ESSAYS ON EMPIRICAL WELFARE ECONOMICS

Abstract

Chapter 1 Over the last decade we have witnessed an improvement in the literature on how to measure and compare the well-being of individuals. While the standard approach in the analysis of individual well-being is based almost exclusively on a material dimension of income, there is a robust evidence that individuals care about non-income dimensions of life which accordingly should be included in such a measure. Nevertheless, even when the non-income dimensions of life are included in the analysis of individual well-being, a majority of studies are predominantly neglecting the notion of individual preferences in order to escape the problem of interpersonal comparisons of well-being. In this chapter we have employed an alternative welfare measure which takes into account income and non-income dimensions of life while at the same time the following measure is sensitive to the individual preferences. This is known as the equivalent income measure. We have compared the equivalent income and income measures over 25 countries of the European Union. The following analysis is done for 2007 and 2011, since these were the only available years that we could use. Our contribution to the literature is reflected in the fact, that to the best of our knowledge, there are no such studies in the literature that compares the well-being with the following two measures for such a large set of countries. Although we found that the welfare rankings across countries change to some extent between the average income and average equivalent income, in spite of that when we ranked countries according to the growth rates of income and equivalent income we observed a substantial change in the country rankings. The following evidence implies that the choice of the welfare metrics is empirically important. The previous results have been broadened by computing the welfare rankings across countries once we raise the concern about the egalitarian principle of justice. We observed that the welfare rankings across countries changed remarkably when we take into account distributional inequalities. The evidence we have found suggest that correlations of disadvantages between life dimensions matter since some countries have sufficiently larger inequalities of equivalent income than inequalities of incomes. The results we have found show that individuals across countries care about material dimension (income) but they also care much about the non-income dimensions. Relatively the most important non-income dimension for almost all countries is health while the least important non-income dimension concerns the (un)employment status. We have identified the worst off individuals according to each welfare measure and we have compared the socio-demographic characteristics of the worst off individuals. We have seen that different well-being measures will identify the worst off individuals with different socio-demographic characteristics. We have also observed that the worst off individuals considerably differ over countries with respect to their average income and non-income dimensions and average socio-demographic characteristics. Finally, we have illustrated the degree of re-ranking between income and equivalent income measures taking into account all individuals of a given country. While we found a similar pattern of re-ranking between income and equivalent income across countries, the degree of re-ranking differs across countries. The pattern that we have observed across all countries indicated that individuals who are income rich can end up as equivalent income poor while the opposite has not been found. ; Chapter 2 An important role of social and public policies, among others, is to provide employment opportunities and to maintain the initiative of people to work. Yet, the role of these policies should not exclusively be concerned with the improvement of job quantity, such as ensuring the optimal employment rate but likewise these policies should be designed to improve the well-being of workers by enhancing job quality. Although, the precise definition of job quality is lacking in the literature, one can reasonably argue that job quality is a multi-dimensional concept which includes the wage dimension but also non-wage job dimensions such as job autonomy, job security, whether a job is interesting, challenging, whether it offers a good career opportunities, etc. Accepting the idea that a job quality is a multi-dimensional concept leads to the important question of aggregating various job dimensions into an overall index of job quality which can be used as a measure of well-being on the job. Moreover, if we agree on the notion that the preferences of workers over job dimensions should be respected, then we have to find a proper way to weight these various job characteristics such that the construction of weights is consistent with the preference orderings over different jobs. A measure that satisfies the previous two requirements is known in the literature as the equivalent wage measure. In this study we have applied the concept of equivalent wage to a specific sub-population of recent graduates (bachelor, master and doctoral students) who are currently participating in the labour market. In addition to the equivalent wage measure, we have used four other well-being measures which are wage, average preferences objective measure, equal weights objective measure and subjective job satisfaction measure. We have compared the job quality using a large scale survey which includes nineteen countries. We have found that individuals with various personal characteristics have different preferences over wage and non-wage job dimensions. This result underlines the importance of considering the heterogeneity of individual preferences seriously. We have shown that different measures of job quality will result in substantially different ranking of countries. In other words, the evidence we have found points out to the fact that the choice of well-being measure is utterly important for measuring job quality. Since, we have observed a considerable re-ranking of countries between different measures, we were encouraged to provide the evidence on the strength and direction of relationship between the ranking for all pairs of measures. We have found that the rank correlation is positive and statistically significant for almost all pairwise correlations. The lowest correlation has been found between pairs of wage and equal weights objective measure while the highest correlation has been found between subjective job satisfaction and average preferences objective measure. As one important issue in creating a reasonable public policies is to identify those individuals who are not faring well or those who are faring very well, we have decided to identify the individuals at the bottom and top end of the distribution according to different measures of job quality. The results have shown that the overlap of the worst off when we use two measures is lowest for wage and equivalent wage measures while the largest overlap occurs when we use two objective measures. These results should not be surprising since the informational requirements between two objective measures are more alike than the informational requirements between wage and equivalent wage. On the other hand, the largest overlap of the best off individuals for a pair of measures occurs between average preferences and equal weights measures while the overlap is lowest between wage and subjective job satisfaction. We have evaluated the gender differential in job quality across countries by using wage and equivalent wage measures. The evidence we have found indicate that in majority of countries, the average quality of jobs is higher for men than for women if we use the wage measure. While even if we use the equivalent wage measure, the quality of jobs are higher among men in most countries, still we have observed that women are either holding jobs of the same quality or they are even faring better than men in some countries. In addition, we have computed the willingness-to-pay for each non-wage job characteristic and we have compared the gender differences in the willingness-to-pay across and within countries. We have found that across almost all countries, the willingness-to-pay is higher among men, which indicates that they are suffering more not reaching the best possible values of non-wage job characteristics. Finally, we have presented the decomposition the total willingness-to-pay on the contributions attributed to each non-wage job characteristic. We have found that across countries for both men and women, having a good career prospect is relatively the most important non-wage job characteristic while job security is relatively the least important non-wage job characteristic. ; Chapter 3 In recent decades there has been a growing number of studies that investigated the effects of personal and job characteristics on the subjective well-being on the job. Besides, the empirical findings reveal that workers who are paid on the piece rates exert more effort and earn more than those workers paid an hourly salary. Nevertheless, it is ambiguous what will be the effect of performance paying schemes, since the well-being on the job can increase in wage but it can decrease with higher level of excreted effort. Since the possible effect of performance paying jobs on the well-being of workers stay hidden, we have tackled the following issue in this chapter and we have provided the empirical evidence on these effects. This chapter contributes to the literature on the subjective well-being by providing the casual effects that the performance pay job schemes have on job satisfaction. We approximate the well-being on the job by the subjective job satisfaction reported by individuals. We have used the Korean Labour and Income Panel Survey which allowed us to distinguish between the workers who are paid by performance and those who are paid by fixed rate. In addition, we could exploit the information about the particular type of performance paying scheme that applies to the workers who are paid by performance. Since the personality traits are possibly correlated with the observed and unobserved explanatory variables, we have decided to estimate the job satisfaction regression using the fixed effects estimator. We have shown that workers in the performance pay job schemes have a higher subjective well-being on the job than workers who are using the non-performance pay job schemes. The following result holds true even after we have controlled for the level of earnings, attitudes toward risk and other personal and job related characteristics. This evidence is also confirmed for both men and women. When it comes down to the effect of wage on job satisfaction, we observe that the effect is not apparent as someone might expect. We have found that among all explanatory variables, health has the strongest effect on the well-being on the job. Finally, we have exploited the information on the type of performance pay schemes in order to analyse how different performance pay schemes affect job satisfaction. The results have shown that workers who are employed on the individual, group and company performance pay job schemes are more satisfied on their job than workers who are paid by the fixed amount. The later result remains even after we have controlled for the difference in earnings, personal and job characteristics.