data from the survey conducted in December of each year from 1999 to 2004, which is the last year for which expectations data is available.

The Barómetro is a stratified-random sample, balanced by gender, age and urban-rural areas in Spain. Specific households in each area were randomly selected and an adult (18 years old and more) member was directly interview in each household. The unit of study is the person and not the family. Over 2,500 questionnaires are properly completed every year. The total sample for the period 1999 to 2004 contains 13,576 individuals.

Two main limitations are faced when using the Barómetro: first, it is a cross-section database. The authors are aware of the potential endogeneity problems that may show up when working with cross-section information; thus, an instrumental-variables technique -which is widely accepted to deal with this kind of potential problem - is used. Second, The Barómetro is a general opinion survey which was not specifically designed to address issues about expectations and happiness. In consequence, the authors are forced to work with somehow imperfect proxies. Nonetheless, the Barómetro provides relevant and exceptional information to understand the situation of expectations and happiness in Spain.

The main variables

a. Life satisfaction

The following question gathers information regarding people's satisfaction with life: "Could you please tell me on a scale from 1 to 5, where "1" means you are "very dissatisfied" and "5" means you are "very satisfied", how satisfied you are with the way your life has turned out so far?" Five response categories are presented: "very dissatisfied", "quite dissatisfied", "neither satisfied nor dissatisfied", "quite satisfied", and "very satisfied." In Spain, more than 11% of people were very satisfied and 69% were quite satisfied with their life during the 1999 to 2004 period; only 8% of Spanish people were quite and very unsatisfied with their life.

b. Life Expectations

The barometer survey gathers information regarding expectations about personal life on the basis of the following question: "Do you think that in the next 12 months you will live better than today, equal or worse?" 51.4% of Spanish people think that their lives will improve next year, whereas 8.3% of them think that their lives will worsen the next year.

c. Other relevant variables: personality, personal economic trajectory, and current situation

Personality proxy. There is no information on personality in the Barometer survey; thus, there is no direct information on the optimistic or pessimistic character of Spanish people. However, the survey has information on people's general perspective about the country's future from which it is possible to proxy people's optimism about the future. Optimism is measured using the information that comes from seven questions with the following general structure: "Do you think that next year the country's (unemployment, public insecurity, terrorism, drugs, the environment, health care, housing) situation will be better than today, equal or worse?" Each response is assigned a value of 1 for a better answer, and a value of 0 otherwise. These seven variables are added and the new value is then re-scaled into a 0 to 1 range; where 0 is associated to a pessimistic 3 personality and 1 to a highly optimistic personality.

As noted above, data limitations forced us to use somehow imperfect proxies; such is the case of the personality proxy. However, we believe that the personality proxy can capture the pessimistic and optimistic character of the people surveyed. Pessimists tend to provide responses with lower scores than optimists; and this is especially true when making assessments about the situation of the country in the future.

Personal economic trajectory. Past events may also influence people's expectations; in specific, life trajectories have been shown to play an important role in the formation of expectations. The Barometer survey asks people about whether their economic situation has improved or not with respect to the previous year. The question is "Do you think that your economic situation is better than one year ago, equal or worse?". Two dichotomous variables are constructed to handle this information: Economic trajectory-worse and economic trajectorybetter. On average, 25% of Spanish people think that their economic situation is better with respect to the previous year, while 13% think that their economic situation is worse.

Current situation. The survey did also ask people about their current situation, the following question is used: "How have things gone with you during the present year?"; the possible responses are "very badly", "badly", "so-so", "well" and "very well.". The categories "very badly" and "badly" are combined into one category due to their small number of observations. Four dichotomous variables are constructed to handle this information about people's appreciation of their current situation: Current situation very well, current situation well, current situation so-so, and current situation badly. About 3% of the people think that things have gone very badly in the present year, 10% think that things have gone badly in the current year, 26% think that things have gone so-so, and 61% that things have gone well or very well.

General overview

Table 1 shows the relationship between life expectations and life satisfaction. It is observed that the mean life satisfaction of Spanish people raises with life expectations; those who have better expectations for the future report, on average, a life satisfaction of 3.87 (associated to quite satisfied with their life), while those who have worse expectations report a life satisfaction of 3.22 (closer to the 'neither satisfied nor dissatisfied' category). Therefore, expectations appear to be positively correlated to satisfaction with life.

Table 1 also reports the relationship between the current-situation variable ("how have things gone this year?") and life satisfaction. Life satisfaction increases considerably if things have gone well compared to things going badly. If things have gone very well this year, the person will have, on average, a high level of life satisfaction (4.44), while if things have gone wrong this year, the average satisfaction level will be 2.64.

Table 2 shows how personal-life expectations relate to the personality and economic trajectory variables. It provides a hint about the positive relationship between life expectations and optimism and upward economic trajectory. More optimism, less pessimism, and upward economic trajectory are associated to better life expectations.

Control variables: socio-economic and demographic characteristics

As it is common in the explanation of life satisfaction, socio-economic and demographic variables need to be introduced into the model in order to reduce any misspecification problem (Argyle, 1999; Ahn & Mochón, 2010; Van Praag & Ferrer-i-Carbonel 2004). The database contains the following variables to control for:

Gender, coded 1 if the person is male and 0 otherwise. 52% of persons in the survey are male.

Age measures the person's age in years. An age-squared variable (Age2) is also generated in order to incorporate nonlinearity in the relation between life satisfaction and age. The average age in the sample is 42 years.

Marital status is measured by four dummies indicating whether the person has never married (Single), is married (Married), divorced (Divorced), or widowed (Widowed). 55% of people in the survey are single, 37% are married, 6% are divorced, and 2% are widowed.

Occupational status. Five dichotomous variables are constructed to study the occupational status: Unemployed, Retired, Studying, Housekeeper, and Working. 10% of people in the sample are unemployed, 15% are retired, 7% are students, and 13% are housekeepers.

Educational level is measured by four dummies: Primary Incomplete, Primary, Secondary, and University. 4% of people in the sample have not completed primary education, 45% have completed primary school, 31% have a secondary level of education, and 20% have a university degree.

Economic status is measured by household net income. The information comes by income ranges, and five income categories are constructed. The vector of dichotomous Income variables provides information about the economic situation in the household. The income intervals range from "less or equal than 300 euros" to "over 6,000 euros". 4 In the sample, 31% of the households have a low income, 24% have an income that is about average, 6% have a high level of income and 1% has the highest level of income.

The attitude towards religion has three response categories: believer, practicing, and nonbeliever. Two dichotomous variables are generated: Believer and Practicing; 80% of people in the sample are believers and 16% are practicing.

Year dummies are also introduced to capture the effect of cyclical macroeconomic fluctuations on satisfaction with life.

Implementing an Instrumental-Variables Approach

The relationship between life expectations and life satisfaction is complex; life expectations may have an influence in current life satisfaction, but it may also happen for life satisfaction to play a role in the formation of life expectations. In other words, it could be that by having better life expectations people tend to be more satisfied with their current life; however, it could also be that those who are more satisfied with their life tend to foresee a better future. Furthermore, it Mcould also be for a third variable, such a personality trait, to influence both current life satisfaction and life expectations.

In order to overcome these limitations associated to the potential existence of endogeneity, as well as of spurious correlations, this paper implements an instrumental-variables technique; so that a group of (instrumental) variables is used to explain life expectations, and this group of variables do not affect other variables that are determinants of life satisfaction and which could be captured in the error term. A two stage estimation process is followed: In the first stage, the life expectation variable is estimated on the basis of a set of explanatory variables used to estimate life satisfaction, as well as on a set of instrumental variables. Two instrumental variables are used to explain life expectations: First, personality; since it is likely for life expectations to depend on people's personalities. Optimists tend to have better expectations for the future than pessimists (Scheier & Carver, 1985 and 1992; Daukantaitè & Zukauskiene, 2012). Second, personal economic trajectory; it is possible for people to form their life expectations on the basis of recent past. In other words, it is reasonable to think that people's expectations are conditioned by what happened in their recent past. Table 2 already provided hints on this regard. The Bound, Jaeger & Baker F statistics test (Bound, Jaeger & Baker, 1995) and the Sheás partial R2 statistics test (Shea, 1997) for accepting personality and personal economic trajectory as instrumental variables were implemented, and it was found that both variables passed the tests.

Life expectation is handled as an ordinal variable with three response categories: better, equal, and worse. Hence, life expectation of person i (LEi) is modeled as a manifestation of the latent and continuous variable ( *); which is estimated as a function of personality trait, personal economic traj*ectory, and a set of control variables:

... (1)

where is the latent and continuous life expectation, Optimisti is a variable with value between 0 and 1 showing people with different levels of optimist personality, and the vector Economic Trajectory includes the variables that measure whether the person's economic situation has improved or worsened with respect to the previous year. xi represents the vector of control variables (socio-economic, demographic variables and year dummies). ^ represents the thresholds increasing in k. γ, and β are parameters to be estimated, and is a normal distributed error-term.5 This assumption makes the model a standard ordered-probit model. An OLS model of life expectation is also estimated; the results using both estimation methodologies are largely consistent. 6

With the estimated parameters β and γ it is possible to calculate the predicted life expectation of person i ( ? ). In the second-stage regression, the predicted life expectation ( ? ) is introduced as an explanatory variable in the life-satisfaction equation. This allows for examining how life expectation affects life satisfaction. Life satisfaction (LSi) is handled as an ordinal variable; hence a latent and continuous life satisfaction varia ble ( *) is assumed, and the following ordered- probit specification is used to explain life satisfaction:

... (2)

represents the predicted life expectation variable, ^ is the thresholds increasing in k, θ and δ are the parameters to be estimated, and ^ is a normal distributed error-term. A chi2 test for checking endogeneity for the role of life expectation is applied; this test confirms the presence of endogeneity in the relationship between life expectation and life satisfaction.

Expectations and Life Satisfaction: Evidence from Spain

Table 3 presents the estimated results for the life expectation equation (equation 1). This econometric exercise addresses our hypothesis about how expectations are associated to people's personality and to recent events in their lives. For the sake of simplicity, the estimated parameters corresponding to demographic and socio-economic variables are omitted in this table. It is observed in Table 3 that the coefficients of the variables corresponding to the Optimist personality and Economic trajectory are significant and show plausible signs.

The Bound, Jaeger & Baker F statistics and the Shea's partial R2 statistics are used to test the relevance of the instrumental variables in the first-stage regression. The value of these statistics show that the instruments used to measure life expectation are of relevance.

The results show that there is a positive relationship between being optimistic and life expectations. As it has been remarked by many authors the optimists tend to have better expectations for the future than the pessimists (Scheier & Caver, 1985; Daukantaitè & Zukauskiene, 2012). The results also show the importance of past events in the formation of expectations. Those who have improved economically with respect to the previous year tend to have better expectations. On the contrary, those whose economic situation has worsened compared to the previous year tend to have worse expectations.

Predictions about life expectations are obtained by replacing the parameters with the estimated coefficients in equation (1). Based on this prediction it is possible to construct two variables: LE_Better represents the prediction of people who think that in the next year they will live better than today. LE_Worse represents the prediction of people who think that in the next year they will live worse than today. These variables are introduced as regressors in the life-satisfaction equation. Equation (2) is estimated by using an ordered-probit method in order to address our other hypothesis: current expectations do positively influence current happiness.

Table 4 reports the results from two regressions: The ordered-probit method using the lifeexpectation variables as exogenous and the ordered-probit method with instrumental variables. It is observed that the coefficients of the expectations variables vary considerably when an instrumental-variables technique is used. Furthermore, a chi2 test9 for checking the endogeneity of the two life-expectation variables is run; it is found that the null hypothesis of exogeneity is rejected; thus, it is possible to conclude that some endogeneity in the relationship between life satisfaction and life expectations does exist and that, in consequence, the instrumental-variables technique allows solving the endogeneity problem in order to better estimate the impact of life expectations on life satisfaction.

The impact of general expectations on life satisfaction supports the previously formulated hypotheses (hypothesis 2); having better life expectations positively contributes to current life satisfaction, while having worse life expectations tends to reduce current life satisfaction. Both estimated coefficients are statistically significant at 1%. Having better life expectations - rather than equal life expectations - increases the probability of being very satisfied in 0.02; while having worse life expectations - rather than equal - tends to reduce the probability of being very satisfied in 0.05. Hence, moving from worse to better life expectations represents an increase in the probability of being very satisfied with life of about 0.07, which is a relatively large and statistically significant difference.

The coefficient of better expectations (0.16) is also significantly lower (in absolute terms) than the coefficient of worse expectations (0.66). This is an interesting result, showing that life satisfaction is more sensitive to negative life expectations than to positive life expectations.

It is also observed in Table 4 that the current situation also plays an important role in explaining people's satisfaction with life.

The empirical research shows that expectations have an impact on people's satisfaction with life; in other words, expectations have intrinsic value in the present, independently of their realization in the future. This result confirms our second hypothesis.

Conclusion

This paper analysed the impact of people's life expectations on their life satisfaction in the Spanish society. The paper recognized the existence of endogeneity problems in the relationship between life expectations and life satisfaction, and an instrumental-variables technique was used to estimate the net impact of expectations on life satisfaction.

It was found that people's expectations about their future are related to their personality (as measured by some optimist and pessimist-personality proxies) as well as to the evolution of their economic situation in the recent past. Optimists tend to have better life expectations, as well as those who have experienced a positive trend in their economic situation. These results are consistent with the general literature on expectations.

Once endogeneity problems are taken into account by using an instrumental-variables technique it is found that having high life expectations positively contributes to people's current satisfaction with life, while having low life expectations tends to reduce people's current satisfaction with life. These findings support Frijters, Liu & Meng's (2008) approach on considering expectations as part of the goods people may have.

Further research is needed to fully understand the channels that link current expectations to current happiness: holding positive expectations may induce behaviors that favor happiness, for example: people may be more inclined of venturing into new activities; they may be more inclined on pursuing some important goals and aspirations; and they may be more open to socializing and to start new relationships. Positive expectations may also change people's patterns of focalization and informational processing; for example: people may focus on future projects rather than on past ones; people may also focus on opportunities rather than on threats. Unfortunately, the information available is not sufficient to address these fundamental issues; a tailor-made survey would be required to fully study these issues, and the results presented in this paper may contribute to justify this kind of study. Meanwhile, the results from this study show that, perhaps, Alexander Pope was wrong when stating some centuries ago that having no expectations was a blessing; in fact, this investigation found that having high or positive expectations may be a blessing. Furthermore, perhaps it is also wrong to state that the secret to happiness is holding low expectations. It may be true that having low expectations may reduce the risk of facing disappointment in the future; however, this paper shows that this strategy may lead to a reduction of happiness in the present.

Acknowlegments

We wish to thank, for their useful comments and suggestions, to the participants at the XI ISQOLS Conference. This research has been funded by the project CICYT CSO2012-39677.

Footnote

2 The strata were formed by the junction of the 17 autonomous communities with habitat size, divided into 7 categories: less than or equal to 2000 inhabitants, from 2001 to 10000, from 10001 to 50000, from 50001 to 100000, from 100001 to 400000, from 400001 to 1000000, and over 1000000 inhabitants. The sampling error for a confidence level of 95.5% (two sigma), and P = Q, is ± 2'0% for the whole sample.

3 It is clear that expectations about some country-level issues obey to a different explanatory structure than personal-life expectations. A person can have great expectations about their personal life while being pessimistic about the future of the country; as a matter of fact this is a common pattern. Furthermore, while many of the factors justifying personal-life expectations are closer to people's own decisions, the expectations about many country-level issues depend on factors which are beyond the control of most people.

4 The income ranges are the following: "Very low" ranges from less or equal than/to 300 euros to 600 euros; "Low" ranges from 601 to 1200 euros; "About average" ranges from 1201 to 2400 euros; "High" ranges from 2401 to 4500 euros; "Very high" ranges from 4501 to over 6000 euros.

5 The notation of this paper is based on the one used by Frijters, Liu & Meng (2008).

6 Results from an OLS model are largely consistent with the ordered-probit model. This is a usual finding in the literature on limited dependent variables, notably for happiness (see Ferrer-i-Carbonell & Frijters, 2004: van Praag Carbonell, 2004).

7 For a detailed description of the model, see among others Chapter 15 of Econometric Analysis of Cross Section and Panel Data by Wooldridge (2002).

8 Two dichotomous variables are constructed on the basis of the estimated value for life expectations. If this estimated value is lower than 1.91 then LE_Worse is equal to 1, otherwise it is equal to 0. If the estimated value is greater than 2.45 then LE_Better is equal to 1, otherwise it is equal to 0. It is clear that the case of LE_Worse = 0 and LE_Better = 0 corresponds to equal life expectations.

9 The chi2 tests, which allow us to check the endogeneity of the two expectations variables, indicate that the null hypothesis is rejected so these variables are endogenous. These results confirm that the econometric method used to estimate the happiness model is correct.

References

References

1. Ahn, N., & Mochon, F. (2010). La Felicidad de los Españoles: Factores Explicativos. Revista de Economía Aplicada, 54(XVIII), pp. 5-31.

2. Argyle, M. (1999). Causes and Correlates of Happiness. In D. Kahneman, E. Diener & N. Schwarz (Eds.), Well-Being: The Foundations of Hedonic Psychology (pp. 353-373), New York: Russell Sage Foundation.

3. Bengtsson S.L., Lau, H., & Passingham, R.E. (2009). Motivation to do Well Enhances Responses to Errors and Self-monitoring. Cerebral Cortex, 19(4), 797 804.

4. Berns, G.S., Chappelow, J., Cekic, M., Pagnoni, G., & Martin-Skurki, M.E. (2006). Neurobiological Substrates of Dread, Science 5, 312(5774), pp. 754-758.

5. Bound, J., Jaeger, D.A., & Baker, R.M. (1995). Problems with Instrumental Variables Estimation when the Correlation between the Instruments and the Endogeneous Explanatory Variable is Weak, Journal of the American Statistical Association, 90 (430), pp. 443-450. http://dx.doi.org/10.2307/2291055

6. Brissette, I., Scheier, M., & Carver, C. (2002). The Role of Optimism in Social Network Development, Coping, and Psychological Adjustment during a Life Transition, Journal of Personality and Social Psychology, 82(1), pp. 102-111. http://dx.doi.org/10.1037/00223514.82.1.102

7. Brown, J.D., & Brown, M.A. (2011). Self-Reflection and Feelings of Self-Worth: When Rosenberg meets Heisenberg, Journal of Experimental Social Psychology, 47(6), pp. 1269-1275. http://dx.doi.org/10.1016/j.jesp.2011.05.019

8. Brunnermeier, M., & Parker, J. (2004). Optimal Expectations, National Bureau of Economic Research Working Paper 10707.

9. Camerer, C.F., Loewenstein, G., & Rabin, M. (Eds.) (2003). Advances in Behavioral Economics, USA: Princeton University Press.

10. Centro de Investigaciones Sociológicas (1999-2005). Barometro Data Files

11. Daukantaite, D., & Zukauskiene, R. (2012). Optimism and Subjective Well-Being: Affectivity Plays a Secondary Role in the Relationship Between Optimism and Global Life Satisfaction in the Middle-Aged Women. Longitudinal and Cross-Cultural Findings. Journal of Happiness Studies, 13(1), pp. 1-16. http://dx.doi.org/10.1007/s10902-0109246-2

12. Davis, W. (1981). A Theory of Happiness, American Philosophical Quarterly, 18(2), pp. 111-120.

13. Dominitz, J., & Manski, C. (1997a). Perceptions of Economic Insecurity: Evidence from the Survey of Economic Expectations, Public Opinion Quarterly, 61(2), pp. 261287. http://dx.doi.org/10.1086/297795

14. Dominitz, J., & Manski, C. (1997b). Using Expectations Data to Study Subjective Income Expectations, Journal of the American Statistical Association, 92(439), pp. 855867. http://dx.doi.org/10.1080/01621459.1997.10474041

15. Elster, J., & Loewenstein, G. (1992). Utility from Memory and Anticipation. In G. Loewenstein & J. Elster (Eds), Choice over Time (pp. 213-234), New York: Russell Sage Foundation.

16. Evans, G., & Honkapohja, S. (2001). Learning and Expectations in Macroeconomics, Princeton, New Jersey: Princeton University Press.

17. Ferrer-i-Carbonel, A., & Frijters, P. (2004). The Effect of Methodology on the Determinants of Happiness, Economic Journal, 114, pp. 641-659.

18. Foster, G., Frijters, P., & Johnston, D. (2012). The Triumph of Hope over Disappointment: A Note on the Utility Value of Good Health Expectations, Journal of Economic Psychology, 33(1), pp. 206- 214. http://dx.doi.org/10.1016/j.joep.2011.09.010

19. Frey, B., & Stutzer , A. (2002). Happiness and Economics, Princeton: Princeton University Press.

20. Frijters, P., Liu, A.Y.C., & Meng, X. (2008). Are Optimistic Expectations keeping the Chinese Happy? NCER Working paper Series, 37, November.

21. Gilbert, D. (2007). Stumbling on Happiness, US: Vintage.

22. Helson, H. (1947). Adaptation-level as a Frame of Reference for Prediction of Psychophysical Data, American Journal of Psychology, 60(1), pp. 1-29. http://dx.doi.org/10.2307/1417326

23. Kahneman, D. (2011). Thinking, Fast and Slow, New York: Farrar, Straus and Giroux.

24. Kahneman, D., & Tversky, A. (1979). Prospect Theory: An Analysis of Decision under Risk, Econometrica. 47(2), pp. 263-291.http://dx.doi.org/10.2307/1914185

25. Katona, G. (1975). Psychological Economics, New York: Elsevier.

26. Keynes, J.M. (1936). The General Theory of Employment, Interest and Money, London: Macmillan.

27. Knight, J., & Gunatilaka, R. (2010). Great Expectations? The Subjective Well-Being of Rural-Urban Migrants in China, World Development, 38(1), pp. 113-124. http://dx.doi.org/10.1016/j.worlddev.2009.03.002

28. Knight, J., & Gunatilaka, R. (2014). Memory and Anticipation: New Empirical Support for an Old Theory of the Utility Function, Department of Economics, University of Oxford, Discussion paper No. 721.

29. Kopalle, P.K., & Lehmann, D.R. (2001). Strategic Management of Expectations: The Role of Disconfirmation Sensitivity and Perfectionism, Journal of Marketing Research, 38(3), pp. 386-394. http://dx.doi.org/10.1509/jmkr.38.3.386.18862

30. Loewestein, G. (1987). Anticipation and the Valuation of Delayed Consumption, The Economic Journal, 97(387), pp. 666-684.http://dx.doi.org/10.2307/2232929

31. Loewenstein, G., & Schkade, D. (1999). Wouldn't it be Nice? Predicting Future Feelings. In D. Kahneman, E. Diener & N. Schwarz (Eds.) Foundations of Hedonic Psychology: Scientific Perspectives on Enjoyment and Suffering (pp. 85-105), New York: Russell Sage Foundation.

32. Loewenstein, G. & Lerner, J. (2003). The Role of Affect in Decision Making. In R.J. Davidson, K.R. Scherer & H.H. Goldsmith, Handbook of Affective Sciences (pp. 619642), New York: Oxford University Press.

33. Lucas, R.E. (1976). Econometric Policy Evaluation: A Critique, Carnegie-Rochester Conference Series on Public Policy, 1, pp. 19-46. http://dx.doi.org/10.1016/S01672231(76)80003-6

34. McCrae, R.R., & Costa, P.T. (1986). Personality, Coping, and Coping Effectiveness in an Adult Sample, Journal of Personality, 54(2), pp. 385-405. http://dx.doi.org/10.1111/j.1467-6494.1986.tb00401.x

35. Michalos, A. C. (1980). Satisfaction and Happiness, Social Indicators Research, 8(4), pp. 385-422. http://dx.doi.org/10.1007/BF00461152

36. Muth, J.F. (1961). Rational Expectations and the Theory of Price Movements, reprinted in The New Classical Macroeconomics, Volume 1, (1992): pp. 3-23 (International Library of Critical Writings in Economics, vol. 19. Aldershot, UK: Elgar).

37. Ojasalo, J. (2001). Managing Customer Expectations in Professional Services, Managing Service Quality, 11(3), pp. 200-213. http://dx.doi.org/10.1108/09604520110391379

38. Oishi. S., Diener, E., & Lucas, R.E. (2007). The Optimum Level of Well-Being. Can People be Too Happy?, Perspectives on Psychology Science, 2(4), pp. 346-360. http://dx.doi.org/10.1111/j.1745-6916.2007.00048.x

39. Pesaran, M.H., & Weale, M. (2006). Survey Expectations. In G. Elliott, C.W.J. Granger, & A. Timmermann (Eds.), Handbook of Economic Forecasting (pp. 715-776), Amsterdam: North-Holland.

40. Sargent, T.J. (1987). Rational expectations, The New Palgrave: A Dictionary of Economics, 4, pp. 76-79.

41. Senik, C. (2008). Is Man Doomed to Progress? Journal of Economic Behavior and Organization, 68(1), pp. 140-152.http://dx.doi.org/10.1016/j.jebo.2008.03.004

42. Scheier, M., & Carver, C. (1985). Optimism, Coping and Health: Assessment and Implications of Generalized Outcome Expectancies, Health Psychology, 4(3), pp. 219247. http://dx.doi.org/10.1037/0278-6133.4.3.219

43. Scheier, M., & Carver, C. (1992). Effects of Optimism on Psychological and Physical Well-Being: Theoretical Overview and Empirical Update, Cognitive Therapy and Research, 16(2), pp. 201 - 228.http://dx.doi.org/10.1007/BF01173489

44. Schwartz, B. (2003). The Paradox of Choice: Why More is Less, New York: Ecco.

45. Sharot, T. (2012). The Optimism Bias: A Tour of the Irrationally Positive Brain, US: Vintage Books.

46. Shea, J. (1997). Instrument Relevance in Multivariate Linear Models: A Simple Measure, Review of Economics and Statistics, 79(2), pp. 348-352. http://dx.doi.org/10.1162/rest.1997.79.2.348

47. Stutzer, A. (2004). The Role of Income Aspirations in Individual Happiness, Journal of Economic Behavior and Organization, 54(1), pp. 89-109. http://dx.doi.org/10.1016/j.jebo.2003.04.003

48. Thomas, W.C. (1981). The Expectation Gap and the Stereotype of the Stereotype: Images of Old People, The Gerontologist, 21(4), pp. 402-407. http://dx.doi.org/10.1093/geront/21.4.402

49. Van Praag, B., & Ferrer-i-Carbonell, A. (2004). Happiness Quantified: A Satisfaction Calculus Approach, Oxford: Oxford University Press.

50. Vermunt, R., Spaans, E. & Zorge, F. (1989). Satisfaction, Happiness and Well-Being of Dutch Students, Social Indicators Research, 21(1), pp. 1-33. http://dx.doi.org/10.1007/BF00302402

51. Walker, J., & Baker, J. (2000). An Exploratory Study of a Multi-Expectation Framework for Services, Journal of Services Marketing, 14(5), pp. 411-431. http://dx.doi.org/10.1108/08876040010340946

52. Wooldridge, J.M. (2002). Econometric Analysis of Cross Section and Panel Data, New York: The MIT Press.