Human capital investment, credentialing, and wage differentials

• Author: Masashi Tanaka
• Published date: 01 August 2020
• Date: 01 August 2020
• DOI: http://doi.org/10.1111/jpet.12443

J Public Econ Theory. 2020;22:992–1016.wileyonlinelibrary.com/journal/jpet992

|

© 2020 Wiley Periodicals, Inc.

Received: 7 August 2017

|

Accepted: 19 March 2020

DOI: 10.1111/jpet.12443

ORIGINAL ARTICLE

Human capital investment, credentialing, and

wage differentials

Masashi Tanaka

Department of Economics, Fukuyama

University, Fukuyama, Japan

Correspondence

Masashi Tanaka, Department of

Economics, Fukuyama University,

Gakuen‐cho 1, Fukuyama,

Hiroshima 729‐0292, Japan.

Email: masashi.tanaka.0203@gmail.com

Funding information

The Osaka University Institute of Social

and Economic Research

Abstract

This study considers how individuals determine at what

ratio they will invest in two different types of education.

The first type contributes to the development of labor

skills, while the other does not. We refer to the former as

human capital investment and the latter as unproductive

investment,whichimprovestestscoresbuthasnoben-

eficial effect on students' human capital. We formulate

an overlapping‐generations economy in which the rich

and poor households invest in both types of education.

We find that the ratio of human capital investment to

unproductive investment is lower in the economy with

medium size of the wage differentials. In a dynamic

analysis, we identify two patterns of stable steady states

for the dynamics of the wage differentials, namely, no‐

inequality and high‐inequality steady states. Further, we

show that a rapid increase in the level of skill‐biased

technology may cause a switch from a steady state with

no‐inequality to one with high inequality. This causes at

least a temporary increase in the ratio of unproductive

investment during the transition to the new steady state.

1|INTRODUCTION

The effect of schooling, the way it alters a man's capacity and will to do things,

depends not only on what he learns, or the way he learns it, but also on why he

learns it. That is at the basis of the distinction between schooling which is education,

and schooling which is only qualification, a mere process of certificating—or

‘credentialing.’

(Dore, 1976,p.8)

As predicted by Ronald P. Dore, people sometimes invest heavily in “credentialing,”

meaning that they only exert effort to improve test scores for the purpose of obtaining a

credential. It is known that higher educational attainment serves as a signal and has a big

impact on students' postgraduation earnings. Using the U.S. data, Jaeger and Page (1996) and

Park (1999) find evidence of the following diploma effects: earning gains of 9–10% associated

with high school graduation, 11% with associate's degrees, and 20–30% with bachelor's degrees.

Thus, higher education plays the dual role of “educating”and “credentialing.”

The main purpose of this paper is to examine the amount of social loss included in higher

education. To investigate this problem, we introduce two types of educational investments:

human capital investment and unproductive investment. Households are concerned not only

with their level of human capital but also with whether they obtain credentials because their

lifetime income is determined by these two elements. Here, unproductive investment is defined

as an education investment that only contributes to obtaining a credential, not to developing

human capital. On the other hand, human capital investment not only contributes to labor

skills, but it is also helpful for obtaining credentials. Thus, human capital investment is seen as

socially desirable education because it directly affects the total output of the economy. In

contrast, since we assume that the number of households obtaining credentials is fixed, that is,

we assume a zero‐sum case, an increase in unproductive investment does not contribute at all to

the total output. Our interest is examining the ratio of socially beneficial (human capital)

investment and socially wasteful (unproductive) investment.

We introduce a three‐period, overlapping‐generations economy, in which households are

initially heterogeneous in their income, being either rich or poor. In their second period of life,

middle‐aged parents give human capital and unproductive investments to their offspring.

Households work for the second (middle‐age) and third (old‐age) periods in their three‐period

lifetime. We suppose that the levels of human capital and/or individual outputs of middle‐aged

workers are not observed by employers, but those of old‐aged workers are observed. Therefore,

credentials play the role of a signal regarding the individual performance of middle‐aged

workers. In contrast, this does not matter for old‐aged workers because their wages are correctly

consistent with their outputs. Further, we suppose that there is uncertainty in obtaining cre-

dentials, the individual probability of which increases with both the level of human capital and

unproductive investment. The zero‐sum property of the credential‐gaining process implies that

the winning probability of each household has a negative influence on others. Thus, the model

assumes competition between households in obtaining credentials, and it becomes an im-

portant factor in determining the ratio of human capital to unproductive investment.

The results of our study are divided into two parts, static and dynamic analyses. In the static

analysis, we examine how the ratio of human capital to unproductive investment changes with

the wage differentials between rich and poor households. We find that, for both rich and poor

households, there is a nonmonotonic relation between the ratio of human capital to un-

productive investment and the wage differentials. That is to say, human capital investment

makes up a large fraction of total education expenditure in an economy with high or low

inequality. While, as the wage differentials approach medium levels, the share of unproductive

investment increases. The nonmonotonic relationship stems from two opposite effects. Lower

wage differentials lead to a greater degree of competition between rich and poor households,

which leads to larger incentives for unproductive investment. Conversely, the value of cre-

dentials increases with the wage differentials because the wage differentials of middle‐aged

workers represents the wage gap between workers with credentials and workers without them.

TANAKA

|

993