View in EDS HTML Full Text PDF Full Text

Why Doesn't Expanding Higher Education Decrease Wage Inequality?

Saved in:
Bibliographic Details
Title: Why Doesn't Expanding Higher Education Decrease Wage Inequality?
Language: English
Authors: Hidefumi Kasuga, Yuichi Morita
Source: Education Economics. 2026 34(1):130-142.
Availability: Routledge. Available from: Taylor & Francis, Ltd. 530 Walnut Street Suite 850, Philadelphia, PA 19106. Tel: 800-354-1420; Tel: 215-625-8900; Fax: 215-207-0050; Web site: http://www.tandf.co.uk/journals
Peer Reviewed: Y
Page Count: 13
Publication Date: 2026
Document Type: Journal Articles
Reports - Descriptive
Education Level: Higher Education
Postsecondary Education
Descriptors: Higher Education, Outcomes of Education, Wages, Skilled Workers, Job Skills, Merit Pay, Access to Education, Cost Effectiveness, Developing Nations, Foreign Countries, College Graduates
DOI: 10.1080/09645292.2025.2477469
ISSN: 0964-5292
1469-5782
Abstract: The supply of college-educated workers has increased in many countries, but wage inequality has not necessarily decreased. Although skill-biased technical change can explain this phenomenon, we often observe that skill premiums increase in developing countries, where skill-biased technical change is less likely. We show that skill premiums can increase as the supply of skilled workers increases if merit-based pay is introduced. We find that from 1970 to 2010, in many countries, the Gini coefficient and skill premium increased as higher education expanded. The result suggests that meritocracy expands higher education but exacerbates wage inequality.
Abstractor: As Provided
Entry Date: 2026
Accession Number: EJ1500863
Database: ERIC
Full text is not displayed to guests.
FullText Links:
  – Type: pdflink
    Url: https://content.ebscohost.com/cds/retrieve?content=AQICAHj0k_4E0hTGH8RJwT4gCJyBsGNe_WN95AvKlDbXJGqwxwF07PFlT4_7CAzNBE-gV64yAAAA4zCB4AYJKoZIhvcNAQcGoIHSMIHPAgEAMIHJBgkqhkiG9w0BBwEwHgYJYIZIAWUDBAEuMBEEDK2UQQnIb_KHG4FgMgIBEICBm1TOvsnRzYlyq50oYEMWQHqmDlGSFdHjXbVFzlRnnoH5_UL4LljwqsB_OmAeT1JHDyE4CVMK8Zh8P_UThf2eMSyZUNjTumwVCkSXF4YyOz3oJivWPnruJKIKXldEoDM_-fUt4s-Yh6a6zlsnFP_MPiomsp_ueMzpSJ5owFC0mud_V4y_P5h9asdj5_CGSkqDJyo6Qr4RG7h9wUvV
Text:
  Availability: 1
  Value: <anid>AN0191012013;ede01feb.26;2026Jan23.07:23;v2.2.500</anid> <title id="AN0191012013-1">Why doesn't expanding higher education decrease wage inequality? </title> <p>The supply of college-educated workers has increased in many countries, but wage inequality has not necessarily decreased. Although skill-biased technical change can explain this phenomenon, we often observe that skill premiums increase in developing countries, where skill-biased technical change is less likely. We show that skill premiums can increase as the supply of skilled workers increases if merit-based pay is introduced. We find that from 1970 to 2010, in many countries, the Gini coefficient and skill premium increased as higher education expanded. The result suggests that meritocracy expands higher education but exacerbates wage inequality.</p> <p>Keywords: Higher education; meritocracy; wage inequality; skill premium</p> <hd id="AN0191012013-2">1. Introduction</hd> <p>More and more people are receiving tertiary education worldwide (Lee and Lee [<reflink idref="bib24" id="ref1">24</reflink>]; Morrisson and Murtin [<reflink idref="bib27" id="ref2">27</reflink>])[<reflink idref="bib1" id="ref3">1</reflink>]. This trend may contribute to reducing education inequality. Lower education inequality is expected to lower income inequality (De Gregorio and Lee [<reflink idref="bib11" id="ref4">11</reflink>]; Lee and Lee [<reflink idref="bib24" id="ref5">24</reflink>]; Park [<reflink idref="bib29" id="ref6">29</reflink>]). However, some studies show that inequality in wage income does not necessarily decrease when education inequality decreases (Baye, Wirba, and Tingum [<reflink idref="bib5" id="ref7">5</reflink>]; Castelló-Climent and Doménech [<reflink idref="bib8" id="ref8">8</reflink>]; Földvári and van Leeuwen [<reflink idref="bib13" id="ref9">13</reflink>]).</p> <p>Empirical evidence from developed countries shows that while the relative supply of college-educated workers has increased for decades, the wage gap between college graduates and high-school graduates has also risen or remained unchanged (Autor, Katz, and Krueger [<reflink idref="bib4" id="ref10">4</reflink>]; Blundell, Green, and Jin [<reflink idref="bib7" id="ref11">7</reflink>]). As Acemoglu ([<reflink idref="bib1" id="ref12">1</reflink>], [<reflink idref="bib2" id="ref13">2</reflink>]) shows, skill-biased technical change (SBTC), which increases the demand for skilled workers, can explain this puzzling phenomenon. While the term 'wage inequality' is often used in the literature, most authors focus on the skill premium (wage gap between skilled and unskilled workers). In this study, we distinguish between skill premium and wage inequality. The former implies the wage gap between skilled and unskilled workers, and the latter is measured by inequality indicators such as the Gini coefficient. The skill premium and Gini coefficient can move differently during education expansion because the Gini coefficient reflects both the wage gap between education groups and population size of each group. For example, a high skill premium can be consistent with a low inequality if there are few skilled (or unskilled) workers. Despite the recent interest in widening inequalities, little research has been conducted on the relationship between wage inequality (rather than the skill premium) and expanding higher education. Most studies focus on the decreasing share of labor income and rise of the super-rich[<reflink idref="bib2" id="ref14">2</reflink>]. Hence, little is known about why wage inequality changes and how expanding higher education and wage inequality are related.</p> <p>The literature from decades ago shows that the Gini coefficient increases at the initial stage of education expansion and declines subsequently, as in the inverted-U curve in Kuznets ([<reflink idref="bib22" id="ref15">22</reflink>]) (Anand and Kanbur [<reflink idref="bib3" id="ref16">3</reflink>]; Knight [<reflink idref="bib19" id="ref17">19</reflink>]; Knight and Sabot [<reflink idref="bib20" id="ref18">20</reflink>]; Robinson [<reflink idref="bib31" id="ref19">31</reflink>]). In these studies, the inverted-U relationship between education expansion and wage inequality is derived from dual economy models in which a population shifts from the traditional sector to the modern sector, which employs skilled labor; wage inequality increases initially because of a large skill premium, and as the supply of skilled workers increases, the pace of increase in wage inequality slows down. However, there are limitations to this explanation. First, these studies consider an exogenous increase in the share of skilled workers and do not explain how and why it increases. Second, they assume that the skill premium is constant or it declines while the supply of skilled workers increases. As discussed above, the skill premium has recently increased despite the increasing supply of skilled workers. Then, the inverted-U relationship will unlikely arise.</p> <p>Figure 1 shows how each country's skill premium (vertical axis) and the share of college graduates (horizontal axis) changed during 1970–2010[<reflink idref="bib3" id="ref20">3</reflink>]. The figure shows that the skill premium increases in half of the sample countries. The observation is not necessarily consistent with the SBTC explanation (which implies that technical change is biased to abundant factors) because most countries where the skill premium increased are developing countries. They do not have a large skilled workforce, which can accelerate SBTC, and hence, SBTC is less likely there. Thus, the trend in skill premium is not fully consistent with the presumption of the SBTC explanation, as well as with that of the model where the inverted-U curve arises. Hence, how and why wage inequality changes is an open question.</p> <p>Graph: Figure 1. Expanding education and changes in skill premiums: 1970–2010.</p> <p>To answer this question, this study uses a model of individuals' education choice and focuses on factors other than a technical change. Our model considers the costs and benefits of receiving higher education. While a lower cost and larger benefit increase the supply of educated workers, they can affect wage inequality differently. When tertiary enrollment increases because of a lower cost of education, more people improve their skills, and wage inequality decreases. However, when a larger benefit increases tertiary enrollment, wage inequality can increase; if improving skills is more rewarding, even a small skill gap leads to a large wage gap and can increase wage inequality. The key is how skills are rewarded. Meritocratic environments can increase tertiary enrollment and income inequality. Our model shows that 'meritocracy' explains why wage inequality increases as the supply of skilled workers increases[<reflink idref="bib4" id="ref21">4</reflink>]. In particular, the model shows two patterns of education expansion: in the first one, the Gini coefficient first increases and subsequently declines as in the inverted-U curve, whereas, in the other one, the Gini coefficient and skill premium increase persistently. In this study, we explain the relationship between the skill premium, education expansion, and wage inequality by focusing on reward systems rather than on technical change.</p> <p>The paper proceeds as follows. Section 2 explains how the cost of education, reward systems, and longevity affect tertiary enrollment in a model where individuals differ in innate ability. Our numerical examples derive two patterns of education expansion. Section 3 classifies each country's education expansion using data from 1970 to 2010. Section 4 concludes the paper.</p> <hd id="AN0191012013-3">2. The model</hd> <p></p> <hd id="AN0191012013-4">2.1. Education choice and the skill premium</hd> <p>We consider an overlapping generations model in a small open economy. Individuals live for three periods at most; all live through their childhood and adulthood, but some die before old age. We introduce longevity in the model because it determines the payback period of education investment and positively affects the benefit of education (Ben-Porath [<reflink idref="bib6" id="ref22">6</reflink>]). In childhood, individuals cannot make decisions. In adulthood, they supply labor, consume, and raise children. Adults study at college if they choose to become skilled workers[<reflink idref="bib5" id="ref23">5</reflink>]. Without college degrees, adults become unskilled workers. Some adults survive into old age; in their old age, they retire and only consume.</p> <p>The expected utility for an adult individual <emph>i</emph> in period <emph>t</emph> depends on consumption</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msubsup><mi>c</mi><mrow><mi>t</mi></mrow><mrow><mi>i</mi></mrow></msubsup></math> </ephtml> in adulthood and</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msubsup><mi>d</mi><mrow><mi>t</mi><mo>+</mo><mn>1</mn></mrow><mrow><mi>i</mi></mrow></msubsup></math> </ephtml> in old age as follows:</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msubsup><mi>U</mi><mrow><mi>t</mi></mrow><mrow><mi>i</mi></mrow></msubsup><mo>=</mo><mi>ln</mi><mo>⁡</mo><msubsup><mi>c</mi><mrow><mi>t</mi></mrow><mrow><mi>i</mi></mrow></msubsup><mo>+</mo><msup><mi>p</mi><mrow><mi>i</mi></mrow></msup><mtext fontfamily="times">β</mtext><mi>ln</mi><mo>⁡</mo><msubsup><mi>d</mi><mrow><mi>t</mi><mo>+</mo><mn>1</mn></mrow><mrow><mi>i</mi></mrow></msubsup><mo>,</mo></math> </ephtml> (<reflink idref="bib1" id="ref24">1</reflink>)</p> <p>where</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><mtext fontfamily="times">β</mtext><mo>></mo><mn>0</mn></math> </ephtml> is the discount factor, and</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msup><mi>p</mi><mrow><mi>i</mi></mrow></msup><mo>∈</mo><mo>(</mo><mn>0</mn><mo>,</mo><mn>1</mn><mo>)</mo></math> </ephtml> denotes the probability of adult survival. Each individual chooses whether to receive higher education and consumption levels (</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msubsup><mi>c</mi><mrow><mi>t</mi></mrow><mrow><mi>i</mi></mrow></msubsup></math> </ephtml> and</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msubsup><mi>d</mi><mrow><mi>t</mi><mo>+</mo><mn>1</mn></mrow><mrow><mi>i</mi></mrow></msubsup></math> </ephtml> ) at the beginning of adulthood. If they receive higher education, they supply skilled labor after graduation; otherwise, they supply unskilled labor. We assume that each adult has one child (i.e. population size is constant across generations).</p> <p>We denote the cost of education by a fraction of individuals' time endowment</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><mi>τ</mi><mo>∈</mo><mo>(</mo><mn>0</mn><mo>,</mo><mn>1</mn><mo>)</mo></math> </ephtml> . The benefit of education depends on wages, and acquired skills determine the wage rate of educated workers. We assume that acquired skills depend on their innate ability levels, which vary among individuals. Accordingly, the wage income is not identical for skilled workers; skilled workers with higher abilities earn more. Meanwhile, all unskilled workers earn the same wage, that is, wage income of unskilled workers does not depend on their innate ability levels. To simplify the notation, we use the superscript <emph>i</emph> for individuals: <emph>i</emph> = <emph>s</emph> for skilled workers and <emph>i</emph> = <emph>u</emph> for unskilled workers. The budget constraint of an adult in period <emph>t</emph> is:</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msubsup><mi>c</mi><mrow><mi>t</mi></mrow><mrow><mi>i</mi></mrow></msubsup><mo>=</mo><msubsup><mi>w</mi><mrow><mi>t</mi></mrow><mrow><mi>i</mi></mrow></msubsup><mo>(</mo><mn>1</mn><mo>−</mo><msup><mi>τ</mi><mrow><mi>i</mi></mrow></msup><mo>)</mo><mo>−</mo><msubsup><mi>s</mi><mrow><mi>t</mi></mrow><mrow><mi>i</mi></mrow></msubsup><mo>,</mo></math> </ephtml> (<reflink idref="bib2" id="ref25">2</reflink>)</p> <p>where</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msubsup><mi>w</mi><mrow><mi>t</mi></mrow><mrow><mi>i</mi></mrow></msubsup></math> </ephtml> and</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msubsup><mi>s</mi><mrow><mi>t</mi></mrow><mrow><mi>i</mi></mrow></msubsup></math> </ephtml> denote wage rates and savings, respectively. Let</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msup><mi>τ</mi><mrow><mi>i</mi></mrow></msup></math> </ephtml> denote the cost of education;</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msup><mi>τ</mi><mrow><mi>s</mi></mrow></msup><mo>=</mo><mi>τ</mi><mo>></mo><mn>0</mn></math> </ephtml> for skilled workers, and</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msup><mi>τ</mi><mrow><mi>u</mi></mrow></msup><mo>=</mo><mn>0</mn></math> </ephtml> for unskilled workers. To examine how adult survival and its gap affect education choice, we denote the probability of adult survival for skilled and unskilled workers by</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msup><mi>p</mi><mrow><mi>s</mi></mrow></msup></math> </ephtml> and</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msup><mi>p</mi><mrow><mi>u</mi></mrow></msup></math> </ephtml> , respectively. We assume that individuals have access to the world interest rate <emph>R</emph>. The budget constraint in <emph>t</emph> + 1 is:</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msubsup><mi>d</mi><mrow><mi>t</mi><mo>+</mo><mn>1</mn></mrow><mrow><mi>i</mi></mrow></msubsup><mo>=</mo><mo>(</mo><mn>1</mn><mo>+</mo><msubsup><mi>r</mi><mrow><mi>t</mi><mo>+</mo><mn>1</mn></mrow><mrow><mi>i</mi></mrow></msubsup><mo>)</mo><msubsup><mi>s</mi><mrow><mi>t</mi></mrow><mrow><mi>i</mi></mrow></msubsup><mo>,</mo></math> </ephtml> (<reflink idref="bib3" id="ref26">3</reflink>)</p> <p>where</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><mn>1</mn><mo>+</mo><msubsup><mi>r</mi><mrow><mi>t</mi><mo>+</mo><mn>1</mn></mrow><mrow><mi>i</mi></mrow></msubsup><mo>=</mo><mo>(</mo><mn>1</mn><mo>+</mo><mi>R</mi><mo>)</mo><mrow><mo>/</mo></mrow><msup><mi>p</mi><mrow><mi>i</mi></mrow></msup></math> </ephtml> . The rate of return</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msubsup><mi>r</mi><mrow><mi>t</mi><mo>+</mo><mn>1</mn></mrow><mrow><mi>i</mi></mrow></msubsup></math> </ephtml> depends on</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msup><mi>p</mi><mrow><mi>i</mi></mrow></msup></math> </ephtml> . This assumption implies that there are separate pension plans by occupation; the rate of return on each plan depends on the probability of adult survival.</p> <p>Output is produced by labor. Each adult endowed with a unit of time supplies labor hours. While unskilled workers supply one unit of labor, skilled workers supply</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><mn>1</mn><mo>−</mo><mi>τ</mi></math> </ephtml> unit. We assume that one unit of labor produces</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><mi>A</mi><mo>(</mo><mi>b</mi><mo>+</mo><mi>hα</mi><mo>)</mo></math> </ephtml> , where <emph>A</emph>>0 denotes the scale factor and</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><mo>(</mo><mi>b</mi><mo>+</mo><mi>hα</mi><mo>)</mo></math> </ephtml> represents how much skills contribute to labor productivity[<reflink idref="bib6" id="ref27">6</reflink>]. Parameters <emph>b</emph> and</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><mi>hα</mi></math> </ephtml> represent the contribution of the minimal skill endowment (i.e. the skill level of uneducated workers) and skills acquired in higher education, respectively. We assume that innate ability</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><mi>α</mi></math> </ephtml> varies across individuals on an interval between 0 and</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><mrow><mover><mi>α</mi><mo>ˆ</mo></mover></mrow></math> </ephtml> and that an educated worker whose innate ability is</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><mi>α</mi></math> </ephtml> acquires the skill level</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><mi>hα</mi></math> </ephtml> , where <emph>h</emph> measures how much tertiary education enhances skills. Accordingly, the wage rate of unskilled workers is <emph>Ab</emph>, and that of skilled workers is</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><mi>A</mi><mo>(</mo><mi>b</mi><mo>+</mo><mi>hα</mi><mo>)</mo></math> </ephtml> . In other words, workers without college degrees only earn a base pay (<emph>Ab</emph>) whereas educated workers earn additional rewards based on their skills (</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><mi>Ahα</mi></math> </ephtml> ). Thus, the reward system of the economy is described by parameters <emph>A</emph>, <emph>b</emph>, <emph>h</emph> and</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><mi>α</mi></math> </ephtml> . A small <emph>b</emph> implies a meritocratic environment. We denote the probability density function of</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><mi>α</mi></math> </ephtml> by</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><mi>f</mi><mo>(</mo><mi>α</mi><mo>)</mo></math> </ephtml> .</p> <p>Each adult individual maximizes the expected utility (<reflink idref="bib1" id="ref28">1</reflink>) subject to Equations (<reflink idref="bib2" id="ref29">2</reflink>) and (<reflink idref="bib3" id="ref30">3</reflink>). The first-order condition is</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msubsup><mi>s</mi><mrow><mi>t</mi></mrow><mrow><mi>i</mi></mrow></msubsup><mo>=</mo><msup><mi>p</mi><mrow><mi>i</mi></mrow></msup><mtext fontfamily="times">β</mtext><msubsup><mi>c</mi><mrow><mi>t</mi></mrow><mrow><mi>i</mi></mrow></msubsup></math> </ephtml> . From (<reflink idref="bib2" id="ref31">2</reflink>), we obtain</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msubsup><mi>c</mi><mrow><mi>t</mi></mrow><mrow><mi>i</mi></mrow></msubsup><mo>=</mo><mrow><mfrac><mrow><mn>1</mn><mo>−</mo><msup><mi>τ</mi><mrow><mi>i</mi></mrow></msup></mrow><mrow><mn>1</mn><mo>+</mo><msup><mi>p</mi><mrow><mi>i</mi></mrow></msup><mtext fontfamily="times">β</mtext></mrow></mfrac></mrow><msubsup><mi>w</mi><mrow><mi>t</mi></mrow><mrow><mi>i</mi></mrow></msubsup><mo>.</mo></math> </ephtml> (<reflink idref="bib4" id="ref32">4</reflink>)</p> <p>Equation (<reflink idref="bib4" id="ref33">4</reflink>) shows that the level of consumption in period <emph>t</emph> depends negatively on the cost of education and probability of survival. The negative effect of</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msup><mi>p</mi><mrow><mi>i</mi></mrow></msup></math> </ephtml> on</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msubsup><mi>c</mi><mrow><mi>t</mi></mrow><mrow><mi>i</mi></mrow></msubsup></math> </ephtml> implies that longevity encourages savings. From Equations (<reflink idref="bib1" id="ref34">1</reflink>)–(<reflink idref="bib4" id="ref35">4</reflink>), we obtain</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msubsup><mi>U</mi><mrow><mi>t</mi></mrow><mrow><mi>s</mi></mrow></msubsup><mo>=</mo><mo>(</mo><mn>1</mn><mo>+</mo><msup><mi>p</mi><mrow><mi>s</mi></mrow></msup><mtext fontfamily="times">β</mtext><mo>)</mo><mi>ln</mi><mo>⁡</mo><mrow><mfrac><mrow><mo>(</mo><mn>1</mn><mo>−</mo><mi>τ</mi><mo>)</mo><mi>A</mi><mo>(</mo><mi>b</mi><mo>+</mo><mi>hα</mi><mo>)</mo></mrow><mrow><mn>1</mn><mo>+</mo><msup><mi>p</mi><mrow><mi>s</mi></mrow></msup><mtext fontfamily="times">β</mtext></mrow></mfrac></mrow><mo>+</mo><msup><mi>p</mi><mrow><mi>s</mi></mrow></msup><mtext fontfamily="times">β</mtext><mi>ln</mi><mo>⁡</mo><mo>(</mo><mn>1</mn><mo>+</mo><mi>R</mi><mo>)</mo><mtext fontfamily="times">β</mtext><mo>,</mo></math> </ephtml> (<reflink idref="bib5" id="ref36">5</reflink>)</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msubsup><mi>U</mi><mrow><mi>t</mi></mrow><mrow><mi>u</mi></mrow></msubsup><mo>=</mo><mo>(</mo><mn>1</mn><mo>+</mo><msup><mi>p</mi><mrow><mi>u</mi></mrow></msup><mtext fontfamily="times">β</mtext><mo>)</mo><mi>ln</mi><mo>⁡</mo><mrow><mfrac><mrow><mi>Ab</mi></mrow><mrow><mn>1</mn><mo>+</mo><msup><mi>p</mi><mrow><mi>u</mi></mrow></msup><mtext fontfamily="times">β</mtext></mrow></mfrac></mrow><mo>+</mo><msup><mi>p</mi><mrow><mi>u</mi></mrow></msup><mtext fontfamily="times">β</mtext><mi>ln</mi><mo>⁡</mo><mo>(</mo><mn>1</mn><mo>+</mo><mi>R</mi><mo>)</mo><mtext fontfamily="times">β</mtext><mo>.</mo></math> </ephtml> (<reflink idref="bib6" id="ref37">6</reflink>)</p> <p>From Equations (<reflink idref="bib5" id="ref38">5</reflink>) and (<reflink idref="bib6" id="ref39">6</reflink>),</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><mi mathvariant="normal">∂</mi><msubsup><mi>U</mi><mrow><mi>t</mi></mrow><mrow><mi>i</mi></mrow></msubsup><mrow><mo>/</mo></mrow><mi mathvariant="normal">∂</mi><msup><mi>p</mi><mrow><mi>i</mi></mrow></msup><mo>=</mo><mtext fontfamily="times">β</mtext><mo>[</mo><mi>ln</mi><mo>⁡</mo><msubsup><mi>c</mi><mrow><mi>t</mi></mrow><mrow><mi>i</mi></mrow></msubsup><mo>−</mo><mn>1</mn><mo>+</mo><mi>ln</mi><mo>⁡</mo><mo>(</mo><mn>1</mn><mo>+</mo><mi>R</mi><mo>)</mo><mtext fontfamily="times">β</mtext><mo>]</mo></math> </ephtml> . Thus, longevity positively affects the expected utility if</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><mi>ln</mi><mo>⁡</mo><msubsup><mi>c</mi><mrow><mi>t</mi></mrow><mrow><mi>i</mi></mrow></msubsup><mo>(</mo><mn>1</mn><mo>+</mo><mi>R</mi><mo>)</mo><mtext fontfamily="times">β</mtext><mo>></mo><mn>1</mn></math> </ephtml> , which holds for a sufficiently large <emph>A</emph> (because</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msubsup><mi>c</mi><mrow><mi>t</mi></mrow><mrow><mi>i</mi></mrow></msubsup></math> </ephtml> depends on <emph>A</emph>). We assume</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><mi mathvariant="normal">∂</mi><msubsup><mi>U</mi><mrow><mi>t</mi></mrow><mrow><mi>i</mi></mrow></msubsup><mrow><mo>/</mo></mrow><mi mathvariant="normal">∂</mi><msup><mi>p</mi><mrow><mi>i</mi></mrow></msup><mo>></mo><mn>0</mn></math> </ephtml> . Adults choose whether to receive higher education by comparing</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msubsup><mi>U</mi><mrow><mi>t</mi></mrow><mrow><mi>s</mi></mrow></msubsup></math> </ephtml> and</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msubsup><mi>U</mi><mrow><mi>t</mi></mrow><mrow><mi>u</mi></mrow></msubsup></math> </ephtml> . If</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msubsup><mi>U</mi><mrow><mi>t</mi></mrow><mrow><mi>s</mi></mrow></msubsup><mo>≥</mo><msubsup><mi>U</mi><mrow><mi>t</mi></mrow><mrow><mi>u</mi></mrow></msubsup></math> </ephtml> , they receive education and become skilled workers. From Equations (<reflink idref="bib5" id="ref40">5</reflink>) and (<reflink idref="bib6" id="ref41">6</reflink>),</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msubsup><mi>U</mi><mrow><mi>t</mi></mrow><mrow><mi>s</mi></mrow></msubsup></math> </ephtml> is increasing in</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><mi>α</mi></math> </ephtml> , whereas</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msubsup><mi>U</mi><mrow><mi>t</mi></mrow><mrow><mi>u</mi></mrow></msubsup></math> </ephtml> is independent of</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><mi>α</mi></math> </ephtml> . Let</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msup><mi>α</mi><mrow><mo>∗</mo></mrow></msup></math> </ephtml> be the level of</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><mi>α</mi></math> </ephtml> such that</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msubsup><mi>U</mi><mrow><mi>t</mi></mrow><mrow><mi>u</mi></mrow></msubsup><mo>=</mo><msubsup><mi>U</mi><mrow><mi>t</mi></mrow><mrow><mi>s</mi></mrow></msubsup></math> </ephtml> . Then, adults with innate ability</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><mi>α</mi></math> </ephtml> equal to and larger than</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msup><mi>α</mi><mrow><mo>∗</mo></mrow></msup></math> </ephtml> receive higher education, and the proportion of skilled workers in the adult population is given by</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><mi>ϕ</mi><mo>=</mo><mn>1</mn><mo>−</mo><msubsup><mo>∫</mo><mrow><mn>0</mn></mrow><mrow><msup><mi>α</mi><mrow><mo>∗</mo></mrow></msup></mrow></msubsup><mi>f</mi><mo>(</mo><mi>α</mi><mo>)</mo><mspace width="thinmathspace" /><mrow><mi mathvariant="normal">d</mi></mrow><mi>α</mi><mo>=</mo><msubsup><mo>∫</mo><mrow><msup><mi>α</mi><mrow><mo>∗</mo></mrow></msup></mrow><mrow><mrow><mover><mi>α</mi><mo>ˆ</mo></mover></mrow></mrow></msubsup><mi>f</mi><mo>(</mo><mi>α</mi><mo>)</mo><mspace width="thinmathspace" /><mrow><mi mathvariant="normal">d</mi></mrow><mi>α</mi><mo>.</mo></math> </ephtml> (<reflink idref="bib7" id="ref42">7</reflink>)</p> <p>We can interpret <emph>ϕ</emph> as the tertiary enrollment rate.</p> <p>Using the average wage for skilled workers</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msup><mrow><mover><mi>w</mi><mo>¯</mo></mover></mrow><mrow><mi>s</mi></mrow></msup></math> </ephtml> , we define the wage gap between skilled and unskilled workers as</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><mrow><mfrac><msup><mrow><mover><mi>w</mi><mo>¯</mo></mover></mrow><mrow><mi>s</mi></mrow></msup><msup><mi>w</mi><mrow><mi>u</mi></mrow></msup></mfrac></mrow><mo>=</mo><mrow><mfrac><mrow><mi>A</mi><mo>(</mo><mi>b</mi><mo>+</mo><mi>H</mi><mo>)</mo></mrow><mrow><mi>Ab</mi></mrow></mfrac></mrow><mo>=</mo><mn>1</mn><mo>+</mo><mrow><mfrac><mi>H</mi><mi>b</mi></mfrac></mrow><mo>,</mo><mspace width="1em" /><mrow><mi mathvariant="normal">where</mi></mrow><mspace width="thinmathspace" /><mi>H</mi><mo>=</mo><msubsup><mo>∫</mo><mrow><msup><mi>α</mi><mrow><mo>∗</mo></mrow></msup></mrow><mrow><mrow><mover><mi>α</mi><mo>ˆ</mo></mover></mrow></mrow></msubsup><mi>hαf</mi><mo>(</mo><mi>α</mi><mo>)</mo><mspace width="thinmathspace" /><mrow><mi mathvariant="normal">d</mi></mrow><mi>α</mi><mo>.</mo></math> </ephtml> (<reflink idref="bib8" id="ref43">8</reflink>)</p> <p>The wage gap (<reflink idref="bib8" id="ref44">8</reflink>) measures the skill premium. Letting</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msubsup><mi>w</mi><mrow><mo movablelimits="true" form="prefix">max</mo></mrow><mrow><mi>s</mi></mrow></msubsup></math> </ephtml> and</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msubsup><mi>w</mi><mrow><mo movablelimits="true" form="prefix">min</mo></mrow><mrow><mi>s</mi></mrow></msubsup></math> </ephtml> denote the highest and lowest wages for skilled workers, respectively, we define the wage gap within skilled workers as</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><mrow><mfrac><msubsup><mi>w</mi><mrow><mo movablelimits="true" form="prefix">max</mo></mrow><mrow><mi>s</mi></mrow></msubsup><msubsup><mi>w</mi><mrow><mo movablelimits="true" form="prefix">min</mo></mrow><mrow><mi>s</mi></mrow></msubsup></mfrac></mrow><mo>=</mo><mrow><mfrac><mrow><mi>b</mi><mo>+</mo><mi>h</mi><mrow><mover><mi>α</mi><mo>ˆ</mo></mover></mrow></mrow><mrow><mi>b</mi><mo>+</mo><mi>h</mi><msup><mi>α</mi><mrow><mo>∗</mo></mrow></msup></mrow></mfrac></mrow><mo>.</mo></math> </ephtml> (<reflink idref="bib9" id="ref45">9</reflink>)</p> <p>The population size of skilled workers (<reflink idref="bib7" id="ref46">7</reflink>) and the wage gaps (<reflink idref="bib8" id="ref47">8</reflink>) and (<reflink idref="bib9" id="ref48">9</reflink>) affect the degree of inequality measured by the Gini coefficient. Next, we examine what factors affect (<reflink idref="bib7" id="ref49">7</reflink>)–(<reflink idref="bib9" id="ref50">9</reflink>).</p> <p>First, we consider a case with no survival gap (i.e.</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msup><mi>p</mi><mrow><mi>s</mi></mrow></msup><mo>=</mo><msup><mi>p</mi><mrow><mi>u</mi></mrow></msup></math> </ephtml> ) and examine the effects of the parameters. From Equations (<reflink idref="bib5" id="ref51">5</reflink>) and (<reflink idref="bib6" id="ref52">6</reflink>),</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><mi>h</mi><msup><mi>α</mi><mrow><mo>∗</mo></mrow></msup><mo>=</mo><mi>bτ</mi><mrow><mo>/</mo></mrow><mo>(</mo><mn>1</mn><mo>−</mo><mi>τ</mi><mo>)</mo></math> </ephtml> holds; then,</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msup><mi>α</mi><mrow><mo>∗</mo></mrow></msup></math> </ephtml> is independent of</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msup><mi>p</mi><mrow><mi>i</mi></mrow></msup></math> </ephtml> . An increase in <emph>τ</emph> leads to a larger</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msup><mi>α</mi><mrow><mo>∗</mo></mrow></msup></math> </ephtml> , and hence, decreases the tertiary enrollment <emph>ϕ</emph> in Equation (<reflink idref="bib7" id="ref53">7</reflink>), raises the skill premium in Equation (<reflink idref="bib8" id="ref54">8</reflink>), and narrows the wage gap within skilled workers in Equation (<reflink idref="bib9" id="ref55">9</reflink>). An increase in <emph>b</emph> also leads to a larger</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msup><mi>α</mi><mrow><mo>∗</mo></mrow></msup></math> </ephtml> and smaller <emph>ϕ</emph>. Thus, a rise in the base pay <emph>Ab</emph> decreases tertiary enrollment; however, it is not clear how the skill premium changes because both <emph>b</emph> and <emph>H</emph> increase. The wage gap within skilled workers narrows because</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msup><mi>α</mi><mrow><mo>∗</mo></mrow></msup></math> </ephtml> increases. As long as <emph>h</emph> is constant, individuals acquire skills in proportion to their innate abilities. By relaxing the assumption that <emph>h</emph> is constant, we can describe a more realistic setting where individuals with high innate abilities acquire more skills at a better university. Suppose that <emph>h</emph> is an increasing function of</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><mi>α</mi></math> </ephtml> . Then, individuals with higher abilities earn higher wages by entering higher-quality universities, but those with lower abilities acquire their skills only slightly by entering lower-tier universities (accordingly, their wages are not much different from those of unskilled workers). Even in this setting, a rise in <emph>τ</emph> increases</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msup><mi>α</mi><mrow><mo>∗</mo></mrow></msup></math> </ephtml> , <emph>H</emph>, and the skill premium. A rise in <emph>b</emph> also increases</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msup><mi>α</mi><mrow><mo>∗</mo></mrow></msup></math> </ephtml> . Thus, considering university quality does not alter the qualitative results of the model. Hence, we assume that <emph>h</emph> is constant and 1 for simplicity.</p> <p>When</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msup><mi>p</mi><mrow><mi>s</mi></mrow></msup><mo>≠</mo><msup><mi>p</mi><mrow><mi>u</mi></mrow></msup></math> </ephtml> , from Equations (<reflink idref="bib5" id="ref56">5</reflink>) and (<reflink idref="bib6" id="ref57">6</reflink>),</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msup><mi>α</mi><mrow><mo>∗</mo></mrow></msup></math> </ephtml> depends on</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msup><mi>p</mi><mrow><mi>s</mi></mrow></msup></math> </ephtml> and</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msup><mi>p</mi><mrow><mi>u</mi></mrow></msup></math> </ephtml> . We assume</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msup><mi>p</mi><mrow><mi>s</mi></mrow></msup><mo>></mo><msup><mi>p</mi><mrow><mi>u</mi></mrow></msup></math> </ephtml> [<reflink idref="bib7" id="ref58">7</reflink>]. To examine how longevity affects tertiary enrollment and the two wage gaps, we consider two cases where life expectancy increases:</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msup><mi>p</mi><mrow><mi>s</mi></mrow></msup></math> </ephtml> increases for a given</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msup><mi>p</mi><mrow><mi>u</mi></mrow></msup></math> </ephtml> (the longevity gap widens), and</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msup><mi>p</mi><mrow><mi>u</mi></mrow></msup></math> </ephtml> increases for a given</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msup><mi>p</mi><mrow><mi>s</mi></mrow></msup></math> </ephtml> (the longevity gap narrows). In the former case,</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msubsup><mi>U</mi><mrow><mi>t</mi></mrow><mrow><mi>s</mi></mrow></msubsup></math> </ephtml> increases for a given</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><mi>α</mi></math> </ephtml> by assumption; hence,</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msup><mi>α</mi><mrow><mo>∗</mo></mrow></msup></math> </ephtml> decreases and <emph>ϕ</emph> increases. Then, the skill premium decreases, and the wage gap within skilled workers increases. In the latter case,</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msubsup><mi>U</mi><mrow><mi>t</mi></mrow><mrow><mi>u</mi></mrow></msubsup></math> </ephtml> increases; hence,</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msup><mi>α</mi><mrow><mo>∗</mo></mrow></msup></math> </ephtml> increases and <emph>ϕ</emph> decreases. Then, contrary to the former case, the skill premium increases, and the wage gap within skilled workers decreases. These results show that the tertiary enrollment <emph>ϕ</emph> depends positively on the difference between</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msup><mi>p</mi><mrow><mi>s</mi></mrow></msup></math> </ephtml> and</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msup><mi>p</mi><mrow><mi>u</mi></mrow></msup></math> </ephtml> . While life expectancy (the average values of</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msup><mi>p</mi><mrow><mi>s</mi></mrow></msup></math> </ephtml> and</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msup><mi>p</mi><mrow><mi>u</mi></mrow></msup></math> </ephtml> ) increases in these two cases, only the former case increases the tertiary enrollment <emph>ϕ</emph> because a large longevity gap increases the benefit of education.</p> <p>The results above contribute to our understanding of the relationship between longevity and education investment. The potential positive relationship between longevity and education investment is known as the Ben-Porath effect (Ben-Porath [<reflink idref="bib6" id="ref59">6</reflink>]; Cohen and Leker [<reflink idref="bib9" id="ref60">9</reflink>]; de la Croix and Licandro [<reflink idref="bib12" id="ref61">12</reflink>]). However, empirical evidence does not necessarily support this effect (Cohen and Soto [<reflink idref="bib10" id="ref62">10</reflink>]; Hazan [<reflink idref="bib16" id="ref63">16</reflink>]; Hazan and Zoabi [<reflink idref="bib17" id="ref64">17</reflink>]). We show that the effect of longevity on education investment can be either positive or negative by focusing on an individual's decision to go to college. Our results suggest that education investment depends on the longevity gap rather than longevity per se.</p> <p>The relationship between higher education and the wage gaps can be summarized as follows. First, education expansion due to a decrease in the cost of education and an increase in the longevity gap leads to a decrease in the skill premium. Second, regardless of the cause, expanding education is associated with an increase in the wage gap within skilled workers. Thus, the model shows that the skill premium decreases as the supply of skilled workers increases. This relationship is the key to explaining the inverted-U curve in the literature as follows: initially, wage inequality increases as the supply of skilled workers increases, because the skill premium is large; thereafter, the pace of the increase in inequality slows down, and finally, inequality decreases, as long as the skill premium decreases. However, our results do not eliminate the possibility that the skill premium increases during education expansion. It is unclear how <emph>b</emph> affects the skill premium. If a decline in <emph>b</emph> increases the share of skilled workers and skill premium, the inverted-U relationship may not hold because a decrease in inequality relies on the decrease in skill premium. To examine this possibility, we specify the distribution of innate ability</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><mi>f</mi><mo>(</mo><mi>α</mi><mo>)</mo></math> </ephtml> in Section 2.2 and show how the education–inequality relationship depends on each parameter.</p> <hd id="AN0191012013-5">2.2. The skill premium and Gini coefficient</hd> <p>This subsection assumes that innate ability</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><mi>α</mi></math> </ephtml> is distributed uniformly to examine how parameters <emph>τ</emph>, <emph>b</emph>,</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msup><mi>p</mi><mrow><mi>s</mi></mrow></msup></math> </ephtml> , and</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msup><mi>p</mi><mrow><mi>u</mi></mrow></msup></math> </ephtml> affect <emph>ϕ</emph>, the skill premium, and Gini coefficient. We use the Gini coefficient to measure wage inequality because it is one of the most commonly used measures and has the scale invariance property[<reflink idref="bib8" id="ref65">8</reflink>]. We focus on the case where <emph>ϕ</emph> increases (i.e. education expansion due to a decrease in <emph>τ</emph>, <emph>b</emph>, and</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msup><mi>p</mi><mrow><mi>u</mi></mrow></msup></math> </ephtml> , and an increase in</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msup><mi>p</mi><mrow><mi>s</mi></mrow></msup></math> </ephtml> ). Table 1 shows how we change parameter values. For example, we set the initial value and minimum of <emph>τ</emph> to 0.2 and 0.02, respectively. We decrease the value of <emph>τ</emph> from 0.2 to 0.02 in small increments.</p> <p>Table 1. Parameters.</p> <p> <ephtml> <table><thead valign="bottom"><tr><td /><td>Initial value</td><td>Minimum</td><td>Maximum</td></tr></thead><tbody><tr><td><italic>τ</italic></td><td char=".">0.2</td><td char=".">0.02</td><td /></tr><tr><td><italic>b</italic></td><td>100</td><td>10</td><td /></tr><tr><td><p><graphic href="cede_a_2477469_ilm0087.gif" content-type="Graph" /><math xmlns="http://www.w3.org/1998/Math/MathML"><msup xmlns=""><mi>p</mi><mrow><mi>s</mi></mrow></msup></math></p></td><td char=".">0.7</td><td /><td char=".">0.75</td></tr><tr><td><p><graphic href="cede_a_2477469_ilm0088.gif" content-type="Graph" /><math xmlns="http://www.w3.org/1998/Math/MathML"><msup xmlns=""><mi>p</mi><mrow><mi>u</mi></mrow></msup></math></p></td><td char=".">0.7</td><td char=".">0.65</td><td /></tr><tr><td>fixed</td><td /><td /></tr><tr><td><italic>A</italic></td><td>10</td><td /></tr><tr><td><italic>β</italic></td><td char=".">1.3</td><td /></tr><tr><td><italic>h</italic></td><td>1</td><td /></tr><tr><td><italic>R</italic></td><td char=".">0.1</td><td /></tr><tr><td><p><graphic href="cede_a_2477469_ilm0089.gif" content-type="Graph" /><math xmlns="http://www.w3.org/1998/Math/MathML"><mi xmlns="">α</mi></math></p></td><td /><td>1</td><td>30</td></tr></tbody></table> </ephtml> </p> <p>Figure 2 shows the effects of parameter changes on the Gini coefficient (left panel), skill premium</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msup><mi>w</mi><mrow><mi>s</mi></mrow></msup><mrow><mo>/</mo></mrow><msup><mi>w</mi><mrow><mi>u</mi></mrow></msup></math> </ephtml> (center panel), and wage gap within skilled workers</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msubsup><mi>w</mi><mrow><mo movablelimits="true" form="prefix">max</mo></mrow><mrow><mi>s</mi></mrow></msubsup><mrow><mo>/</mo></mrow><msubsup><mi>w</mi><mrow><mo movablelimits="true" form="prefix">min</mo></mrow><mrow><mi>u</mi></mrow></msubsup></math> </ephtml> (right panel) in each row. The horizontal axis of each panel is <emph>ϕ</emph>. The first row shows the effect of a decrease in <emph>τ</emph>. The Gini coefficient first increases and subsequently declines as <emph>ϕ</emph> increases. The skill premium decreases as <emph>ϕ</emph> increases. An inverted-U relationship arises because of the downward-sloping skill premium. As the model suggests, the wage gap within skilled workers increases as <emph>ϕ</emph> increases. The second row shows the effect of a decrease in <emph>b</emph>. The Gini coefficient and skill premium slope upward. This is the case where the skill premium increases as the share of skilled workers increases. A decrease in <emph>b</emph> implies that skills are rewarded more highly (i.e. the share of the base pay decreases). Note that a decrease in <emph>b</emph> leads to an increase in the Gini coefficient, but this is not because of a decline in the earnings. Even if <emph>b</emph> decreases, wage income can increase when <emph>A</emph> increases. As the Gini coefficient is scale-invariant, an increase in <emph>A</emph> does not affect it. As long as <emph>b</emph> decreases, the Gini coefficient rises regardless of the wage level. Thus, a declining base pay increases the supply of skilled workers and Gini coefficient (adults go to college to avoid low-income jobs even if education improves their skills only slightly). The third and fourth rows show the effects of the longevity gap. An increase in</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msup><mi>p</mi><mrow><mi>s</mi></mrow></msup></math> </ephtml> and decrease in</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msup><mi>p</mi><mrow><mi>u</mi></mrow></msup></math> </ephtml> raise the share of skilled workers because both of them decrease</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msup><mi>α</mi><mrow><mo>∗</mo></mrow></msup></math> </ephtml> . Then, as in the first row, the Gini coefficient initially increases and subsequently declines because of the downward-sloping skill premium.</p> <p>Graph: Figure 2. Effects of parameter changes. (a) Decrease in education cost τ from 0.2 to 0.02. (b) Decrease in the base pay b from 100 to 10. (c) Increase in survival ps from 0.7 to 0.75 and (d) Decrease in survival pu from 0.7 to 0.65.</p> <p>The relationship between education expansion, the Gini coefficient, and skill premium can be summarized as follows. First, education expansion is associated with a rise in the Gini coefficient when the share of skilled workers is small. Second, the inverted-U relationship between education expansion and the Gini coefficient arises if the skill premium is downward-sloping. Third, the Gini coefficient persistently increases as the share of skilled workers increases if the skill premium is upward-sloping. The second and third results show that the relationship between education expansion and the Gini coefficient depends on the slope of the skill premium[<reflink idref="bib9" id="ref66">9</reflink>].</p> <p>We summarize the key results as follows.</p> <hd id="AN0191012013-6">Proposition 1</hd> <p>Education expansion can be consistent with a persistent rise in the skill premium.</p> <p>As shown in Figure 2, the relationship between education expansion and the Gini coefficient depends on how the skill premium changes. When the skill premium is downward-sloping, the Gini coefficient first rises (because of a large skill premium) and subsequently declines (after the skill premium becomes sufficiently small). Thus, we observe an inverted-U relationship. Conversely, when the skill premium is upward-sloping, the Gini coefficient increases persistently as the share of skilled workers increases[<reflink idref="bib10" id="ref67">10</reflink>]. This is observed only when the parameter <emph>b</emph> decreases. A small <emph>b</emph> represents a meritocratic environment, where rewards are based on ability and effort (Young [<reflink idref="bib33" id="ref68">33</reflink>]). In our model, ability is</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><mi>α</mi></math> </ephtml> , and effort refers to acquiring skills at college. The common view is that wage inequality decreases when more people acquire skills at college. However, a decrease in <emph>b</emph> implies that rewards for college graduates become more dependent on their skills</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><mi>α</mi></math> </ephtml> (i.e. strengthening meritocracy); the skill premium increases even if all workers become college graduates. Thus, strengthening meritocracy leads to a large wage difference among college graduates (degrees in science and engineering have higher earning potential than degrees in other fields). Our numerical examples show that because of the increasing wage difference among college graduates, wage inequality can increase even if everyone has a university degree[<reflink idref="bib11" id="ref69">11</reflink>].</p> <hd id="AN0191012013-7">3. Empirical relationship between the skill premium and Gini coefficient</hd> <p>Contrary to the common view, the previous section shows that the skill premium can increase persistently as the supply of skilled workers grows. The skill premium, which measures the value of skill, reflects the degree of meritocracy. In a more meritocratic pay system, workers would be compensated based on their skills and education. Kunovich and Slomczynski ([<reflink idref="bib21" id="ref70">21</reflink>]), based on the joint distribution of education and earnings for 14 countries, show that East Germany, the United States, and Czechoslovakia are highly meritocratic, while Russia is the least meritocratic (interestingly, some former communist countries are highly meritocratic in their results). In our model, there are two types of education expansion; only if strengthening meritocracy leads to education expansion, the skill premium and the Gini coefficient simultaneously increase (if the number of skilled workers increases for reasons other than strengthening meritocracy, the skill premium is downward-sloping, and the inverted-U curve arises). This section examines whether the empirical relationship between the skill premium and the Gini coefficient during education expansion is consistent with strengthening meritocracy.</p> <p>We use the same wage and education data as in Figure 1 but limit the sample to countries where the share of college graduates increased in 1970–1990 and 1990–2010. To classify types of education expansion, we use two correlation coefficients:</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msub><mi>R</mi><mrow><mn>1</mn></mrow></msub></math> </ephtml> is the correlation coefficient between the skill premium and Gini coefficient, and</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msub><mi>R</mi><mrow><mn>2</mn></mrow></msub></math> </ephtml> is the correlation coefficient between the skill premium and <emph>changes</emph> in the Gini coefficient. Figure 2 shows that there are two patterns of education expansion:</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msub><mi>R</mi><mrow><mn>1</mn></mrow></msub><mo>></mo><mn>0</mn></math> </ephtml> and</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msub><mi>R</mi><mrow><mn>2</mn></mrow></msub><mo>></mo><mn>0</mn></math> </ephtml> correspond to the second and first rows, respectively. When the skill premium is upward-sloping, as in the second row, the Gini coefficient is also upward-sloping. Then, the skill premium and Gini coefficient increase as <emph>ϕ</emph> increases. The upper panel of Figure 3 shows a positive correlation between the two variables (</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msub><mi>R</mi><mrow><mn>1</mn></mrow></msub><mo>></mo><mn>0</mn></math> </ephtml> ). This numerical example implies that a more meritocratic pay system (a smaller <emph>b</emph>) increases both the skill premium and Gini coefficient. When the skill premium is downward-sloping, as in the first row, the slope of the Gini coefficient is positive for a large skill premium (for a small <emph>ϕ</emph>) and becomes negative for a small skill premium (for a large <emph>ϕ</emph>). The decline in the slope of the Gini coefficient as the skill premium becomes smaller (as <emph>ϕ</emph> rises) implies that there is a positive correlation between the skill premium and a change in the Gini coefficient. The lower panel of Figure 3 shows this positive correlation (</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msub><mi>R</mi><mrow><mn>2</mn></mrow></msub><mo>></mo><mn>0</mn></math> </ephtml> ): as <emph>ϕ</emph> increases (from top right to bottom left), the skill premium and change in the Gini coefficient decrease, except when <emph>ϕ</emph> is close to 1[<reflink idref="bib12" id="ref71">12</reflink>]. This example shows that an inverted-U relationship arises between the Gini coefficient and share of skilled workers[<reflink idref="bib13" id="ref72">13</reflink>]. In summary,</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msub><mi>R</mi><mrow><mn>1</mn></mrow></msub><mo>></mo><mn>0</mn></math> </ephtml> holds if the skill premium is increasing, and</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msub><mi>R</mi><mrow><mn>2</mn></mrow></msub><mo>></mo><mn>0</mn></math> </ephtml> holds if the skill premium is decreasing during education expansion.</p> <p>Graph: Figure 3. Education, the skill premium, and Gini coefficient. (a) Case 1: meritocracy (second row of Figure 2) and (b) Case 2: the inverted-U curve (first row of Figure 2).</p> <p>Figure 4 shows</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msub><mi>R</mi><mrow><mn>1</mn></mrow></msub></math> </ephtml> (left panel) and</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msub><mi>R</mi><mrow><mn>2</mn></mrow></msub></math> </ephtml> (right panel) for each country. The calculation is based on 14 and 13 time points for</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msub><mi>R</mi><mrow><mn>1</mn></mrow></msub></math> </ephtml> and</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msub><mi>R</mi><mrow><mn>2</mn></mrow></msub></math> </ephtml> , respectively[<reflink idref="bib14" id="ref73">14</reflink>]. Countries are arranged in a descending order of</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msub><mi>R</mi><mrow><mn>1</mn></mrow></msub></math> </ephtml> . The left panel shows that</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msub><mi>R</mi><mrow><mn>1</mn></mrow></msub></math> </ephtml> is positive for most countries. The positive correlation is necessary (but not sufficient) for the upward-sloping skill premium in Figure 2. Hence, we cannot guarantee that the upward-sloping skill premium holds for these countries; however, we can say that the upward-sloping skill premium does not apply to countries with a negative</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msub><mi>R</mi><mrow><mn>1</mn></mrow></msub></math> </ephtml> , such as Russia and Italy. The right panel shows that</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msub><mi>R</mi><mrow><mn>2</mn></mrow></msub></math> </ephtml> is negative for most countries. Since</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msub><mi>R</mi><mrow><mn>2</mn></mrow></msub><mo>></mo><mn>0</mn></math> </ephtml> is necessary for the downward-sloping skill premium in the model, the result suggests that the inverted-U curve does not arise for these countries. Interestingly,</p> <p>Graph</p> <p> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msub><mi>R</mi><mrow><mn>2</mn></mrow></msub><mo>></mo><mn>0</mn></math> </ephtml> for Russia and Italy is consistent with the result that these countries do not satisfy the necessary conditions for the upward-sloping skill premium. Overall, Figure 4 shows that many countries satisfy the necessary conditions for a meritocratic pay system but do not satisfy the necessary conditions for an inverted-U curve. We only discuss the necessary conditions. However, given that skill premiums increased for countries such as India, Mexico, Egypt, and Peru in Figure 1, the empirical relationship does not fully support the view that wage inequality decreases as the supply of college-educated workers increases (as suggested by the inverted-U curves).</p> <p>Graph: Figure 4. Correlation coefficients R1 and R2: 1970–2010. (a) R1>0 if merit-based pay is strengthened and (b) R2>0 if an inverted-U arises.</p> <p>The observed correlations in Figure 4 show that the skill premium and Gini coefficient increased persistently during education expansion in many countries. This result per se does not exclude the explanation based on SBTC[<reflink idref="bib15" id="ref74">15</reflink>]. However, it is important to note that the SBTC explanation applies the most to the United States and other developed countries. Figure 4 demonstrates that the skill premium and Gini coefficient increase in many developing countries, where SBTC is less likely (they do not have a sufficient skilled workforce). While the SBTC explanation focuses on technology, our explanation is based on worker reward systems. The result in this section suggests that meritocratic reward systems can be an explanation for the rise in the skill premium.</p> <hd id="AN0191012013-8">4. Conclusions</hd> <p>Wage inequality has been studied extensively since the 1990s. The literature has focused on SBTC to explain the puzzling phenomenon that skill premium increases as the supply of college-educated workers increases. However, how the Gini coefficient and skill premium interact with the supply of skilled workers has been an open question. To answer this question, we investigate an individual's choice to receive higher education by focusing on factors other than technical change. Our model shows that the Gini coefficient and skill premium can increase persistently during education expansion. We derive the conditions under which wage inequality widens as the supply of skilled workers increases. If more people go to college because of lower education costs, the Gini coefficient increases initialy but decreases later on. However, if more people go to college because of higher rewards for skills, the Gini coefficient keeps increasing. In this case, the skill premium increases as the supply of skilled workers increases. Using data on wage inequality from 1970 to 2010, we show that the skill premium and Gini coefficient increased persistently in many developing countries, which typically lack a sufficient skilled workforce. The result suggests that SBTC is not necessarily the main reason behind the rise in wage inequality.</p> <p>Our model shows that when more businesses introduce merit-based pay, more people become skilled workers; however, the Gini coefficient rises. In other words, meritocracy enhances education investment but exacerbates wage inequality. Inequality caused by merit-based pay has been often justified. Nonetheless, when strengthening meritocracy leads to a persistent rise in wage inequality, concerns remain that the level of inequality may exceed what society can tolerate.</p> <hd id="AN0191012013-9">Disclosure statement</hd> <p>No potential conflict of interest was reported by the author(s).</p> <hd id="AN0191012013-10">Supplemental Material</hd> <p>Supplemental data for this article can be accessed online at <ulink href="http://dx.doi.org/10.1080/09645292.2025.2477469">http://dx.doi.org/10.1080/09645292.2025.2477469</ulink>.</p> <ref id="AN0191012013-11"> <title> Notes </title> <blist> <bibl id="bib1" idref="ref3" type="bt">1</bibl> <bibtext> Gross tertiary school enrollment has reached approximately 40 percent worldwide and almost 80 percent in high-income countries. See data from the UNESCO Institute for Statistics (https://data.worldbank.org/indicator/SE.TER.ENRR).</bibtext> </blist> <blist> <bibl id="bib2" idref="ref13" type="bt">2</bibl> <bibtext> For example, Karabarbounis and Neiman ([18]) show that labor share has been declining worldwide; Piketty and Zucman ([30]) and Saez and Zucman ([32]) show that asset income is a key driver of a growing income inequality.</bibtext> </blist> <blist> <bibl id="bib3" idref="ref16" type="bt">3</bibl> <bibtext> We use the wage dataset by Hammar and Waldenström ([15]), which is based on the UBS Prices and Earnings reports. Data on the share of college graduates are obtained from Lee and Lee ([23]). Three-letter ISO codes indicate countries. See the Data Appendix for details.</bibtext> </blist> <blist> <bibl id="bib4" idref="ref10" type="bt">4</bibl> <bibtext> The term 'meritocracy' describes a social system in which rewards are based on ability and effort rather than on social class (Young [33]).</bibtext> </blist> <blist> <bibl id="bib5" idref="ref7" type="bt">5</bibl> <bibtext> The assumption is similar to that of Galor and Moav ([14]), who examine the effect of technological progress on the skill premium.</bibtext> </blist> <blist> <bibl id="bib6" idref="ref22" type="bt">6</bibl> <bibtext> Scale factor <emph>A</emph> determines the wage level but does not affect wage inequality below. None of our results depend on <emph>A</emph>.</bibtext> </blist> <blist> <bibl id="bib7" idref="ref11" type="bt">7</bibl> <bibtext> Educated workers tend to have a better health because of their knowledge and lifestyles. See, for example, Lleras-Muney ([25]), Mackenbach et al. ([26]), and Murtin et al. ([28]).</bibtext> </blist> <blist> <bibl id="bib8" idref="ref8" type="bt">8</bibl> <bibtext> The Gini coefficient is unaffected by a fixed percentage raise (e.g. a 10 percent increase in income) for the whole population.</bibtext> </blist> <blist> <bibl id="bib9" idref="ref45" type="bt">9</bibl> <bibtext> The results are not dependent on the use of particular inequality measures. The inverted-U relationship arises when we use the variance of the log of income (Knight and Sabot [20]; Robinson [31]) and the coefficient of variation. Using these two measures, we also show a persistent increase in wage inequality during education expansion.</bibtext> </blist> <blist> <bibtext> The same results hold when we assume that individuals with higher ability levels enter higher-quality universities and acquire higher skills (</bibtext> </blist> <blist> <bibtext>Graph</bibtext> </blist> <blist> <bibtext> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><mi>h</mi><mo>=</mo><mi>h</mi><mo>(</mo><mi>α</mi><mo>)</mo></math> </ephtml> and</bibtext> </blist> <blist> <bibtext>Graph</bibtext> </blist> <blist> <bibtext> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msup><mi>h</mi><mo>′</mo></msup><mo>></mo><mn>0</mn></math> </ephtml> ). Assuming</bibtext> </blist> <blist> <bibtext>Graph</bibtext> </blist> <blist> <bibtext> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><mi>h</mi><mo>=</mo><msup><mi>α</mi><mrow><mn>0.05</mn></mrow></msup></math> </ephtml> (instead of <emph>h</emph> = 1), we obtain the upward-sloping skill premium and the downward-sloping skill premium, as in Figure 2.</bibtext> </blist> <blist> <bibtext> Skill mismatches may also play a role when most people go to college. Individuals with relatively low innate abilities have difficulty finding high-paying jobs. They end up in jobs requiring lower qualifications, and hence, the wage difference among college graduates increases.</bibtext> </blist> <blist> <bibtext> When <emph>ϕ</emph> is close to 1, the relationship does not hold because the Gini coefficient never goes below zero.</bibtext> </blist> <blist> <bibtext> If the supply of skilled workers is sufficiently high,</bibtext> </blist> <blist> <bibtext>Graph</bibtext> </blist> <blist> <bibtext> <ephtml> <math xmlns="http://www.w3.org/1998/Math/MathML"><msub><mi>R</mi><mrow><mn>1</mn></mrow></msub><mo>></mo><mn>0</mn></math> </ephtml> holds even when the inverted-U curve arises. For example, as long as <emph>ϕ</emph> is larger than 0.5, the Gini coefficient and skill premium are positively correlated in the first row of Figure 2. However, this was not the case for many countries before the 1990s (<emph>ϕ</emph> was much smaller).</bibtext> </blist> <blist> <bibtext> See the Data Appendix for details.</bibtext> </blist> <blist> <bibtext> An SBTC explains why the skill premium can increase as the supply of skilled workers increases. While it does not discuss how the Gini coefficient changes, the rise in the skill premium can be consistent with an increase in the Gini coefficient.</bibtext> </blist> </ref> <ref id="AN0191012013-12"> <title> References </title> <blist> <bibtext> Acemoglu, D. 1998. " Why Do New Technologies Complement Skills? Directed Technical Change and Wage Inequality." Quarterly Journal of Economics 113 (4): 1055 – 1089. https://doi.org/10.1162/003355398555838.</bibtext> </blist> <blist> <bibtext> Acemoglu, D. 2002. " Directed Technical Change." Review of Economic Studies 69 (4): 781 – 809. https://doi.org/10.1111/roes.2002.69.issue-4.</bibtext> </blist> <blist> <bibtext> Anand, S., and S. R. Kanbur. 1993. " The Kuznets Process and the Inequality-Development Relationship." Journal of Development Economics 40 (1): 25 – 52. https://doi.org/10.1016/0304-3878(93)90103-T.</bibtext> </blist> <blist> <bibtext> Autor, D. H., L. F. Katz, and A. B. Krueger. 1998. " Computing Inequality: Have Computers Changed the Labor Market? " Quarterly Journal of Economics 113 (4): 1169 – 1213. https://doi.org/10.1162/003355398555874.</bibtext> </blist> <blist> <bibtext> Baye, F. M., E. L. Wirba, and E. N. Tingum. 2024. " Impact of Education on Inequality across the Unconditional Wage Distribution in Cameroon." Education Economics 32 (3): 298 – 316. https://doi.org/10.1080/09645292.2023.2208783.</bibtext> </blist> <blist> <bibtext> Ben-Porath, Y. 1967. " The Production of Human Capital and the Life Cycle of Earnings." Journal of Political Economy 75 (4, Part 1): 352 – 365. https://doi.org/10.1086/259291.</bibtext> </blist> <blist> <bibtext> Blundell, R., D. A. Green, and W. Jin. 2022. " The UK as a Technological Follower: Higher Education Expansion and the College Wage Premium." Review of Economic Studies 89 (1): 142 – 180. https://doi.org/10.1093/restud/rdab034.</bibtext> </blist> <blist> <bibtext> Castelló-Climent, A., and R. Doménech. 2021. " Human Capital and Income Inequality Revisited." Education Economics 29 (2): 194 – 212. https://doi.org/10.1080/09645292.2020.1870936.</bibtext> </blist> <blist> <bibtext> Cohen, D., and L. Leker. 2016. " Testing the Ben-Porath Effect through the Educational Patterns of Young Cohorts." Journal of Macroeconomics 48 : 252 – 262. https://doi.org/10.1016/j.jmacro.2016.04.001.</bibtext> </blist> <blist> <bibtext> Cohen, D., and M. Soto. 2007. " Growth and Human Capital: Good Data, Good Results." Journal of Economic Growth 12 (1): 51 – 76. https://doi.org/10.1007/s10887-007-9011-5.</bibtext> </blist> <blist> <bibtext> De Gregorio, J., and J.-W. Lee. 2002. " Education and Income Inequality: New Evidence from Cross-Country Data." Review of Income and Wealth 48 (3): 395 – 416. https://doi.org/10.1111/roiw.2002.48.issue-3.</bibtext> </blist> <blist> <bibtext> de la Croix, D., and O. Licandro. 2013. " The Child is Father of the Man: Implications for the Demographic Transition." Economic Journal 123 (567): 236 – 261. https://doi.org/10.1111/j.1468-0297.2012.02523.x.</bibtext> </blist> <blist> <bibtext> Földvári, P., and B. van Leeuwen. 2011. " Should less Inequality in Education Lead to a More Equal Income Distribution? " Education Economics 19 (5): 537 – 554. https://doi.org/10.1080/09645292.2010.488472.</bibtext> </blist> <blist> <bibtext> Galor, O., and O. Moav. 2000. " Ability-Biased Technological Transition, Wage Inequality, and Economic Growth." Quarterly Journal of Economics 115 (2): 469 – 497. https://doi.org/10.1162/003355300554827.</bibtext> </blist> <blist> <bibtext> Hammar, O., and D. Waldenström. 2020. " Global Earnings Inequality, 1970–2018." Economic Journal 130 (632): 2526 – 2545. https://doi.org/10.1093/ej/ueaa109.</bibtext> </blist> <blist> <bibtext> Hazan, M. 2009. " Longevity and Lifetime Labor Supply: Evidence and Implications." Econometrica 77 : 1829 – 1863. https://doi.org/10.3982/ECTA8107.</bibtext> </blist> <blist> <bibtext> Hazan, M., and H. Zoabi. 2006. " Does Longevity Cause Growth? a Theoretical Critique." Journal of Economic Growth 11 (4): 363 – 376. https://doi.org/10.1007/s10887-006-9008-5.</bibtext> </blist> <blist> <bibtext> Karabarbounis, L., and B. Neiman. 2014. " The Global Decline of the Labor Share." Quarterly Journal of Economics 129 (1): 61 – 103. https://doi.org/10.1093/qje/qjt032.</bibtext> </blist> <blist> <bibtext> Knight, J. B. 1976. " Explaining Income Distribution in less Developed Countries: A Framework and An Agenda." Oxford Bulletin of Economics and Statistics 38 (3): 161 – 177. https://doi.org/10.1111/obes.1976.38.issue-3.</bibtext> </blist> <blist> <bibtext> Knight, J. B., and R. H. Sabot. 1983. " Educational Expansion and the Kuznets Effect." American Economic Review 73 : 1132 – 1136.</bibtext> </blist> <blist> <bibtext> Kunovich, S., and K. M. Slomczynski. 2007. " Systems of Distribution and a Sense of Equity: A Multilevel Analysis of Meritocratic Attitudes in post-Industrial Societies." European Sociological Review 23 (5): 649 – 663. https://doi.org/10.1093/esr/jcm026.</bibtext> </blist> <blist> <bibtext> Kuznets, S. 1955. " Economic Growth and Income Inequality." American Economic Review 45 : 1 – 28.</bibtext> </blist> <blist> <bibtext> Lee, J.-W., and H. Lee. 2016. " Human Capital in the Long Run." Journal of Development Economics 122 : 147 – 169. https://doi.org/10.1016/j.jdeveco.2016.05.006.</bibtext> </blist> <blist> <bibtext> Lee, J.-W., and H. Lee. 2018. " Human Capital and Income Inequality." Journal of the Asia Pacific Economy 23 (4): 554 – 583. https://doi.org/10.1080/13547860.2018.1515002.</bibtext> </blist> <blist> <bibtext> Lleras-Muney, A. 2022. "Education and Income Gradients in Longevity: The Role of Policy." NBER Working Paper Series, National Bureau of Economic Research.</bibtext> </blist> <blist> <bibtext> Mackenbach, J. P., J. R. Valverde, B. Artnik, M. Bopp, H. Brønnum-Hansen, P. Deboosere, R. Kalediene, K. Kovács, M. Leinsalu, and P. Martikainen. 2018. " Trends in Health Inequalities in 27 European Countries." Proceedings of the National Academy of Sciences 115 (25): 6440 – 6445. https://doi.org/10.1073/pnas.1800028115.</bibtext> </blist> <blist> <bibtext> Morrisson, C., and F. Murtin. 2013. " The Kuznets Curve of Human Capital Inequality: 1870–2010." Journal of Economic Inequality 11 (3): 283 – 301. https://doi.org/10.1007/s10888-012-9227-2.</bibtext> </blist> <blist> <bibtext> Murtin, F., J. Mackenbach, D. Jasilionis, and M. M. d'Ercole. 2017. "Inequalities in Longevity by Education in OECD Countries: Insights from New OECD Estimates." OECD Statistics Working Papers.</bibtext> </blist> <blist> <bibtext> Park, K. H. 1996. " Educational Expansion and Educational Inequality on Income Distribution." Economics of Education Review 15 (1): 51 – 58. https://doi.org/10.1016/0272-7757(95)00000-3.</bibtext> </blist> <blist> <bibtext> Piketty, T., and G. Zucman. 2014. " Capital is Back: Wealth-Income Ratios in Rich Countries 1700–2010." Quarterly Journal of Economics 129 (3): 1255 – 1310. https://doi.org/10.1093/qje/qju018.</bibtext> </blist> <blist> <bibtext> Robinson, S. 1976. " A Note on the U Hypothesis Relating Income Inequality and Economic Development." American Economic Review 66 (3): 437 – 440.</bibtext> </blist> <blist> <bibtext> Saez, E., and G. Zucman. 2016. " Wealth Inequality in the United States since 1913: Evidence from Capitalized Income Tax Data." Quarterly Journal of Economics 131 (2): 519 – 578. https://doi.org/10.1093/qje/qjw004.</bibtext> </blist> <blist> <bibtext> Young, M. 1958. The Rise of the Meritocracy 1870–2033 An Essay on Education and Equality. London : Thames and Hudson.</bibtext> </blist> </ref> <aug> <p>By Hidefumi Kasuga and Yuichi Morita</p> <p>Reported by Author; Author</p> </aug> <nolink nlid="nl1" bibid="bib24" firstref="ref1"></nolink> <nolink nlid="nl2" bibid="bib27" firstref="ref2"></nolink> <nolink nlid="nl3" bibid="bib11" firstref="ref4"></nolink> <nolink nlid="nl4" bibid="bib29" firstref="ref6"></nolink> <nolink nlid="nl5" bibid="bib13" firstref="ref9"></nolink> <nolink nlid="nl6" bibid="bib22" firstref="ref15"></nolink> <nolink nlid="nl7" bibid="bib19" firstref="ref17"></nolink> <nolink nlid="nl8" bibid="bib20" firstref="ref18"></nolink> <nolink nlid="nl9" bibid="bib31" firstref="ref19"></nolink> <nolink nlid="nl10" bibid="bib12" firstref="ref61"></nolink> <nolink nlid="nl11" bibid="bib10" firstref="ref62"></nolink> <nolink nlid="nl12" bibid="bib16" firstref="ref63"></nolink> <nolink nlid="nl13" bibid="bib17" firstref="ref64"></nolink> <nolink nlid="nl14" bibid="bib33" firstref="ref68"></nolink> <nolink nlid="nl15" bibid="bib21" firstref="ref70"></nolink> <nolink nlid="nl16" bibid="bib14" firstref="ref73"></nolink> <nolink nlid="nl17" bibid="bib15" firstref="ref74"></nolink>
Header DbId: eric
DbLabel: ERIC
An: EJ1500863
AccessLevel: 3
PubType: Academic Journal
PubTypeId: academicJournal
PreciseRelevancyScore: 0
IllustrationInfo
Items – Name: Title
  Label: Title
  Group: Ti
  Data: Why Doesn't Expanding Higher Education Decrease Wage Inequality?
– Name: Language
  Label: Language
  Group: Lang
  Data: English
– Name: Author
  Label: Authors
  Group: Au
  Data: <searchLink fieldCode="AR" term="%22Hidefumi+Kasuga%22">Hidefumi Kasuga</searchLink><br /><searchLink fieldCode="AR" term="%22Yuichi+Morita%22">Yuichi Morita</searchLink>
– Name: TitleSource
  Label: Source
  Group: Src
  Data: <searchLink fieldCode="SO" term="%22Education+Economics%22"><i>Education Economics</i></searchLink>. 2026 34(1):130-142.
– Name: Avail
  Label: Availability
  Group: Avail
  Data: Routledge. Available from: Taylor & Francis, Ltd. 530 Walnut Street Suite 850, Philadelphia, PA 19106. Tel: 800-354-1420; Tel: 215-625-8900; Fax: 215-207-0050; Web site: http://www.tandf.co.uk/journals
– Name: PeerReviewed
  Label: Peer Reviewed
  Group: SrcInfo
  Data: Y
– Name: Pages
  Label: Page Count
  Group: Src
  Data: 13
– Name: DatePubCY
  Label: Publication Date
  Group: Date
  Data: 2026
– Name: TypeDocument
  Label: Document Type
  Group: TypDoc
  Data: Journal Articles<br />Reports - Descriptive
– Name: Audience
  Label: Education Level
  Group: Audnce
  Data: <searchLink fieldCode="EL" term="%22Higher+Education%22">Higher Education</searchLink><br /><searchLink fieldCode="EL" term="%22Postsecondary+Education%22">Postsecondary Education</searchLink>
– Name: Subject
  Label: Descriptors
  Group: Su
  Data: <searchLink fieldCode="DE" term="%22Higher+Education%22">Higher Education</searchLink><br /><searchLink fieldCode="DE" term="%22Outcomes+of+Education%22">Outcomes of Education</searchLink><br /><searchLink fieldCode="DE" term="%22Wages%22">Wages</searchLink><br /><searchLink fieldCode="DE" term="%22Skilled+Workers%22">Skilled Workers</searchLink><br /><searchLink fieldCode="DE" term="%22Job+Skills%22">Job Skills</searchLink><br /><searchLink fieldCode="DE" term="%22Merit+Pay%22">Merit Pay</searchLink><br /><searchLink fieldCode="DE" term="%22Access+to+Education%22">Access to Education</searchLink><br /><searchLink fieldCode="DE" term="%22Cost+Effectiveness%22">Cost Effectiveness</searchLink><br /><searchLink fieldCode="DE" term="%22Developing+Nations%22">Developing Nations</searchLink><br /><searchLink fieldCode="DE" term="%22Foreign+Countries%22">Foreign Countries</searchLink><br /><searchLink fieldCode="DE" term="%22College+Graduates%22">College Graduates</searchLink>
– Name: DOI
  Label: DOI
  Group: ID
  Data: 10.1080/09645292.2025.2477469
– Name: ISSN
  Label: ISSN
  Group: ISSN
  Data: 0964-5292<br />1469-5782
– Name: Abstract
  Label: Abstract
  Group: Ab
  Data: The supply of college-educated workers has increased in many countries, but wage inequality has not necessarily decreased. Although skill-biased technical change can explain this phenomenon, we often observe that skill premiums increase in developing countries, where skill-biased technical change is less likely. We show that skill premiums can increase as the supply of skilled workers increases if merit-based pay is introduced. We find that from 1970 to 2010, in many countries, the Gini coefficient and skill premium increased as higher education expanded. The result suggests that meritocracy expands higher education but exacerbates wage inequality.
– Name: AbstractInfo
  Label: Abstractor
  Group: Ab
  Data: As Provided
– Name: DateEntry
  Label: Entry Date
  Group: Date
  Data: 2026
– Name: AN
  Label: Accession Number
  Group: ID
  Data: EJ1500863
PLink https://search.ebscohost.com/login.aspx?direct=true&site=eds-live&db=eric&AN=EJ1500863
RecordInfo BibRecord:
  BibEntity:
    Identifiers:
      – Type: doi
        Value: 10.1080/09645292.2025.2477469
    Languages:
      – Text: English
    PhysicalDescription:
      Pagination:
        PageCount: 13
        StartPage: 130
    Subjects:
      – SubjectFull: Higher Education
        Type: general
      – SubjectFull: Outcomes of Education
        Type: general
      – SubjectFull: Wages
        Type: general
      – SubjectFull: Skilled Workers
        Type: general
      – SubjectFull: Job Skills
        Type: general
      – SubjectFull: Merit Pay
        Type: general
      – SubjectFull: Access to Education
        Type: general
      – SubjectFull: Cost Effectiveness
        Type: general
      – SubjectFull: Developing Nations
        Type: general
      – SubjectFull: Foreign Countries
        Type: general
      – SubjectFull: College Graduates
        Type: general
    Titles:
      – TitleFull: Why Doesn't Expanding Higher Education Decrease Wage Inequality?
        Type: main
  BibRelationships:
    HasContributorRelationships:
      – PersonEntity:
          Name:
            NameFull: Hidefumi Kasuga
      – PersonEntity:
          Name:
            NameFull: Yuichi Morita
    IsPartOfRelationships:
      – BibEntity:
          Dates:
            – D: 01
              M: 01
              Type: published
              Y: 2026
          Identifiers:
            – Type: issn-print
              Value: 0964-5292
            – Type: issn-electronic
              Value: 1469-5782
          Numbering:
            – Type: volume
              Value: 34
            – Type: issue
              Value: 1
          Titles:
            – TitleFull: Education Economics
              Type: main
ResultId 1