Skip to main content

Structural Changes in Economic Growth and Well-Being: The Case of Italy’s Parabola


The controversies on the relationship (called ‘gradient’) between the time trend of GDP and of subjective well-being oppose those who claim that the gradient is positive, to those who argue that it is nil. The possible existence of significant changes of the two trends and of the gradient within the same country is a challenge to both views. By focusing on the case of Italy, we show that the long-run trends of GDP and of well-being turned from increasing to decreasing, and that the gradient exhibits a rise through two structural breaks. This evidence is consistent with the ‘loss aversion’ hypothesis. From this macro-analysis we further go into micro-analysis to explain why subjective well-being declined so steeply. We find that the erosion of trust in others, the increase of financial dissatisfaction, worsened health, and greater importance of income contribute to the decline of subjective well-being. As far as we know, this is the first attempt to test the role of loss aversion in the long run using both macro- and micro-economic approaches.

This is a preview of subscription content, access via your institution.

Fig. 1
Fig. 2
Fig. 3
Fig. 4


  1. 1.

    ‘Loss aversion’ occurs when “the aggravation that one experiences in losing a sum of money appears to be greater than the pleasure associated with gaining the same amount” ( Kahneman and Tversky 1979, p. 279). The need to further investigate the ‘loss aversion’ effect within the Economics of Happiness has been recently raised in the survey by Clark (2018).

  2. 2.

    GDP per capita is at PPP constant 2011 international dollars (World Bank data).

  3. 3.

    The data of the figure are drawn from three sources: from the Eurobarometer Survey, which provides the share of ‘very’ and ‘fairly’ satisfied with life for the entire period, from the European Values Survey, which provides the average life satisfaction over 0-10 scale for only 1981, 1990, 1999 and 2005, and from the European Quality of Life Survey, which provides the same index for 2003, 2007, 2009 and 2016.

  4. 4.

    In the micro-economic analysis, our measure of income is household income after taxes (except in 1981 when the question did not specify whether income should be before or after taxes). For comparability over time, household income before 1999 has been convered in real Euro of 2010. The conversion to Euro follows the fixed exchange rate of 1936.27 Lire per 1 Euro, whereas deflation uses GDP deflator from World Development Indicators (World Bank).

  5. 5.

    These two data sources report the average of people’s life satisfaction. The Gallup World Poll, as fourth data source, reports a significant decline of life evaluation in Italy from 2005-2007 to 2014-2016 (Helliwell et al. 2017). ISTAT, as fifth data source, confirms and updates the decline for 2017 with respect to both 2016 and 2010. For the details of the five sources see:

  6. 6.

    The ‘Easterlin paradox’ arose from the pioneering 1974 and 1995 Easterlin’ s articles, and it states that 'at a point in time happiness varies directly with income both among and within nations, but over time happiness does not trend upward as income continues to grow. “Happiness” is used here interchangeably with subjective well-being as a proxy for all evaluative measures of selfreported feelings of well-being, including life satisfaction [\(\cdots\)]. “Income” is a proxy for real GDP per capita, the standard unitary measure of economic growth’ (Easterlin 2017).

  7. 7.

    The most significant break of the gradient within the entire period is detected in 2007 (Wald test=60.8, p=0.0000), but the test for the break in 2005 is very similar (Wald test=58.9, p=0.0000). The most significant break of the gradient between 1973 and 2005 is detected in 1989 (as reported in the text), which is the same year as the most significant break of the gradient between 1973 and 2007.

  8. 8.

    If the gradient is estimated by using one-year lagged GDP, the structural changes are confirmed, although the coefficient in the third period has p=0.14. If GDP growth is added as a regressor in the life satisfaction equation, then it is never significant. If other macroeconomic regressors are introduced, like inflation, unemployment, etc., the few observations hamper significant results.

  9. 9.

    In the same year, the traditional parties lost the elections, many politicians were prosecuted for corruption, and two magistrates who led investigations on the mafia were killed.

  10. 10.

    The social consequences of these shocks are worsened by the increased generational inequality that began in the mid-1990s. In fact, equivalent income of older cohorts of household heads improved, while that of younger cohorts deteriorated (Berloffa and Villa 2010).

  11. 11.

    Although EVS data are available until 2009, we limited the analysis of EVS data to the period 1981-2005 because income data are not available in 2009. However, EQLS provides data for 2003, 2007, 2011 and 2016 which complement and extend EVS data.

  12. 12.

    Ferrer-i Carbonell and Frijters (2004, p. 655).

  13. 13.

    Recall that we omit the year 2009 of the EVS because data about household income are not available for Italy in that year.

  14. 14.

    For more details, please, see

  15. 15.

    Measuring social comparisons using financial dissatisfaction may raise concerns when used to explain life satisfaction. To alleviate this concerns, we run additional estimates in which we replace financial dissatisfaction with a classic measure of relative income. Results do not change if we use the alternative specification. For details, please, refer to Appendix B.

  16. 16.

    The equivalised income is calculated by dividing the household’s total income by its equivalent size, which is computed by attaching specific weights to each member of the household. EVS, however, provides neither equivalised income nor household composition, at least for Italy in 1981 and 2005, thus differing from EQLS. We are nevertheless able to compute equivalised household income for 1981 and show that, if replaced with non-equivalised income, the results do not change significantly. The correlation between the two measures is 73% (significant at 1%; N = 1008). If we linearize income using the logarithm, the correlation is 78% (significant at 1%; N = 1008). The inclusion of equivalised household income in Eq. 1 yields a coefficient of -0.447 significant at 5%, which is only marginally different from the coefficient of non-equivalised household income (-0.498 significant at 5%) (in both cases N = 999).

  17. 17.

    For details about the use of a classic measure of relative income, please, refer to Appendix B.

  18. 18.

    The complete set of results from the OLS model is available in Appendix C.

  19. 19.

    Tables with the complete set of results are available in Appendix C.

  20. 20.

    The detailed results from the decomposition are available in Tables 16, 18 and 20 in Appendices D and E.

  21. 21.

    The trend of income inequality is confirmed if we use the standard deviation of income as a measure of income inequality (see Fig. 6 in Appendix F). The standard deviation of income is computed by year on EVS and EQLS samples.

  22. 22.

    More formally, if financial dissatisfaction (F) depends on others’ income (\(Y^o\)) so that \(F = Y^o - Y\), then the life satisfaction regression can be written as: \(LS = \alpha + \beta \cdot Y - \gamma \cdot (Y^o - Y) + \varepsilon\) which is equivalent to: \(LS = \alpha + (\beta + \gamma ) \cdot Y - \gamma \cdot Y^o + \varepsilon\) so that the coefficient of income (\(\theta = \beta + \gamma\)) is positive even if \(\beta < 0\), provided that \(\gamma > \beta\). This is true in our regressions in which financial dissatisfaction attracts large coefficients.


  1. Bartolini, S., Bilancini, E., & Pugno, M. (2013). Did the decline in social connections depress Americans‘ happiness? Social Indicators Research, 110(3), 1033–1059.

    Article  Google Scholar 

  2. Bartolini, S., & Bonatti, L. (2008). Endogenous growth, decline in social capital and expansion of market activities. Journal of Economic Behavior and Organization, 67(3), 917–926.

    Article  Google Scholar 

  3. Bartolini, S., & Sarracino, F. (2015). The dark side of Chinese growth: Declining social capital and well-being in times of economic boom. World Development, 74, 333–351.

    Article  Google Scholar 

  4. Becchetti, L., Massari, R., & Naticchioni, P. (2014). The drivers of happiness inequality: Suggestions for promoting social cohesion. Oxford Economic Papers, 66(2), 419–442.

    Article  Google Scholar 

  5. Berloffa, G., & Villa, P. (2010). Differences in equivalent income across cohorts of households: Evidence from Italy. Review of Income and Wealth, 56(4), 693–714.

    Article  Google Scholar 

  6. Blinder, A. (1973). Wage discrimination: Reduced-form and structural estimates. Journal of Human Resources, 8, 436–455.

    Article  Google Scholar 

  7. Bolt, J., Inklaar, R., de Jong, H., and van Zanden, J. (2018). Maddison project database, version 2018. for documentation and explanation of the data series.

  8. Brockmann, H., Delhey, J., Welzel, C., & Yuan, H. (2009). The China puzzle: Falling happiness in a rising economy. Journal of Happiness Studies, 10, 387–405.

    Article  Google Scholar 

  9. Clark, A., Frijters, P., & Shields, M. (2008). Relative income, happiness and utility: An explanation for the Easterlin paradox and other puzzles. Journal of Economic Literature, 46(1), 95–144.

    Article  Google Scholar 

  10. Clark, A. E. (2018). Four decades of the economics of happiness: Where next? Review of Income and Wealth, 64(2), 245–269.

    Article  Google Scholar 

  11. Clark, A. E., Flèche, S., Layard, R., Powdthavee, N., & Ward, G. (2017). The key determinants of happiness and misery. In J. F. Helliwell, R. Layard, & J. D. Sachs (Eds.), World Happiness Report 2017 (pp. 122–143). New York: Sustainable Development Solutions Network.

    Google Scholar 

  12. D'Ambrosio, C., & Frick, J. (2007). Income satisfaction and relative deprivation: An empirical link. Social Indicators Research, 81(3), 497–519.

    Article  Google Scholar 

  13. D'Ambrosio, C., & Frick, J. (2012). Individual well-being in a dynamic perspective. Economica, 79, 284–302.

    Article  Google Scholar 

  14. De Neve, J.-E., Ward, G., De Keulenaer, F., Van Landeghem, B., Kavetsos, G., & Norton, M. I. (2018). The asymmetric experience of positive and negative economic growth: Global evidence using subjective well-being data. Review of Economics and Statistics, 100(2), 362–375.

    Article  Google Scholar 

  15. Diener, E. (2006). Guidelines for national indicators of subjective well-being and ill-being. Journal of Happiness Studies, 7, 397–404.

    Article  Google Scholar 

  16. Easterlin, R. (1974). Does economic growth improve the human lot? Some empirical evidence. In P. David & W. Melvin (Eds.), Nations and households in economic growth (pp. 98–125). Palo Alto: Stanford University Press.

    Google Scholar 

  17. Easterlin, R. (2017). Paradox lost? Review of Behavioral Economics, 4(4), 311–339.

    Article  Google Scholar 

  18. Easterlin, R. A. (1995). Will raising the incomes of all increase the happiness of all? Journal of Economic Behavior and Organization, 27(1), 35–47.

    Article  Google Scholar 

  19. Easterlin, R. A., Angelescu, L., Switek, M., Sawangfa, O., & Zweig, J. S. (2010). The happiness-income paradox revisited. Proceedings of the National Academy of Sciences, 107(52), 1–6.

    Article  Google Scholar 

  20. Easterlin, R. A., Wang, F., & Wang, S. (2017). Growth and happiness in China, 1990–2015. In J. F. Helliwell, R. Layard, & J. D. Sachs (Eds.), World Happiness Report 2017 (pp. 48–83). New York: Sustainable Development Network Solutions.

    Google Scholar 

  21. EVS (2011). European values study 1981–2008, longitudinal data file. Technical report, GESIS Data Archive, Cologne. ZA4804 Data File Version 2.0.0.

  22. Ferrer-i Carbonell, A., & Frijters, P. (2004). How important is methodology for the estimates of the determinants of happiness? The Economic Journal, 114(497), 641–659.

    Article  Google Scholar 

  23. Helliwell, J., & Barrington-Leigh, C. (2010). Viewpoint: measuring and understanding subjective well-being. Canadian Journal of Economics, 43, 729–753.

    Article  Google Scholar 

  24. Helliwell, J., Huang, H., & Wang, S. (2017). The social foundations of world happiness. In J. Helliwell, R. Layard, & J. Sachs (Eds.), World happiness report 2017 (pp. 8–47). New York: Sustainable Development Solutions Network.

    Google Scholar 

  25. Herbst, C. M. (2011). 'paradoxical' decline? Another look at the relative reduction in female happiness. Journal of Economic Psychology, 32(5), 773–788.

    Article  Google Scholar 

  26. Jann, B. (2008). A Stata implementation of the Blinder-Oaxaca decomposition. Stata journal, 8, 453–459.

    Article  Google Scholar 

  27. Jenkins, S. P. (2015). World income inequality databases: An assessment of WIID and SWIID. The Journal of Economic Inequality, 13(4), 629–671.

    Article  Google Scholar 

  28. Kahneman, D., & Tversky, A. (1979). Prospect theory: An analysis of decision under risk. Econometrica, 47, 263–291.

    Article  Google Scholar 

  29. Kaiser, C., & Vendrik, M. (2019). How threatening are transformations of reported happiness to subjective wellbeing research? IZA Discussion Paper no. 13905.

  30. Knack, S., & Keefer, P. (1997). Does social capital have an economic payoff? A cross-country investigation. The Quarterly Journal of Economics, 112(4), 1251–1288.

    Article  Google Scholar 

  31. Becker, G., Rayo, L., & Krueger, A. (2008). Economic growth and subjective well-being: reassessing the easterlin paradox. Comments and Discussion. Brookings Papers on Economic Activity, 2008, 88–102.

    Article  Google Scholar 

  32. Layard, R., Mayraz, G., & Nickell, S. (2010). Does relative income matter? Are the critics right? In E. Diener, D. Kahneman, & J. Helliwell (Eds.), International differences in well-being (Vol. 6, pp. 139–166). New York: Oxford University Press.

    Chapter  Google Scholar 

  33. Mikucka, M., Sarracino, F., & Dubrow, J. K. (2017). When does economic growth improve life satisfaction? Multilevel analysis of the roles of social trust and income inequality in 46 countries, 1981–2012. World Development, 93, 447–459.

    Article  Google Scholar 

  34. Oaxaca, R. (1973). Male-female wage differentials in urban labor markets. International Economic Review, 14, 693–709.

    Article  Google Scholar 

  35. Oishi, S., & Kesebir, S. (2015). Income inequality explains why economic growth does not always translate to an increase in happiness. Psychological science, 26(10), 1630–1638.

    Article  Google Scholar 

  36. Proto, E., & Rustichini, A. (2013). A reassessment of the relationship between GDP and life satisfaction. PloS One, 8(11), e79358.

    Article  Google Scholar 

  37. Sacks, D. W., Stevenson, B., & Wolfers, J. (2012). Subjective well-being, income, economic development and growth. In P. Booth (Ed.), The pursuit of happiness (Vol. 3, pp. 59–98). London: The Institute of Economic Affairs.

    Google Scholar 

  38. Solt, F. (2016). The standardized world income inequality database. Social Science Quarterly, 97(5), 1267–1281.

    Article  Google Scholar 

  39. Stevenson, B., & Wolfers, J. (2008). Economic growth and subjective well-being: Reassessing the Easterlin Paradox. Brookings Papers on Economic Activity, 1, 1–87.

    Article  Google Scholar 

Download references


We wish to thank the participants at the Conference on “Happiness, Capabilities, and Opportunities” (15–17 November 2018, Rome), who discussed a previous version of this paper. Our especial thanks go to Richard Easterlin, Carol Graham, Stefano Bartolini, Marteen Vendrick, Kelsey O’Connor, and Gennaro Zezza whose suggestions have helped us to improve the paper. Any remaining errors remain obviously ours.

Author information



Corresponding author

Correspondence to Francesco Sarracino.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.


A Appendix

(See Tables 7, 8, 9, 10).

Table 7 The gradient between life satisfaction and the trend-component of real GDP per capita in three sub-periods of the 1973–2016 period in Italy
Table 8 The trends of life satisfaction, of real GDP per capita, and of their gradient in two sub-periods between 1973 and 2016 in Europe-8 (Belgium, Denmark, France, Germany, Ireland, Luxembourg, Netherland, and the United Kingdom)
Table 9 Test of structural break in the relationship between life satisfaction and real GDP per capita in the period 1973-2016 in the countries of Europe-8
Table 10 The trends of life satisfaction, of real GDP per capita, and of the gradient of life satisfaction with respect to per capita GDP in two sub-periods in Greece, Portugal, and Spain

B Alternatives to financial dissatisfaction

Financial dissatisfaction is our favourite proxy for social comparisons. The reason is that we do not need to make any hypothesis about what defines the reference group. However, financial dissatisfaction is subjective and this can give rise to endogeneity issues, in particular when the dependent variable is also subjective. Endogeneity would bias our coefficients in unpredictable ways, thus altering our conclusions. To what extent does this affect our conclusions? To address this issue, we run a new set of regressions in which we replaced financial dissatisfaction with a non-subjective measure of relative income.

Relative income is computed as the ratio between respondent’s income and the average income of her reference group. One of the limits of this approach is that reference groups can be defined in many ways. Here we adopted a classic definition of reference group as people belonging to the same age bracket, of the same gender, with a similar education level, and surveyed in the same year. To retain a reasonable number of groups (more than 70) with at least 10 observations per group we regrouped age in three categories (less than 30; between 30 and 65; above 65); and three education levels (primary, secondary and tertiary). We then created a dummy variable set to one if respondent’s income is below the reference income, and zero otherwise.

Results are presented graphically in Fig. 5 (see also Tables 11 and 12 for the detailed coefficients from EVS and EQLS data respectively). The coefficients of the income variable, and its interaction over time differ across models, but qualitatively we derive the same conclusions (see Fig. 5). This indicates that our conclusions about income are relatively unaffected by the choice of the proxy of social comparisons.

Fig. 5

Marginal effects of income on life satisfaction using financial dissatisfaction and a measure of relative income (respondent’s income below the reference income). Results from the two models are qualitatively similar. Note: average marginal effects after OLS regression with sample weight and clustered standard errors by year. The complete list of control variables is available in Sect. 4.1

Table 11 Estimates using European Values Study data. OLS with sample weights and robust standard errors clustered by year
Table 12 Estimates using European Quality of Life Survey data. OLS with sample weights and robust standard errors clustered by year

C Relationship between income and life satisfaction over time

(See Tables 13, 14, 15).

Table 13 OLS regression of life satisfaction over income, year, and their interaction
Table 14 OLS regression of life satisfaction over income, year, and their interaction: detailed results from the restricted model
Table 15 OLS regression of life satisfaction over income, year, and their interaction: detailed results from the complete model

D Detailed Results From the European Values Study

(See Tables 16, 17, 18, 19).

Table 16 Detailed decomposition of the life satisfaction gap betwen 1981 and 1990
Table 17 \(\beta\) coefficients and X-values for 1981 and 1990
Table 18 Detailed decomposition of the life satisfaction gap betwen 1990 and 2005
Table 19 \(\beta\) coefficients and X-values for 1990 and 2005

E Detailed Results From the European Quality of Life Survey

(See Tables 20, 21).

Table 20 Detailed decomposition of the life satisfaction gap between 2003 and 2016
Table 21 \(\beta\) coefficients and X-values for 2003 and 2016

F Standard Deviation of Income as a Measure of Inequality

(See Fig. 6).

Fig. 6

The Gini index follows closely the trajectory of the standard deviation of income. Note: Gini index of disposable income is issued from the Standardized World Income Inequality Database (Solt 2016). The standard deviation of income is computed using survey data from the EVS and the EQLS

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Pugno, M., Sarracino, F. Structural Changes in Economic Growth and Well-Being: The Case of Italy’s Parabola. Soc Indic Res (2021).

Download citation


  • Structural breaks
  • Life satisfaction
  • Economic growth
  • Loss aversion
  • GDP per capita

JEL Classification

  • I31
  • O11
  • O52