Who Is Affected by the Minimum Wage?

• Author: Paul Wolfson,Kritkorn Nawakitphaitoon,Dale Belman
• Published date: 01 October 2015
• DOI: http://doi.org/10.1111/irel.12107
• Date: 01 October 2015

Who Is Affected by the Minimum Wage?*

DALE BELMAN, PAUL WOLFSON, and KRITKORN

NAWAKITPHAITOON

Prior surveys of empirical research on the minimum wage have been organized

around the question “What does the minimum wage affect?”This survey is orga-

nized around the question “Who is affected by the minimum wage?”We review

the consequences of the minimum wage for teens and young workers, men and

women, African Americans and Hispanics, the less educated, workers in low-

wage industries, and low-wage/low-income populations. Although there is almost

universal agreement that the minimum wage boosts earnings, evidence for a nega-

tive employment effect varies between mixed and nonexistent. An important gap

in the literature is the paucity of research on low-wage/low-income groups.

Introduction

Surveys and literature reviews of empirical research on the minimum wage

are typically organized around specific outcomes that the minimum wage may

influence. Employment is the most common (Neumark and Wascher 2007),

perhaps in partnership with variables such as unemployment (Brown, Gilroy,

and Kohen 1982); the distribution of income (Brown 1999); and wages,

poverty, and the costs to employees and shareholders (Card and Krueger

1995). Neumark and Wascher’s (2008) book has chapters devoted to most of

these variables as well as skills and prices, as does our book, What Does the

Minimum Wage Do? (Belman and Wolfson 2014). In examining specific out-

comes, the analyses considered in the surveys above almost always consider

them not in the abstract, not with respect to entire economies, but for specific

subpopulations or industries within an economy; common examples are teen-

agers and the food-and-drink sector. The result is a patchwork of effects with

limited insight into the comprehensive impact of the minimum wage on

*The authors’affiliations are, respectively, Michigan State University, East Lansing, Michigan. Email:

drdale@msu.edu; The Tuck School of Business at Dartmouth College, Hanover, New Hampshire. Email:

paul.wolfson@tuck.

dartmouth.edu; Renmin University of China, Beijing, China. Email: kengha_di@hotmail.com.

The W. E. Upjohn Institute for Employment Research provided support for the work from which this

survey developed.

INDUSTRIAL RELATIONS, Vol. 54, No. 4 (October 2015). ©2015 Regents of the University of California

Published by Wiley Periodicals, Inc., 350 Main Street, Malden, MA 02148, USA, and 9600 Garsington

Road, Oxford, OX4 2DQ, UK.

582

segments of the population. What follows is a survey organized around the

question, “who?”rather than “what”; a survey in which the primary dimension

for organizing and discussing analyses is the group or industry under study.

There are two benefits to this approach. From a policy perspective, the

benefit is that this article brings together in one place the known effects of

minimum wage policy on specific groups of people and on specific industries

in the economy: teenagers, women, people of color, the restaurant industry,

and employees from low-income families. From a research perspective, the

article highlights gaps in our knowledge. We use data from the 2012 CPS to

understand the importance of the minimum wage to each group and the

importance of each group to assessing outcomes of the minimum wage. The

combination of the data and research not only clearly reveals the presence of

gaps, it also displays the relative importance of each group and industry to

understanding the effects of the minimum wage. The most obvious example

is that of prime age and older workers and teenagers. Teens are the most

studied group in the minimum-wage literature; more than forty studies (one-

third of this literature) provide estimates of the effect of the minimum wage

on teens. This provides rich information on consequences for employment,

accessions and separations, earnings, and the response of school enrollment.

Teenagers, however, comprise less than one-fifth of workers who earn no

more than the minimum wage and barely one-tenth of those who earn no

more than 1.5 times the minimum wage.

1

Older adults, those at least 25

years old, are a majority of these workers. By now, examination of conse-

quences of the minimum wage for prime age and older adults, for whom

there is little specific research, is certain to be more revealing than yet

another study of teenagers.

After beginning with a discussion of some statistical issues necessary to

evaluate this research, we follow the literature, slicing the working-age popula-

tion along different dimensions in order to focus on demographic groups sus-

pected of being vulnerable to unintended consequences of the minimum wage.

In the second section, we begin by splitting the working-age population along

standard demographic lines: first by age into teenagers, young adults, and older

adults; second by gender; and third by race and ethnicity into Hispanics and

non-Hispanic blacks. In the fourth classification, nonstudent workers are distin-

guished according to the highest level of education achieved, in order to con-

sider the interplay between the minimum wage and low levels of education.

The next section is a consideration of the very limited research of effects on

1

In point of fact, 103 percent rather than 100 percent of the actual minimum wage. This approximation,

not uncommon in the literature, allows for spikes in the measured wage distribution thought to reflect round-

ing and mistakes of respondent recall, without substantially changing the measurements.

Who Is Affected by the Minimum Wage? / 583

full- and part-time workers, followed by a section laying out the more

extensive research on specific low-wage industries. Finally, earners are classi-

fied by family income categories to examine the concern that in focusing on

low-wage workers, the minimum wage is insufficiently focused on low-income

workers.

2

Methodological Issues

Identification. The issue identified by the phrase “Correlation is not cau-

sation”has been of interest to economics, a largely observational science, at

least since Working (1927).

3

Morgan (1990), and CHRIST, CARL F., Christ

(1994). Experimental sciences can resolve the issue either explicitly by sys-

tematic variation in causal variables or implicitly by randomly assigning

experimental subjects to treatment or control groups. Although these solu-

tions are possible in the study of government participatory programs and

occasionally those of private organizations, statistical solutions have received

much attention because those of the experimental sciences are unavailable

for most topics of economic interest. Prior to the 1991 conference that

launched the New Minimum Wage Research (NMWR), minimum wage

research had not shared the increasing interest in the possibility of more

nearly mimicking controlled experiments (Meyer 1995). Instead, it relied on

estimated partial correlations from one-dimensional observational data, pri-

marily time series, and less often cross-sectional, data (Brown, Gilroy, and

Kohen 1982).

4,5

The conference papers contrasted sharply with earlier practice. Neumark and

Wascher (1992) introduced panel data and models to the field, adapting the

analytic framework previously applied to time-series data in the hope that

greater variation in the data would allow for more careful measurement of the

relationship between employment and the minimum wage. Card (1992) and

Katz and Kreuger (1992) introduced what we would now recognize as early

versions of quasi-experiments (QEs): the latter, a differences analysis, the former

2

This review is not exhaustive. We limit ourselves to articles on developed countries, with a heavy

emphasis on studies of the U.S. experience. How far back we reach into the literature varies according to

how much recent work there has been on a topic; in discussing employment effects, we generally limit our

discussion to work published since 2000, but for topics where there are few articles, such as work on educa-

tion and training, we include earlier studies.

3

See also Christ (1994), Epstein (1987), Morgan (1990).

4

In October 1992, Industrial and Labor Relations Review (46:1) included a symposium consisting of

five papers presented at the conference and Ehrenberg’s (1992) introduction.

5

Wooldridge (2010: Chapter 21) contains a more recent discussion of this quasi-experimental literature.

584 / BELMAN,WOLFSON,AND NAWAKITPHAITOON