The thesis consists of three essays on urban economies. The first essay investigates the relationship between proximity to larger markets and wage distribution within local labor markets. In this essay I derive a theoretical spatial skill demand equation that positively links skill premiums to market access. Using data from U.S. metropolitan areas, I provide evidence that while average wages are higher in metropolitan areas with higher market access, as suggested in the existing literature, the wage differential is unequally distributed across the metropolitan workers. That is, greater access to markets is linked to relatively weaker outcomes for those at the bottom of the wage distribution.<p> The second essay examines the extent of urban sprawl with respect to the volatility of local economies. Specifically, it investigates how uncertainty over future land rents explains changes in the extent of urban sprawl. To theoretically study this relationship, I develop a theoretical model that links sprawl to shocks to changes in land development rent, among other factors. The econometric analysis draws upon panel data from U.S. metropolitan areas over the 1980-2000 censuses. To measure urban sprawl, I construct a distinctive measure that better captures the distribution of population density within metropolitan areas. Using suitable proxy that accounts for uncertainty over future land rents, I provide robust evidence confirming the theoretical prediction. That is, metropolitan areas with higher levels of uncertainty have a lower level of sprawl.<p> Finally, the third essay uses theories from urban production economics to empirically investigate the relationship between the economic performance of U.S. metropolitan areas and their respective amounts of sprawl. Specifically, this essay provides a comprehensive empirical analysis on the impact of urban sprawl on labor productivity. The main finding suggests that higher levels of urban sprawl are negatively associated with average labor productivity. Interestingly, this negative association is even stronger in smaller metropolitan areas. Still, there is evidence that the significance of the negative impact of sprawl is not homogenous across major industries.

This thesis explores several aspects of the labour market in Serbia and Croatia during the process of transition from socialism to a market economy. First, it examines how women??s position in the labour market has changed in Serbia. Using five annual Labour Force Surveys (2001-2005), I find that the gender wage gap is still very low in Serbia, and is even decreasing during this period. However, decompositions that apply the Oaxaca (1974) methodology reveal that the unexplained component of the gap is very large, and is increasing. Likewise, quantile decompositions suggest that while the raw gap is falling at each of the quantiles analysed, the unexplained component is increasing at most quantiles at the same time. Thus, the relatively small gap in earnings could be masking considerable discrimination in the labour market. Second, changes in men??s wage inequality in Serbia in the period from 2001 to 2005 are analysed using five annual Labour Force Surveys. Changes in the distribution of earnings are examined using the Lemieux (2002) decomposition methodology. I find that the change in wage inequality is mostly driven by changes in wage premiums, while the effect of changes in the composition of the labour force is very small. Isolating the effect of the emerging private sector reveals that changes in the private sector size and wage premium account for an average 25 percent of the changes in inequality during this period. Third, the effect that the recent war in Croatia (1991-1995) had on the educational and employment trajectories of the 1971 birth cohort of men is investigated. This birth cohort was most affected by the armed forces draft. I treat the occurrence of the war as a natural experiment and use data from the Croatian and Slovenian Labour Force Surveys. Applying the difference-in-difference framework and comparing this cohort to adjacent cohorts, women, and to respective cohorts in Slovenia, a neighbouring country that did not experience war, I find that the war has had a negative effect on educational outcomes and a small positive effect on the employment and earnings outcomes of this cohort of men.

This thesis explores several aspects of the labour market in Serbia and Croatia during the process of transition from socialism to a market economy. First, it examines how women??s position in the labour market has changed in Serbia. Using five annual Labour Force Surveys (2001-2005), I find that the gender wage gap is still very low in Serbia, and is even decreasing during this period. However, decompositions that apply the Oaxaca (1974) methodology reveal that the unexplained component of the gap is very large, and is increasing. Likewise, quantile decompositions suggest that while the raw gap is falling at each of the quantiles analysed, the unexplained component is increasing at most quantiles at the same time. Thus, the relatively small gap in earnings could be masking considerable discrimination in the labour market. Second, changes in men??s wage inequality in Serbia in the period from 2001 to 2005 are analysed using five annual Labour Force Surveys. Changes in the distribution of earnings are examined using the Lemieux (2002) decomposition methodology. I find that the change in wage inequality is mostly driven by changes in wage premiums, while the effect of changes in the composition of the labour force is very small. Isolating the effect of the emerging private sector reveals that changes in the private sector size and wage premium account for an average 25 percent of the changes in inequality during this period. Third, the effect that the recent war in Croatia (1991-1995) had on the educational and employment trajectories of the 1971 birth cohort of men is investigated. This birth cohort was most affected by the armed forces draft. I treat the occurrence of the war as a natural experiment and use data from the Croatian and Slovenian Labour Force Surveys. Applying the difference-in-difference framework and comparing this cohort to adjacent cohorts, women, and to respective cohorts in Slovenia, a neighbouring country that did not experience war, I find that the war has had a negative effect on educational outcomes and a small positive effect on the employment and earnings outcomes of this cohort of men.

Technological progress, human capital, and tax policies play an important role in growth. Recent models of endogenous growth based on technological progress predict that high technological progress and growth are associated with a high relative supply of skilled workers who earn constant or relatively low wages. Chapter 1 of this dissertation reviews recent models of endogenous growth. The 1980s, however, are associated with high technological progress, high relative supply and increasing relative wages of skilled workers. Chapter 2 of this dissertation shows that, unlike most recent endogenous growth models, high rates of technological change can be accompanied by a high relative supply and a high relative wage for skilled workers. This chapter looks at the relative wage of educated to uneducated individuals within the same generation in an overlapping generations model. Individuals live for two periods and decide whether to invest in education in the first period of their lives. As more individuals invest in education, the wage of unskilled workers increases, increasing the opportunity cost of education. At the equilibrium, to make the individuals who invest in education indifferent between education and work, the intra generational relative wage of educated individuals must increase Chapter 3 studies the local stability of the relative wage model. It shows that the unique equilibrium can be a sink, source, or saddle point. The numerical examples study the effects of an increase in the productivity of education on the entire trajectory of investment in education. Chapter 4 looks at the effects of different types of taxes in an economy in which the allocation of resources is inefficient. It shows that different types of taxes affect the long run growth rate differently. In our setting, taxing income from human capital employed in final good production allocates more human capital to R&D, and increases the growth rate of the economy. However, this is a very selective tax, and the conclusion depends on the production function. / Ph. D.

京都大学 / 0048 / 新制・課程博士 / 博士(経済学) / 甲第20147号 / 経博第545号 / 新制||経||280(附属図書館) / 京都大学大学院経済学研究科経済学専攻 / (主査)准教授 遊喜 一洋, 教授 劉 徳強, 教授 神事 直人 / 学位規則第4条第1項該当 / Doctor of Economics / Kyoto University / DFAM

This thesis explores how and why gender wage inequality persisted in Australia during the period 1990 to 2003. A review of the wage data indicated that Australian inequality persisted during this period and even increased in Western Australia. An analysis of the literature and published tribunal decisions showed that the focus of action relating to gender wage equality was on the formal industrial relations system, through legislation and tribunal decisions. As the literature also indicated that these measures had failed to assist many female-dominated occupations, it was felt useful to examine the views of key stakeholders in the formal industrial relations system. Given the mainly quantitative and aggregate data analysis of previous empirical literature and the theoretical literature’s growing focus on less quantifiable social factors, it was felt that qualitative research would add to the analysis of gender wage determination. This thesis uses in-depth interviews with stakeholders in the federal and Western Australian industrial relations systems to examine the period 1990 to 2003. Their perceptions and observations are triangulated with published wage data from selected occupations. / Interviewees felt that the persistence of gender wage inequality was a complex artifact. An artifact of economic, industrial relations, social and cultural factors, combined with the biological and psychological attributes of all involved in decisions before and in the workplace. While much of the gender wage gap literature has been focused on the economic, industrial relations, or legal aspects of wage inequality, this thesis uniquely demonstrates why and how social and cultural influences also act to create persistent gender wage inequality. As much as action is taken in the legal and political arena to create equality, the players in the industrial relations system, consciously or not, both male and female, say that they have helped to (re)construct the gender wage gap. The implications are that in order to achieve gender wage equality, action cannot be left only to legislation and activity in the courts, but is also needed at the social and cultural level. Such action could be taken within the enterprise as well as in the school, community and home.

Doctor of Philosophy / Department of Economics / William F. Blankenau / This dissertation consists of three essays focusing on wage inequality and education policy. Essay 1 considers growth in the variance of wages. Prior work has documented that the college premium plays a major role in explaining wage variance growth. This essay examines the extent to which this role can be attributed to an increase in the dispersion of occupation-specific returns to post-secondary education. Using the variance components approach and CPS data between 1979-1981 and 2003-2005, the essay shows that the variation in the college premium across occupations has increased over time, and this variation expansion explains about five percent of the growth in wage variance across the two periods. By dividing the sample workforce into professional and nonprofessional groups, the results suggest that the increased variation in the return to post-secondary education particularly caused the wage gap between the professional and non-professional workers to increase. Essay 2 applies quantile regression methodology to the study of the determinants of the wage distribution among natives and immigrants in the U.S., using PUMS from 1990 and 2000, and ACS from 2006. Among other findings, the immigrant/native wage gap is concentrated at the lower end to the median of the wage distribution, and the primary source of the wage gap is the relative lack of labor market skills among immigrants. A cross-time comparison shows that the recent immigrant/native wage gap after controlling for skill variables first decreased from 1990 to 2000 and then expanded from 2000 to 2006. The growth is concentrated at the two ends of the wage distribution, and the reason for growth is that the recent immigrants in 2006 are younger and thus have less market experience than their counterparts of 1990. Essay 3 is coauthored with Dr. Blankenau. We analyze the impact of changes in college admission standards on the skilled labor distribution, skilled firm distribution, and the match of skilled labor with skilled firms. We propose a model of schooling with heterogeneous labor and firms, in which firms’ decisions in creating skilled jobs are conditioned on the supply of skilled labor. The model shows that lowering standards without providing incentives to acquire skills does not necessarily motivate accumulation of human capital or expansion of skilled industry. Lower standards tend to create a mismatch of educated labor with unskilled positions. In some specifications, lower standards can lower firms’ willingness to create skilled positions, leaving more skilled workers underemployed.

This thesis discusses the changes and corresponding causes of the wage distribution in China from the late 1990s to the 2000s. According to various data sources, real wage inequality in China has been increasing over time. People have become increasingly concerned about such a phenomenon, which can potentially cause economic instability and further social unrest. From the analysis of household survey data, a significant part of the the increase in wage dispersion in China can be attributed to changes in the institutional changes. Having gone through the institutional reform of state-owned enterprises in the late 1990s, many Chinese firms have become more privatized and smaller in size. That is to say, the Chinese labour market becomes less affected by the government intervention (through public enterprises). Changes in the supply side of the labour market have also been examined. The increase in the number of university graduates slows down the growing wage dispersion. A comparison between the household survey data and the industrial enterprises data tells a slightly different story about Chinas wage structure. As the firm-level data omits within-firm wage inequalities and excludes data of primary sectors, the service sectors, and the small businesses, a decrease in the logarithm of the wage variation has been found. The inconsistency between the changes of real wage dispersion and the dispersion of log wages has been discussed in depth in the thesis. Nonetheless, since China set the new minimum wage in 2004, the wage distribution in the countrys industrial sector has been reshaped, which is not obviously shown in the household data. The impact of increasing the national minimum wage has been evaluated under a set of relatively conservative assumptions. Further analysis has been conducted to quantify the effect of trade liberalization on wage dispersion. It turns out that starting to export on the part of the firms has a significant positive effect on firm-level wages and employments, but the impact of an increasing export exposure remains debatable.

In recent decades, there have been increased investments in automation-related technology across several European industries. Parallel to this, industry-level labour share declines and differences in wage growth between demographic groups can be observed. This thesis examines the relationship between the increased technological investments and labour market outcomes of different demographic groups to uncover how increased exposure to automation technology has contributed to wage inequality between 2002-2018. The empirical strategy follows Acemoglu &amp; Restrepo (2022) and links the wage change of demographic groups across ten European countries to their expected task displacement, quantified through their specialization in routine tasks and exposure to industries with labour share declines, both in a reduced form and instrumental variable specification. The main empirical findings suggest that there has been no significant effect of increased exposure to automation, specifically computers and software, on relative wages between groups in Europe between 2002-2018, contrasting previous results from the U.S.

(Un-)gleichheit und Gerechtigkeit waren schon immer Kernbestandteil eines jeden politischen Konzeptes, welches das Wohl der Menschen in den Mittelpunkt rückte. Die Idee der Gleichheit ist so alt wie die der Demokratie selbst und neben instrumentellen Gründen, aus denen man (Un-)gleichheit schätzen oder ablehnen kann, ist die Existenz einer inhärenten Abneigung von Ungleichheit inzwischen ein allgemein akzeptiertes Faktum innerhalb der volkswirtschaftlichen Disziplin. Diese Dissertation trägt zur daraus folgenden Debatte um die Ursachen, Auswirkungen, und die Messung von Ungleichheit bei. Die Arbeit besteht aus drei Aufsätzen, welche empirische Untersuchungen in einem breit definierten Spektrum von wirtschaftlicher Ungleichheit – welches sowohl Löhne aus nicht-selbstständiger Arbeit, als auch breiter gefasste Einkommenskonzepte umfasst – enthalten. Was sind die Faktoren, welche zu einem Anstieg oder einer Verringerung von Ungleichheit innerhalb eines Landes führen und welche Auswirkungen hat diese höhere oder geringere Ungleichheit für dessen langfristige wirtschaftliche Entwicklung? Es wäre anmaßend zu behaupten, allgemeingültige Antworten auf diese wichtigen Fragen innerhalb der folgenden drei Aufsätze geben zu können. Diese Dissertation versucht vielmehr, Erklärungen für bestimmte Teilaspekte zu liefern. In Aufsatz 1 beschäftige ich mich mit dem Wirtschaftswachstum, eine der wohl wichtigsten Konsequenzen von Ungleichheit was die langfristige Wirkung auf das materielle Wohlbefinden eines Landes angeht. Im Zuge dessen befasse ich mich eingehend mit der noch immer ungelösten Kontroverse darüber, ob Ungleichheit zu mehr Wachstum beiträgt. Im Folgenden vertiefe ich dann die Diskussion über die Messung von Ungleichheit in Aufsatz 2 und führe damit eine Debatte fort, auf die bereits Essay 1 unvermeidlich stößt. Ich konzentriere mich dabei auf ein Maß für die Lohnverteilung in einer der Hauptsektoren der Wirtschaft, der verarbeitenden Industrie. Die Debatte über die richtige Messung von Ungleichheit ist nicht trivial: Weder die Ursachen, noch die Folgen von Ungleichheit können untersucht werden, ohne in die Details verschiedener zur Verfügung stehender Maße gehen zu müssen. Die Verfügbarkeit geeigneter Daten auf Länderebene und über die Zeit ist Voraussetzung für empirische Studien zu diesen Themen. Tatsächlich ist einer der Gründe dafür, dass es noch immer keinen Konsens darüber gibt, ob die Ungleichheit zwischen den Ländern in den letzten 50 Jahren gestiegen oder gefallen ist, die Vielfalt nicht nur der Maße, sondern auch der diesen zugrundeliegenden Konzepten, die in unterschiedlichen Studien verwendet werden. Die verschiedenen Maße reagieren ungleich stark auf bestimmte Bereiche der Einkommensverteilung und können daher zu unterschiedlichen Schlussfolgerungen darüber, in welchen Ländern es mehr und welche weniger Ungleichheit gibt, oder über Änderungen von Ungleichheit im Laufe der Zeit führen. Aus einem ähnlichen Grund ist die Literatur noch immer uneins über den Einfluss von Ungleichheit auf das Wirtschaftswachstum: Aufgrund der Nutzung unterschiedlicher Datenquellen sowie durch Mängel in den zugrundeliegenden Daten selbst ist die Vergleichbarkeit der verschiedenen Studien stark eingeschränkt. Unberührt davon, dass die Debatte um den Einfluss von Ungleichheit auf das Wirtschaftswachstum ungelöst bleibt, ist es wichtig zu wissen, welche Faktoren Ungleichheit verursachen oder verschlimmern (und welche dies nicht tun). Aufsatz 3 untersucht einen der in der öffentlichen Debatte häufiger zitierten Gründe für den Anstieg der Ungleichheit, den Handel. Ich untersuche empirisch die potentiellen Faktoren, die mit Handel einhergehen und zu steigender Ungleichheit führen können. Ich beschränke mich hierbei auf den Bestandteil von Ungleichheit, der laut Handelstheorien unmittelbar relevant für die Auswirkungen von Handel ist: die Verteilung der Löhne. Dabei verwende ich den in Aufsatz 2 behandelten Index der Lohnungleichheit in der verarbeitenden Industrie in einem der wenigen Kontexte, in denen dieser, wie ich in Aufsatz 2 argumentiere, angemessen ist.