Abstract

One might expect economic migrants to experience an increase in happiness after migration: life in wealthier countries might be better, particularly for migrants who succeed in improving their financial situation. From the perspective of ‘happiness studies', however, migration motivated by the prospect of economic gain is perhaps a misguided endeavor. In general, people do not gain happiness from an increase in their incomes, and migration as a means of gaining an increased income might not amount to an exception to that general pattern. This article explores happiness among migrants and stayers in a number of European countries, investigating individuals from eastern European countries who went to western Europe. Migrants generally appear to be happier than those who have remained in the countries of origin—but there is evidence that this difference is the result of a greater tendency towards migration among people with higher levels of happiness (thus not a matter of happiness increasing as a consequence of migration). In addition, there is significant variation by country: migrants from Russia, Turkey, and Romania are happier than stayers, but migrants from Poland are significantly less happy than stayers. Models that determine whether a correction for endogeneity is necessary suggest that those country-level differences represent increases and decreases (respectively) in happiness.

1. Introduction

Do migrants from poorer countries who move to wealthier countries end up happier for having migrated? Insofar as migration is voluntary, one might assume that they do; at a minimum, it seems obvious that migrants expect to be better off after migration. The question addressed here is whether that expectation is borne out in migrants’ experiences, with ‘better off’ specified as happiness.

Several recent articles on the topic have raised reasonable doubts that migration to a wealthier country brings greater happiness. At the very least, there are good reasons not to assume that it does; instead the point should be addressed empirically (Safi 2010; Bartram 2011). This research is rooted in ‘happiness studies’ (e.g. Layard 2005; Diener et al. 2009; Thin 2012)—and a major finding of this field is that, as long as one is above a certain threshold, gaining an increase in income does not generally lead to greater happiness. That finding can be read to imply that migration as a specific means of gaining an increased income might not lead the migrants to become happier.1 Migrants (i.e. ‘economic migrants’) might well believe that gaining a higher income will make them happier, but that belief might simply be misguided; having mistaken beliefs about paths to happiness is not at all uncommon (Gilbert 2006; Bartolini 2007).

Empirical research on the question is hampered by absence of panel data on migrants; there are no suitable panel datasets that capture migrants before as well as after (international) migration. Existing research, then, is limited to cross-sectional comparisons, which are vulnerable to selection bias and which in any event cannot tell us directly about changes in migrants’ happiness. Researchers have mainly compared migrants’ happiness to that of natives in the destination country. The fact that migrants are generally less happy than natives is then sometimes read as suggesting that migrants did not become happier after migration.

That form of comparison, while interesting for certain purposes, is quite limited in what it can tell us about the consequences of migration for the migrants. The key problem is that it fails to account for differences in happiness levels between the origin and destination countries. If a destination country is significantly happier on average than an origin country, one might well find that migrants experience an increase in happiness that nonetheless leaves them at a level below the average for the destination country. (One might imagine that if destination countries have higher happiness levels, then migrants are indeed likely to gain greater happiness by going to such countries—but that assumption amounts to an obvious ecological fallacy.)

A better approach to analysis of cross-sectional data for this purpose is a comparison of migrants in wealthier countries with ‘stayers’ in the countries the migrants left (cf. Bayram et al. 2007). With that form of comparison, it becomes possible to use techniques that correct for endogeneity: if (for example) migrants are happier than stayers, we can try to determine whether that difference is a matter of happiness increasing with migration or results instead from a greater tendency towards migration among happier people.

This article explores these issues via analysis of data from the European Social Survey (ESS). The analysis compares (1) immigrants originating in Eastern European countries who are currently living in Western European countries with (2) stayers in the former. The premise of focusing on migration flows in that frame is that migrants were likely motivated by the prospect of higher incomes available in the western countries (‘economic migration’).2 The main finding reported below is that migrants are happier than stayers in the countries from which the migrants emigrated when individual characteristics are controlled—but this difference disappears in a model that corrects for the possibility that happiness causes migration rather than vice versa. Beyond that broad finding, the analysis also finds notable variation in experiences relating to the different origin countries: emigrants from some countries are indeed happier than stayers in those particular countries, but migrants from Poland are significantly less happy than stayers. Variation in migrants’ experiences is also apparent when investigating their happiness by country of destination.

2. Previous research on happiness and migration

The burgeoning field of happiness studies starts with a reasonably straightforward premise: to understand well-being, it is not sufficient to investigate only its ‘objective’ forms (income, health, ‘social capital’, etc.), nor can one simply assume that ‘subjective well-being’ (i.e. happiness) follows directly from having high levels of objective forms of well-being.3 Instead, one must study happiness directly, for example by asking people how happy they are. Following this approach, it becomes quite clear that the assumption is often false: some people with high levels of objective well-being are not particularly happy, and some common beliefs about what brings happiness are not well supported in people’s experiences.

A now-familiar finding along those lines concerns income. Although people with higher incomes are typically happier than people with less, it is not apparent that gaining an increase in income brings greater happiness (the ‘Easterlin paradox’, Easterlin 1973, 2001). The happiness consequences of income follow to a great extent from the way income signals status: people with more can derive satisfaction from comparing themselves to people with less. An increase in income, then, might fail to bring greater happiness for several reasons. Although a favourable downward comparison is possible for people who gain higher incomes, it is not inevitable: a high-earning individual might prefer to compare to people who earn even more, and that preference might persist even following an increase in income (Clark et al. 2008; Boyce et al. 2010). The point can be expressed also in terms of aspirations: one who keenly wants a higher income might, after gaining an increase, aspire to further increases (Easterlin 2001; Stutzer 2003); the aspiration might be more a personality trait than a preference that can be satisfied at a particular level. At a minimum, the status-signalling function of income appears to mean that economic growth (at least for relatively wealthy countries) cannot lead to increases in average levels of happiness (Easterlin et al. 2010). That point persists even when individual mobility is considered (if some move up, others must move down). Income does matter for happiness, but not in the way many people believe it does, and it is reasonably considered a mistake (at least in connection with happiness) to pursue greater income for its own sake (Kasser 2003; Kahneman et al. 2006).

These ideas have wide appeal, no doubt in part because they accord with a certain folk wisdom about not being able to buy happiness. But they are also disputed. Stevenson and Wolfers (2008) assert that one can find evidence of a positive ‘longitudinal’ association even in wealthy countries (though see Easterlin and Angelescu, 2009, for their rebuttal). Fischer (2008) argues for looking beyond per capita (average) GNP figures, noting that mean income in the USA has risen but most of the gains have gone to the wealthy; the stagnating wages experienced by many are quite consistent with the observation of flat happiness trends there. In addition, there has arguably been excessive attention to income and insufficient research on other elements of people’s economic situations, particularly wealth (Headey et al. 2008; Christoph 2010). Veenhoven (1991, 2008) provides a more comprehensive contrasting view here: he argues that people care less about comparisons than is commonly believed (by many happiness researchers, at any rate) and care more about the ‘absolute’ quality of their situations. That point is especially important for international comparisons: wealthy countries are, Veenhoven writes, more ‘liveable’ than poor countries, and so perhaps economic growth (as well as income increases for individuals) can make countries of both types more liveable.

2.1 Implications for migration

The discussion above of income and happiness leads to contrasting predictions about the happiness consequences of economic migration, i.e. migration motivated by desire for increased income or other economic benefits (as against family reunification or lifestyle migration). Veenhoven’s ‘liveability theory’ grounds an expectation that migrants to wealthy countries are likely to be happier there, even if they were not motivated by desire for economic gain. In many respects modern wealthy societies enable individuals to meet their needs to a greater extent than in poorer societies. While many sociologists embrace a critique of modern societies that emphasizes alienation, anomie, and the loss of close-knit communities, Maryanksi and Turner’s (1992) perspective on the ‘social cage’ is more optimistic about even the social aspects of modern societies—and on this basis as well Veenhoven (2008) argues that modern societies are more conducive to happiness. This perspective has the advantage of not implying that migrants are mistaken in their beliefs about the benefits of the very significant choice they have made.

On the other hand, sometimes people do give ‘excessive’ emphasis to desire for economic gain, and economic migration might be a specific instance of this type of choice (perhaps akin to the decision to become an investment banker). While average happiness is higher in wealthier countries, this does not mean that happiness will increase for anyone who can move there. To the extent that comparisons weigh heavily in people’s happiness (as in the Easterlin perspective), then, migrants to wealthy countries might face some unanticipated difficulties. Migrants who held middle-class positions in the origin country might find it difficult to gain middle-class occupations and incomes in the destination country, given unrecognized qualifications, prejudiced employers, insufficient language abilities, etc. (e.g. Portes and Bach 1985). In such cases one might perceive a downward trajectory of status, perhaps with negative consequences for happiness. For migrants who originate in a lower origin-country position, obstacles to upward mobility in a wealthy destination country are likely to be substantial. Even if status does not decline with migration, aspirations might be intensified (and then frustrated) via direct exposure to the consumption standards of wealthy societies.

As noted above, research on these propositions is hampered by the lack of panel data that would enable the appropriate forms of analysis (at a minimum, observing how migrants’ happiness changes over time, beginning with the pre-migration period). A number of studies therefore consider inferences drawn from cross-sectional analyses. A common finding is that migrants are less happy than natives in destination countries (Bălţătescu 2005, 2007), and that that gap does not diminish with time as per a prediction regarding immigrant integration (Safi 2010). Immigrants in the USA—even those originating in poorer countries—are less happy than natives perhaps because they are less satisfied with their financial situations despite having earnings roughly on par with those of natives (Bartram 2011); in other words, their incomes in an ‘absolute’ sense have manifestly increased relative to pre-migration levels but they are nonetheless even more dissatisfied with those incomes than natives at comparable levels. In addition, research that attempts to explain why some immigrants are unhappy points to difficulties of integration, particularly unemployment and lower-status employment (Aycan and Berry 1996) and social factors such as discrimination and isolation (Ying 1996). In more general terms, immigrants sometimes experience discrepancies between expectations of migration and realities after migration (Vohra and Adair 2001; see Michalos 1985 on discrepancies in general).4 Though not an instance of ‘happiness studies’, Dreby’s (2010) work demonstrates the enormous sacrifices some migrants make (extended separation from close family, in particular), and if the ostensible benefits of increased income prove chimerical, then the sacrifices might amount to pure loss (cf. Matt 2007, on migrant homesickness).

There are similar findings in research on internal migration, at least in certain poorer countries. Knight and Gunatilaka (2010) explore the lower happiness of rural-to-urban migrants in China (relative to rural stayers and those of urban origin) and find evidence indicating that a plausible explanation is that migrants’ aspirations (e.g. for income) rise after migration, in a way they failed to anticipate prior to their decision to migrate. De Jong et al. (2002) found that a significant proportion of migrants within Thailand reported lower satisfaction on key dimensions (employment, living environment, and community facilities) after migration. While these studies use cross-sectional data, internal migration in some wealthier countries can be analysed via panel data. Nowok et al. (2011), for example, find that migrants within Britain typically experience decreased happiness in a period immediately prior to migration; happiness then rises towards an earlier pre-migration level and subsequently ‘stagnates’.

Research by Melzer (2011) uses German panel data to track the experiences of migrants from eastern to western Germany in the period following German reunification. While this is internal migration in a strictly legal sense, it is arguably akin to international migration insofar as the two regions were separate countries with important differences that persisted even after reunification in 1990; the higher development level of the west also supports the assumption that migration to the west was commonly motivated by economic aspiration. Melzer finds that migrants to western Germany generally experienced an increase in happiness; an important reason, particularly for men, is the improved employment opportunities in the west. Though an important result for its methodological merits, it is doubtful that this particular migration flow gives us a good basis for knowing the consequences of international migration more generally: eastern German migrants to the west might have experienced their migration as akin to that between two different countries in certain respects, but they also shared a common language and many elements of a common culture, in contrast to the experiences of many international migrants.

In short, there is mixed evidence regarding the likely happiness consequences of migration. A noteworthy characteristic of research on migration and happiness to date is that researchers tend (at least implicitly) to approach the question in its most general form, i.e. does migration generally lead to an increase/decrease in happiness? It is unlikely that an answer to such a question will be accepted as enlightening; it is surely better to consider immigrants in different contexts, originating from particular types of situations. A recent trend in happiness studies is to consider how variables might have different consequences for different types of individuals (a question readily addressed in a quantitative framework via interaction terms—see Kroll 2011). That logic certainly merits extension to questions relating to migration, a field marked by great variation in people’s experiences, particularly by virtue of migration from and to different countries.

The present article, then, advances beyond existing research in two respects. First, it compares migrants to stayers (i.e. stayers in the origin country), rather than migrants to natives in the destination country. It also takes an initial step in exploring variation in happiness outcomes by considering country-level migration flows rather than resting content with an analysis of ‘economic migration’ in an undifferentiated sense. Small sample sizes inhibit a more thorough analysis of that sort (e.g. there are not enough Slovenian migrants in the data gathered from western European countries to enable meaningful comparisons for that flow), but the results presented below are sufficient to establish that there is indeed variation in migrants’ happiness outcomes.

3. Data and methods

Data for this analysis are taken from Rounds 4 and 5 of the European Social Survey (Jowell 2007).5 The goal is to analyse migrants originating in eastern European countries; the countries in question are Bulgaria, Croatia, Czech Republic, Estonia, Hungary, Latvia, Poland, Romania, Russia, Slovenia, Slovakia, Turkey, and Ukraine. The sample consists primarily of residents in those countries, i.e. ‘stayers’ (immigrants there were removed). The other portion of the sample comprises emigrants from those countries currently (at the time of the survey) living in the Western European countries covered by the ESS. This ‘slice’ captures individuals who have moved to wealthier countries, presumably with the intention of gaining the higher incomes available there. The total sample size is 42,380, of which 1,071 are migrants.

The dependent variable here is happiness. A common practice in happiness studies is to use answers to a single survey question; here that question (C1) is ‘Taking all things together, how happy would you say you are?’, with answers ranging from 0 to 10 (‘extremely unhappy’ to ‘extremely happy’, with the intervening numbers unlabelled). There is a variety of more elaborate approaches to the measurement of happiness (on which see e.g. Diener et al. 1985 for their multi-item ‘satisfaction with life scale’), but it is not apparent that using multi-item scales generally results in a different summary measure of happiness. For research that encompasses different countries, there are advantages in using a single measure insofar as it minimizes potential for different interpretations of specified dimensions; it is also preferable for this purpose to refrain from including a definition of happiness in the question (Graham 2009).

The variables that provide the mechanism for comparing migrants with stayers are straightforward. To identify migrants in Western European countries, I used questions asking respondents whether they were born in the country in which they currently live (C28) as well as a direct question asking for their country of birth (C29). For those flagged by these questions as immigrants, there is also a question asking how long ago they came to live in their current country of residence (C30); using that variable, the analysis below distinguishes those who arrived within the previous 5 years (‘Recent migrants’) from those who arrived 6 or more years ago (‘Established migrants’). The analysis also considers differences among migrants living in different destination countries, using the ‘country’ variable.

Variables used as controls in regression models below include a number of factors well established in existing research as significantly associated with happiness (for reviews regarding these and other factors, see Dolan et al. 2008; Diener et al. 2009; Bartram 2012). Respondents’ evaluation of their health is given in response to a question (C15) with five options: very good, good, fair, bad, and very bad. A question asking how religious one is (C21) gives respondents eleven options on a scale similar to that of the happiness question (0 for not at all religious, 10 for very religious). ‘Unemployment’ is indicated by a question (F8) asking about ‘main activity’; respondents are coded here as unemployed if they say they are unemployed whether actively looking for a job or not. A post-coded question (F35) indicates directly (using responses to questions about relationships to others living in the household) whether someone is living with a spouse or partner. A separate question (C3) asks respondents whether they have someone with whom they can discuss intimate and personal matters (yes/no), labelled in tables below as ‘friends’. A categorical variable for education indicates whether respondents have only a primary education (or less) as against attainment at lower secondary, upper secondary, vocational, or tertiary levels. There are also straightforward questions for gender and age (F2 and F3); the square of age (divided by 100) is included to capture its typically U-shaped association with happiness.

As the discussion in the previous section makes clear, income is significantly associated with happiness in cross-sectional analyses. Income in the ESS is measured by presenting respondents with deciles derived from income distributions pertaining to the country of residence (F32). Using this type of measure here corresponds well to two features of the relationship between happiness and income, which (as noted in the previous section) is primarily a matter of comparing one’s relative position with that of others rather than the absolute magnitude of one’s income. One feature is that most people compare themselves with others in a relatively ‘local’ frame (cf. Firebaugh and Schroeder 2009); in this context the relevant point is that comparisons to others are unlikely to reach beyond national borders (though that issue is perhaps less clear in the case of international migrants—a matter for future research). Another point is that regression models gauging the association of happiness with income typically find that a logarithmic function fits best when income is given in ‘absolute’ amounts. On both grounds the ESS measure of income is manifestly preferable to any attempt to ‘standardize’ income by turning the ranges into numbers (e.g. via midpoints) and then converting to a single currency equivalent. As with most surveys, non-response rates for the income question are sometimes substantial, as discussed in the next section.

Basic descriptive summaries of these variables appear in Table 1, and data on happiness by country are presented in Table 2. As is clear from both parts of Table 2, the number of migrants by country of origin and country of destination are in certain instances quite small, a fact that inhibits some forms of fine-grained analysis. Also evident in Table 2 is the well-known fact that happiness is quite low in many Eastern European countries, certainly in comparison to Western Europe.

Table 1.

Characteristics of migrants and non-migrants

 Total Migrants Non-migrants 
Happiness 6.34 7.31 6.31 
    SD 2.29 1.97 2.29 
Age 47.1 39.5 47.3 
    SD 18.5 15.2 18.5 
Health 2.50 2.06 2.51 
    SD 0.95 0.90 0.95 
Religiosity 4.87 5.37 4.86 
    SD 3.03 2.89 3.03 
Income 4.89 4.59 4.90 
    SD 2.77 2.57 2.78 
Education    
    % Primary 8.0 9.9 8.0 
    % Lower secondary 19.4 18.6 19.5 
    % Upper secondary 44.4 35.8 44.6 
    % Vocational 6.8 10.1 6.7 
    % Tertiary 21.3 25.5 21.2 
% Male 43.6 46.6 43.5 
% Partner 56.9 67.1 56.6 
% Unemployed 7.4 12.1 7.3 
% Friends 84.6 84.5 82.6 
N 42,380 1,071 41,309 
 Total Migrants Non-migrants 
Happiness 6.34 7.31 6.31 
    SD 2.29 1.97 2.29 
Age 47.1 39.5 47.3 
    SD 18.5 15.2 18.5 
Health 2.50 2.06 2.51 
    SD 0.95 0.90 0.95 
Religiosity 4.87 5.37 4.86 
    SD 3.03 2.89 3.03 
Income 4.89 4.59 4.90 
    SD 2.77 2.57 2.78 
Education    
    % Primary 8.0 9.9 8.0 
    % Lower secondary 19.4 18.6 19.5 
    % Upper secondary 44.4 35.8 44.6 
    % Vocational 6.8 10.1 6.7 
    % Tertiary 21.3 25.5 21.2 
% Male 43.6 46.6 43.5 
% Partner 56.9 67.1 56.6 
% Unemployed 7.4 12.1 7.3 
% Friends 84.6 84.5 82.6 
N 42,380 1,071 41,309 
Table 2.

Happiness in Europe by country

Eastern Stayers Emigrants to Western Europe n (migrants) 
Bulgaria 5.3 7.3 37 
Croatia 6.7 7.9 27 
Czech Republic 6.7 7.3 33 
Estonia 6.9 7.3 21 
Hungary 6.2 8.1 36 
Latvia 6.5 7.5 31 
Poland 7.2 7.5 337 
Romania 6.1 7.2 151 
Russia 6.1 7.2 156 
Slovakia 6.6 7.3 16 
Slovenia 7.3 9.5 
Turkey 5.5 7.0 183 
Ukraine 5.5 7.4 41 
Eastern Stayers Emigrants to Western Europe n (migrants) 
Bulgaria 5.3 7.3 37 
Croatia 6.7 7.9 27 
Czech Republic 6.7 7.3 33 
Estonia 6.9 7.3 21 
Hungary 6.2 8.1 36 
Latvia 6.5 7.5 31 
Poland 7.2 7.5 337 
Romania 6.1 7.2 151 
Russia 6.1 7.2 156 
Slovakia 6.6 7.3 16 
Slovenia 7.3 9.5 
Turkey 5.5 7.0 183 
Ukraine 5.5 7.4 41 
Western All Eastern European migrants n (migrants) 
Belgium 7.7 7.2 55 
Britain 7.4 7.6 69 
Denmark 8.3 8.4 31 
Finland 8.0 7.5 50 
France 7.1 6.6 18 
Germany 7.3 7.3 267 
Greece 6.3 6.3 70 
Ireland 7.1 7.3 201 
Netherlands 7.8 6.9 42 
Norway 8.0 8.1 52 
Portugal 6.5 5.8 11 
Spain 7.6 7.5 66 
Sweden 7.9 7.4 64 
Switzerland 8.0 7.6 83 
Western All Eastern European migrants n (migrants) 
Belgium 7.7 7.2 55 
Britain 7.4 7.6 69 
Denmark 8.3 8.4 31 
Finland 8.0 7.5 50 
France 7.1 6.6 18 
Germany 7.3 7.3 267 
Greece 6.3 6.3 70 
Ireland 7.1 7.3 201 
Netherlands 7.8 6.9 42 
Norway 8.0 8.1 52 
Portugal 6.5 5.8 11 
Spain 7.6 7.5 66 
Sweden 7.9 7.4 64 
Switzerland 8.0 7.6 83 

4. Analysis and results

Taking the sample as a whole, a bivariate comparison as per the first row of Table 1 shows that migrants are significantly happier than stayers (in a t-test, P < 0.000), with a happiness level one point higher on average (7.3 vs. 6.3). A key question is then whether these differences are robust to the inclusion of other variables associated with happiness, as per the review of previous research above—a question addressed here via ordinary least squares (OLS) regression. The dependent variable originates in a question with eleven options for response and is thus ordinal; in principle, then, the appropriate form of analysis would be an ordered logistic model. However, given the relatively large number of response options, OLS regression can be expected to produce equivalent results (Ferrer-i-Carbonell and Frijters 2004 make this point with particular reference to research on happiness), and indeed an ordered logit analysis would lead to identical conclusions here. OLS regression results are therefore presented: interpretation is more straightforward and it is also possible to display standardized coefficients. Sampling weights are used in all models except as noted.

The regression model in Table 3 demonstrates that migrants are indeed happier than stayers even when controlling for other determinants of happiness. Recent migrants are one-quarter of a point happier than stayers, and more established migrants are almost three-quarters of a point happier than stayers.6 That latter difference is statistically significant at a demanding level (admittedly not difficult to achieve with a sample size of 28,634—an issue addressed in more detail below). The model also includes a set of dummy variables (not shown) for the different countries of origin. All other variables in the model behave as expected and all are statistically significant at conventional levels (except for the difference between those with a lower-secondary education and those with a primary education).

Table 3.

Determinants of happiness

 b P β 
Recent migrants 0.25 0.035 0.02 
Established migrants 0.72 0.001 0.03 
Age −0.07 0.000 −0.58 
Age2/100 0.07 0.000 0.55 
Female 0.13 0.000 0.03 
Religiosity 0.07 0.000 0.09 
Partner 0.52 0.000 0.11 
Unemployed −0.57 0.000 −0.07 
Friends 0.60 0.000 0.10 
Health −0.69 0.000 −0.28 
Income 0.09 0.000 0.11 
Education    
    Lower secondary 0.14 0.123 0.02 
    Upper secondary 0.28 0.002 0.06 
    Vocational 0.27 0.007 0.03 
    Tertiary 0.41 0.000 0.08 
ESS Round 5 0.10 0.001 0.02 
Constant 6.34 0.000  
N 28,634   
Prob > F 0.000   
Adjusted R2 0.226   
 b P β 
Recent migrants 0.25 0.035 0.02 
Established migrants 0.72 0.001 0.03 
Age −0.07 0.000 −0.58 
Age2/100 0.07 0.000 0.55 
Female 0.13 0.000 0.03 
Religiosity 0.07 0.000 0.09 
Partner 0.52 0.000 0.11 
Unemployed −0.57 0.000 −0.07 
Friends 0.60 0.000 0.10 
Health −0.69 0.000 −0.28 
Income 0.09 0.000 0.11 
Education    
    Lower secondary 0.14 0.123 0.02 
    Upper secondary 0.28 0.002 0.06 
    Vocational 0.27 0.007 0.03 
    Tertiary 0.41 0.000 0.08 
ESS Round 5 0.10 0.001 0.02 
Constant 6.34 0.000  
N 28,634   
Prob > F 0.000   
Adjusted R2 0.226   

Note: Model includes origin country dummies.

One area of concern relates to the income variable. As noted above, large proportions of respondents declined to answer the income question; the result here is that the sample used in the model is quite a bit smaller (via listwise deletion) than the overall sample (68 per cent). A model (not shown) excluding the income variable uses 39,602 respondents (93 per cent); in that model there is no evidence of a happiness difference between recent migrants and stayers, while more established migrants were 0.71 points happier than stayers (there are only small changes for the other variables in the equation). A model was also created using multiple imputation for income (Rubin 1987; Royce 2004), resulting in small differences for the two main coefficients of interest: recent migrants are 0.34 points happier than stayers and established migrants are 0.83 points happier (as against 0.25 and 0.72 in Table 3).

The regression model above compares migrants with stayers across the sample as a whole. We can conduct separate regressions for countries (of origin) where there are sufficient number of migrants (n > 100), to determine whether the overall difference between migrants and stayers is replicated at the level of individual countries. Table 4 shows that migrants from Romania, Turkey, and Russia are happier than those who have remained, with a difference of a full point for Romania and Turkey (controlling for other variables). For Poland, on the other hand, there is no sign of a difference between migrants and stayers at all. (Models are the same as in Table 3 except that ‘recent’ and ‘established’ migrants are combined, given the already small size of the migrant samples; in addition, the country dummy variables are of course absent.)

Table 4.

Regression coefficients for ‘migrant’ in OLS models of happiness, by country of origin

 b migrant P n 
Poland 0.04 0.728 2749 
Romania 1.00 0.000 1625 
Russia 0.68 0.000 3957 
Turkey 0.98 0.001 2052 
 b migrant P n 
Poland 0.04 0.728 2749 
Romania 1.00 0.000 1625 
Russia 0.68 0.000 3957 
Turkey 0.98 0.001 2052 

Variation in migrants’ experiences is also apparent via analysis that considers separate coefficients corresponding to migrants’ country of residence. The coefficients for countries listed in Table 5 (with regression analysis otherwise equivalent to that in Table 3, including the use of origin country dummies, but without distinguishing between recent and established migrants) indicate the happiness difference between migrants in the specified country and stayers overall. For migrants moving from Eastern Europe to Denmark, the happiness difference (comparing to stayers overall) is large (1.72) and statistically significant even though there are only 31 such migrants (as per Table 2). Differences (happiness advantages) are also evident for Eastern European migrants in Finland, Germany, Norway, Spain, Sweden, and Switzerland. In other cases we must take care not to ‘accept the null hypothesis’—but there is no sign of a happiness difference (relative to stayers) for migrants who have moved to Ireland and Greece (the sustained economic difficulties of these countries is likely relevant). A similar statement might be made about migrants to France, but the number of such migrants in this sample is so small (17) that we are safer concluding that no conclusion can be drawn.7

Table 5.

Determinants of happiness, by country of destination

 b P 
Belgium 0.48 0.091 
Britain 0.38 0.069 
Denmark 1.72 0.000 
Finland 0.87 0.003 
France 0.07 0.899 
Germany 0.65 0.000 
Greece −0.10 0.694 
Ireland −0.05 0.780 
Netherlands 0.53 0.109 
Norway 0.90 0.001 
Spain 1.28 0.000 
Sweden 0.73 0.005 
Switzerland 1.01 0.000 
Age −0.07 0.000 
Age2/100 0.07 0.000 
Female 0.13 0.000 
Religiosity 0.07 0.000 
Partner 0.51 0.000 
Unemployed −0.56 0.000 
Friends 0.60 0.000 
Health −0.69 0.000 
Income 0.09 0.000 
Education   
    Lower secondary 0.14 0.129 
    Upper secondary 0.28 0.002 
    Vocational 0.27 0.006 
    Tertiary 0.41 0.000 
ESS Round 5 0.10 0.001 
Constant 6.36 0.000 
N 28,636  
Prob > F 0.000  
Adjusted R2 0.227  
 b P 
Belgium 0.48 0.091 
Britain 0.38 0.069 
Denmark 1.72 0.000 
Finland 0.87 0.003 
France 0.07 0.899 
Germany 0.65 0.000 
Greece −0.10 0.694 
Ireland −0.05 0.780 
Netherlands 0.53 0.109 
Norway 0.90 0.001 
Spain 1.28 0.000 
Sweden 0.73 0.005 
Switzerland 1.01 0.000 
Age −0.07 0.000 
Age2/100 0.07 0.000 
Female 0.13 0.000 
Religiosity 0.07 0.000 
Partner 0.51 0.000 
Unemployed −0.56 0.000 
Friends 0.60 0.000 
Health −0.69 0.000 
Income 0.09 0.000 
Education   
    Lower secondary 0.14 0.129 
    Upper secondary 0.28 0.002 
    Vocational 0.27 0.006 
    Tertiary 0.41 0.000 
ESS Round 5 0.10 0.001 
Constant 6.36 0.000 
N 28,636  
Prob > F 0.000  
Adjusted R2 0.227  

Note: Model includes origin country dummies.

Given that these are cross-sectional models, an additional area of concern is the possibility of endogeneity. Again, as there are no panel data for people who migrate across national borders, we have no direct way of knowing about the happiness of migrants prior to their migration. It is possible that the generally higher level of happiness among migrants represents nothing more than higher levels of happiness among people who choose to become migrants (and not an increase in happiness following migration). A recent article by Graham and Markowitz (2011) casts doubt on this possibility: they investigated the happiness of potential migrants in Latin American countries, comparing those expressing an intention to migrate with those lacking such an intention. The analysis, using data from the Latinobarametro, shows that intending migrants are less happy on average, despite having higher levels of ‘objective’ well-being on various dimensions; Graham and Markowitz conclude that migrants are ‘frustrated achievers’, inclined toward migration because of dissatisfaction with their circumstances. A similar finding emerges in research on intention to migrant among people in Central/Eastern Europe (Popova and Ostrachshenko 2011): those with lower life satisfaction are more likely to express an intention to migrate (see also Blanchflower et al. 2007).

An analysis designed to correct for endogeneity with respect to actual migrants (as compared with stayers) suggests a different conclusion, however. In a two-stage ‘treatment effects’ equation (with the first stage equivalent to a probit model of the migration decision, using respondent’s age, gender, education, mother’s education, and father’s education), the coefficient indicating the difference in happiness between migrants and stayers is not statistically significant (Table 6).8 (The migrant variable for this analysis must be binary, so it is not possible to distinguish here between recent and established migrants; in all other respects the model is equivalent to that reported in Table 3.) The implication, then, is that in general migrants moving from Eastern Europe to Western Europe are positively selected on happiness (a point evident also in the positive figure for ‘ρ’ in Table 6)—such that there is no support in this model for the conclusion that migration brought an increase in happiness.

Table 6.

Determinants of happiness, two-stage ‘treatment’ model

 b P 
Migrant −0.53 0.383 
Age −0.07 0.000 
Age2/100 0.07 0.000 
Female 0.14 0.000 
Religiosity 0.09 0.000 
Partner 0.49 0.000 
Unemployed −0.52 0.000 
Friends 0.64 0.000 
Health −0.70 0.000 
Income 0.09 0.000 
Education   
    Lower secondary 0.21 0.000 
    Upper secondary 0.34 0.000 
    Vocational 0.33 0.000 
    Tertiary 0.46 0.000 
ESS Round 5 0.09 0.003 
Constant 6.19 0.000 
Migration equation   
    Age −0.01 0.000 
    Gender −0.07 0.026 
    Education   
        Lower secondary 0.06 0.417 
        Upper secondary 0.08 0.239 
        Vocational 0.32 0.000 
        Tertiary 0.22 0.003 
 Father's education  
        Lower secondary −0.11 0.095 
        Upper secondary 0.00 0.987 
        Vocational 0.16 0.069 
        Tertiary 0.33 0.000 
 Mother's education  
        Lower secondary −0.38 0.000 
        Upper secondary −0.45 0.000 
        Vocational −0.59 0.000 
        Tertiary −0.52 0.000 
    Constant −1.06 0.000 
λ 0.51 0.056 
ρ 0.26  
σ 1.98  
N 26,013  
Wald χ2 (33) 8658.89  
 b P 
Migrant −0.53 0.383 
Age −0.07 0.000 
Age2/100 0.07 0.000 
Female 0.14 0.000 
Religiosity 0.09 0.000 
Partner 0.49 0.000 
Unemployed −0.52 0.000 
Friends 0.64 0.000 
Health −0.70 0.000 
Income 0.09 0.000 
Education   
    Lower secondary 0.21 0.000 
    Upper secondary 0.34 0.000 
    Vocational 0.33 0.000 
    Tertiary 0.46 0.000 
ESS Round 5 0.09 0.003 
Constant 6.19 0.000 
Migration equation   
    Age −0.01 0.000 
    Gender −0.07 0.026 
    Education   
        Lower secondary 0.06 0.417 
        Upper secondary 0.08 0.239 
        Vocational 0.32 0.000 
        Tertiary 0.22 0.003 
 Father's education  
        Lower secondary −0.11 0.095 
        Upper secondary 0.00 0.987 
        Vocational 0.16 0.069 
        Tertiary 0.33 0.000 
 Mother's education  
        Lower secondary −0.38 0.000 
        Upper secondary −0.45 0.000 
        Vocational −0.59 0.000 
        Tertiary −0.52 0.000 
    Constant −1.06 0.000 
λ 0.51 0.056 
ρ 0.26  
σ 1.98  
N 26,013  
Wald χ2 (33) 8658.89  

Note: Model includes origin country dummies.

Using the same correction with respect to the models in Table 4 analysing the four main migrant-sending countries separately (models not shown), one arrives at a more forceful statement regarding variation in the experience of the migrants investigated here. For Romania, Russia, and Turkey, the test comparing the OLS model and the treatment model fails to provide support for preferring the treatment model—and so the results reported in Table 4 (indicating higher happiness among migrants) are not evidently a consequence of positive selection on happiness. For Poland, however, the correction for endogeneity is required (‘λ’ is significant at P = 0.004)—and the happiness coefficient for migrants is −1.03 (P = 0.012). A model that corrects for selection into migration, then, suggests that migrants to Western Europe from Poland experience a decrease in happiness of more than a point on the 11-point scale.

These results must be taken with a certain degree of caution. Treatment effects models (like two-stage models generally) are sensitive to mis-specification at the first stage; the concern here, then, is that there is limited information in the dataset for predicting the migration decision. On the other hand, all results reported here are not notably different when using multiple imputation with respect to the income variable; in the treatment effects model with multiple imputation for Poland, the analysis uses 97 per cent of the sample and produces an almost identical result (b = −1.10).

5. Discussion

The most interesting result here emerges in the evident variation in migrants’ experiences considered at country (of origin) level. For several countries there is clear evidence of greater happiness among migrants compared with stayers, but Polish emigrants are significantly less happy than stayers; when we adjust for positive selection (happier people are more likely to migrate), there is support for the conclusion that this difference represents a decrease for Polish emigrants. Even when migrants are happier than stayers, they can be happier to different extents (as the Russian result indicates, in comparison to the larger differences for Romania and Turkey). For the sample overall, when we address the possibility of endogeneity with the treatment-effects model, there is no support for the conclusion that happiness increased for migrants following migration—and that statement might well pertain to some of the origin countries where sample sizes were too small for separate analysis.

To the extent that analysis of the entire sample leads us to perceive no increase in happiness for migrants overall, we would surely want to ask: why not? Average levels of happiness in eastern European countries are notably lower than averages for countries in western Europe—even lower than would be expected on the basis of ‘objective’ characteristics (Deaton 2008, Inglehart et al. 2008). International comparisons of this sort sometimes provoke the question of whether different averages in reported happiness actually reflect different levels of happiness. Reviews on this topic consider the possibility that people in different cultures might use numbers on questionnaires differently or might engage in different modes of self-presentation (Oishi 2010). Oishi concludes that issues of this type do not play a major role in cross-national variation in happiness; Veenhoven (2001) asking ‘Are the Russians as unhappy as they say they are?’ concludes that the answer is yes.

Even so, these issues have been addressed mainly in a ‘static’ way, at least insofar as there is very little research exploring how migrants might adapt to a different ‘happiness culture’ in the destination country (perhaps with consequences for the happiness they report on surveys). Senik’s (2011) investigation of the ‘French unhappiness puzzle’ (lower happiness in France relative to what one would expect from objective circumstance there) suggests that ‘culture’ is significant and stable (i.e. resistant to change in circumstances such as migration to another country) with respect to happiness. People who are educated in French schools (including those who subsequently emigrate) appear to develop a ‘mentality’ that contributes to a persistently lower level of happiness. Mentality and culture have a persistent influence on happiness in general (even if the degree of persistence varies by country, perhaps via the nature of the educational system); as Sheldon and Lyubomirsky (2006) suggest, efforts to change one’s circumstances typically have only a limited impact on one’s happiness.

Again, however, that general statement coexists with variation: migrants from some countries do appear (via analysis here) to have increased their happiness via migration. The divergent outcomes at country-level evident help show some limits of quantitative analysis for understanding migration and its happiness consequences. Statistical models can help us understand the role of individual-level factors, but individuals are situated in—and migrate between—particular countries. Alternative quantitative approaches could point to some broad hypotheses for understanding the happiness outcomes of different migration streams. But a good account of different patterns would require an analysis drawing as much on history and context as on survey data. That sort of analysis would help us progress beyond the approach, apparent in research to date, that asks whether migration generally leads to increased or decreased happiness.

The low average happiness among people in Eastern European countries (particularly in comparison to happiness in Western Europe) suggests that east-to-west migration in Europe might amount to an instance where ‘optimism’ about happiness consequences would have been a reasonable expectation. Intuitively, it might have been expected that migrants from countries with low average happiness would experience an increase after moving to a happier country. That expectation is borne out to an extent by the country-level analyses: happiness appears to have increased for migrants from Romania, Russia, and Turkey (where average levels of happiness are quite low even for Eastern Europe), but not for migrants from Poland (where average levels of happiness are comparable to those in many Western European countries). But it is not borne out for east-to-west migrants as a whole. We might then be less optimistic about finding favourable happiness outcomes for migrants where origin-country happiness is already relatively high (the analysis for Poland here is especially pertinent to that suggestion). That latter pattern describes migration from Latin America to the USA, for example: average happiness among Mexicans is already on par with happiness in the USA. The contrast (origin/destination countries on par vs. migration from countries with low average happiness to countries where it is higher) reinforces the notion that migration streams are likely to differ in their happiness consequences.

A number of questions about migration and happiness thus merit further attention. Three issues stand out for the analysis presented here. To determine whether it is necessary to correct for endogeneity of happiness with respect to the migration decision, we would want a better model for the first-stage equation, to predict the migration decision. The analysis above suggests that migrants from some countries evince a happiness advantage while migrants from other countries don’t—but we do not have research that attempts to explain this variation. And the focus here is on migration framed as motivated by aspiration for a specific type of economic gain; in many instances migration flows consist more of refugees and/or people seeking to join family members who migrated earlier, and the ideas discussed in this article might have limited relevance to their experience (even if they too are affected by the economic differences between origin and destination countries and by the transition to a new situation). In each case, we would benefit from having better data than that currently available.

Acknowledgements

I am grateful to Patrick White, Maarja Lühiste, and Stephen Wood for very helpful comments and advice on this paper.

Conflict of interest statement. None declared.

Notes

1. The happiness consequences of migration for purposes of risk diversification and access to credit (Stark 1991) would form a separate question.
2. The data do not give information on migrants’ motivations for migration, and some east-to-west migrants might have had non-economic motivations. Even in such cases, however, the economic aspects of migrants’ experiences might be similar to those motivated by aspiration for economic gain.
3. This assumption is essential to conventional perspectives in economics—‘revealed preferences’—but it is found elsewhere as well.
4. Another question at the intersection of migration and happiness studies is the effect of migration on average happiness in destination countries; Polgreen and Simpson (2011) find that net migration raises happiness for relatively unhappy countries but lowers it for relatively happy countries.
5. Use of earlier rounds of the ESS was not possible, given that the income variable was changed significantly with Round 4 (and was quite unsatisfactory in the earlier rounds).
6. The latter figure is determined via an interaction term between migrant status and a dichotomous variable indicating ‘recent’ vs. ‘established’ migrants. Rather than report the coefficient for the interaction term (0.47), Table 3 gives the difference between established migrants and stayers (0.72) by adding the interaction term coefficient to the coefficient for ‘recent migrants’ (0.25).
7. For Table 5, migrants to Portugal were dropped from the analysis: there are only 11 in the entire sample, and only 3 of these answered the income question (and so could enter the regression).
8. See Cong and Drukker (2001) for details of Stata’s ‘treatreg’ routine and Maddala (1983) for the underlying model. The analysis described here uses the more conservative ‘two-step’ option in place of full-information maximum likelihood; use of sample weights is then not possible. ‘λ’ is significant only at a less demanding threshold of P < 0.10; even so, at P = 0.056 it is quite reasonable to conclude that the errors at the two stages are correlated and so correction for endogeneity is necessary.

References

Aycan
Z
Berry
J W
‘Impact of Employment-Related Experiences on Immigrants’ Psychological Well-Being and Adaptation to Canada’
Canadian Journal of Behavioural Science
 , 
1996
, vol. 
28
 
3
(pg. 
240
-
51
)
Bălţătescu
S
Pop
L
Matiuţă
C
‘Subjective well-being of immigrants in Europe and their evaluation of societal conditions: an exploratory study’
European Identity and Free Movement of Persons in Europe
 , 
2005
Oradea
University of Oradea Publishing House
(pg. 
128
-
43
)
Bălţătescu
S
‘Central and Eastern Europeans Migrants’ Subjective Quality of Life: A Comparative Study’
Journal of Identity and Migration Studies
 , 
2007
, vol. 
1
 
2
(pg. 
67
-
81
)
Bartolini
S
Bruni
L
Porta
P L
‘Why are people so unhappy? Why do they strive so hard for money? Competing explanations of the broken promises of economic growth’
Handbook on the Economics of Happiness
 , 
2007
Cheltenham
Edward Elgar
(pg. 
337
-
64
)
Bartram
D
‘Economic Migration and Happiness: Comparing Immigrants’ and Natives’ Happiness Gains from Income’
Social Indicators Research
 , 
2011
, vol. 
103
 
1
(pg. 
57
-
76
)
Bartram
D
‘Elements of a Sociological Contribution to Happiness Studies’
Sociology Compass
 , 
2012
, vol. 
6
 
8
(pg. 
644
-
56
)
Bayram
N
Thorburn
D
Demirhan
H
Bilgel
N
‘Quality of Life among Turkish Immigrants in Sweden’
Quality of Life Research
 , 
2007
, vol. 
16
 
8
(pg. 
1319
-
33
)
Blanchflower
D G
Saheleen
J
Shadforth
C
‘The Impact of the Recent Migration from Eastern Europe on the UK Economy’
2007
 
IZA Discussion Paper No. 2615
Boyce
C J
Brown
G D A
Moore
S C
‘Money and Happiness: Rank of Income, Not Income, Affects Life Satisfaction’
Psychological Science
 , 
2010
, vol. 
21
 
4
(pg. 
471
-
5
)
Christoph
B
‘The Relation Between Life Satisfaction and the Material Situation: A Re-Evaluation Using Alternative Measures’
Social Indicators Research
 , 
2010
, vol. 
98
 
3
(pg. 
475
-
99
)
Clark
A E
Frijters
P
Shields
M
‘Relative Income, Happiness and Utility: An Explanation for the Easterlin Paradox and Other Puzzles’
Journal of Economic Literature
 , 
2008
, vol. 
46
 
1
(pg. 
95
-
144
)
Cong
R
Drucker
D M
‘Treatment Effects Model’
Stata Technical Bulletin
 , 
2001
, vol. 
10
 
55
(pg. 
25
-
33
)
De Jong
G F
Chamratrithirong
A
Tran
Q-G
‘For Better, for Worse: Life Satisfaction Consequences of Migration’
International Migration Review
 , 
2002
, vol. 
36
 
3
(pg. 
838
-
63
)
Deaton
A
‘Income, Health, and Well-Being Around the World: Evidence from the Gallup World Poll’
Journal of Economic Perspectives
 , 
2008
, vol. 
22
 
2
(pg. 
53
-
72
)
Diener
E
Emmons
R A
Larsen
R J
Griffin
S
‘The Satisfaction with Life Scale’
Journal of Personality Assessment
 , 
1985
, vol. 
49
 
1
pg. 
71
 
Diener
E
Lucas
R
Schimmack
U
Helliwell
J
Well-Being for Public Policy
 , 
2009
Oxford
Oxford University Press
Dolan
P
Peasgood
T
White
M
‘Do We Really Know What Makes Us Happy? A Review of the Economic Literature on the Factors Associated with Subjective Well-Being’
Journal of Economic Psychology
 , 
2008
, vol. 
29
 
1
(pg. 
94
-
122
)
Dreby
J
Divided By Borders: Mexican Migrants and their Children
 , 
2010
Berkeley
University of California Press
Easterlin
R A
‘Does Money Buy Happiness?’
Public Interest
 , 
1973
, vol. 
30
 (pg. 
3
-
10
)
Easterlin
R A
‘Income and Happiness: Towards a Unified Theory’
The Economic Journal
 , 
2001
, vol. 
111
 
473
(pg. 
465
-
84
)
Easterlin
R A
Angelescu
L
‘Happiness and Growth the World Over: Time Series Evidence on the Happiness-Income Paradox’
2009
 
IZA Discussion Paper, No. 4060
Easterlin
R A
, et al.  . 
‘The Happiness-Income Paradox Revisited’
Proceedings of the National Academy of Sciences
 , 
2010
, vol. 
107
 
52
(pg. 
22463
-
8
)
Ferrer-i-Carbonell
A
Frijters
P
‘How Important is Methodology for the Estimates of the Determinants of Happiness?’
The Economic Journal
 , 
2004
, vol. 
114
 
497
(pg. 
641
-
59
)
Firebaugh
G
Schroeder
M B
‘Does Your Neighbor’s Income Affect Your Happiness?’
American Journal of Sociology
 , 
2009
, vol. 
115
 
3
(pg. 
805
-
31
)
Fischer
C
‘What Wealth-Happiness Paradox? A Short Note on the American Case’
Journal of Happiness Studies
 , 
2008
, vol. 
9
 
2
(pg. 
219
-
26
)
Gilbert
D
Stumbling on Happiness
 , 
2006
New York
HarperCollins
Graham
C
Happiness Around the World: The Paradox of Happy Peasants and Miserable Millionaires
 , 
2009
Oxford
Oxford University Press
Graham
C
Markowitz
J
‘Aspirations and Happiness of Potential Latin American Immigrants’
Journal of Social Research and Policy
 , 
2011
, vol. 
2
 
2
(pg. 
9
-
25
)
Headey
B
Muffels
R
Wooden
M
‘Money Does Not Buy Happiness: Or Does It? A Reassessment Based on the Combined Effects of Wealth, Income and Consumption’
Social Indicators Research
 , 
2008
, vol. 
87
 
1
(pg. 
65
-
82
)
Inglehart
R
Foa
R
Peterson
C
Welzel
C
‘Development, Freedom, and Rising Happiness: A Global Perspective (1981–2007)’
Perspectives on Psychological Science
 , 
2008
, vol. 
3
 
4
(pg. 
264
-
85
)
Jowell
R
European Social Survey, Technical Report
 , 
2007
London
Centre for Comparative Social Surveys, City University
Kahneman
D
, et al.  . 
‘Would You Be Happier if You Were Richer? A Focusing Illusion’
Science
 , 
2006
, vol. 
312
 
5782
(pg. 
1908
-
10
)
Kasser
T
The High Price of Materialism
 , 
2003
Boston
MIT Press
Knight
J
Gunatilaka
R
‘Great Expectations? The Subjective Well-Being of Rural-Urban Migrants in China’
World Development
 , 
2010
, vol. 
38
 
1
(pg. 
113
-
24
)
Kroll
C
‘Different Things Make Different People Happy: Examining Social Capital and Subjective Well-Being by Gender and Parental Status’
Social Indicators Research
 , 
2011
, vol. 
104
 
1
(pg. 
157
-
77
)
Layard
R
Happiness: Lessons from a New Science
 , 
2005
New York
Penguin Press
Maddala
G S
Limited Dependent and Qualitative Variables in Economics
 , 
1983
New York
Cambridge University Press
Maryanski
A
Turner
J H
The Social Cage: Human Nature and the Evolution of Society
 , 
1992
Stanford
Stanford University Press
Matt
S J
‘You Can’t Go Home Again: Homesickness and Nostalgia in U.S. History’
The Journal of American History
 , 
2007
, vol. 
94
 
2
(pg. 
469
-
97
)
Melzer
S M
‘Does Migration Make You Happy? The Influence of Migration on Subjective Well-Being’
Journal of Social Research and Policy
 , 
2011
, vol. 
2
 
2
(pg. 
73
-
92
)
Michalos
A C
‘Multiple Discrepancy Theory (MDT)’
Social Indicators Research
 , 
1985
, vol. 
16
 
4
(pg. 
347
-
417
)
Nowok
B
van Ham
M
Findlay
A M
Gayle
V
‘Does Migration Make You Happy? A Longitudinal Study of Internal Migration and Subjective Well-Being’
2011
 
IZA Discussion Paper, No. 6140
Oishi
S
Diener
E
Kahneman
D
Helliwell
J
‘Culture and Well-Being: Conceptual and Methodological Issues’
International Differences in Well-Being
 , 
2010
Oxford
Oxford University Press
(pg. 
34
-
69
)
Polgreen
L
Simpson
N
‘Happiness and International Migration’
Journal of Happiness Studies
 , 
2011
, vol. 
12
 
5
(pg. 
819
-
40
)
Popova
O
Ostrachshenko
V
‘Life (Dis)Satisfaction and Decision to Migrate: Evidence from Central and Eastern Europe’
2011
 
Osteuropa-Institut Regensburg Working Paper No. 306
Portes
A
Bach
R L
Latin Journey: Cuban and Mexican Immigrants in the United States
 , 
1985
Berkeley
University of California Press
Royce
P
‘ICE: Stata Module for Multiple Imputation of Missing Values’
Stata Journal
 , 
2004
, vol. 
4
 
3
(pg. 
227
-
41
)
Rubin
D B
Multiple Imputation for Nonresponse in Surveys
 , 
1987
New York
John Wiley
Safi
M
‘Immigrants’ Life Satisfaction in Europe: Between Assimilation and Discrimination’
European Sociological Review
 , 
2010
, vol. 
26
 
2
(pg. 
159
-
71
)
Senik
C
‘The French Unhappiness Puzzle: The Cultural Dimension of Happiness’
2011
 
IZA Discussion Paper 6175, Institute for the Study of Labor
Sheldon
K M
Lyubomirsky
S
‘Achieving Sustainable Gains in Happiness: Change Your Actions, Not Your Circumstances’
Journal of Happiness Studies
 , 
2006
, vol. 
7
 
1
(pg. 
55
-
86
)
Stark
O
The Migration of Labor
 , 
1991
Oxford
Basil Blackwell
Stevenson
B
Wolfers
J
‘Economic Growth and Subjective Well-Being: Reassessing the Easterlin Paradox’
2008
 
IZA Discussion Paper 3654, Institute for the Study of Labor
Stutzer
A
‘The Role of Income Aspirations in Individual Happiness’
Journal of Economic Behavior and Organization
 , 
2003
, vol. 
54
 
1
(pg. 
89
-
109
)
Thin
N
Social Happiness: Theory into Policy and Practice
 , 
2012
Bristol
The Policy Press
Veenhoven
R
‘Is Happiness Relative?’
Social Indicators Research
 , 
1991
, vol. 
24
 
1
(pg. 
1
-
34
)
Veenhoven
R
‘Are the Russians As Unhappy As They Say They Are?’
Journal of Happiness Studies
 , 
2001
, vol. 
2
 
2
(pg. 
111
-
36
)
Veenhoven
R
Eid
M
Larsen
R
‘Sociological Theories of Subjective Well-Being’
The Science of Subjective Well-Being: A Tribute to Ed Diener
 , 
2008
New York
Guilford Publications
(pg. 
44
-
61
)
Vohra
N
Adair
J
‘Life Satisfaction of Indian Immigrants in Canada’
Psychology and Developing Societies
 , 
2000
, vol. 
12
 
2
(pg. 
109
-
38
)
Ying
Y-W
‘Immigration Satisfaction of Chinese Americans: An Empirical Examination’
Journal of Community Psychology
 , 
1996
, vol. 
24
 
1
(pg. 
3
-
16
)