Impact of the Minimum Wage on Youth Labor Markets
Shanshan Liu1, Thomas J. Hyclak2, Krishna Regmi3
Abstract
We study the effect of the minimum wage on labor market outcomes for young workers using
U.S. county-level panel data from the first quarter of 2000 to the first quarter of 2009. We go
beyond the usual estimates of earnings and employment effects to consider how differences
across states in the minimum wage affect worker turnover via separations and accessions and job
turnover through new job creation and job losses. We find that a higher minimum wage level is
associated with higher earning, lower employment and reduced worker turnover for those in the
14 to18 age group. These effects are consistent and robust to different specifications. For
workers aged 19 to 21 and 22 to 24, we find less consistent evidence of important minimum
wage effects on earnings and employment. But, even for these age groups, a higher minimum
wage is found to reduce accessions, separations and the turnover rate. For all three groups it does
not appear that these effects on worker flows in the labor market are matched by effects on job
creation and job losses at the establishment level.
JEL Codes: J21, J23, J38, J63.
Keywords: Minimum Wage, Youth Employment, Worker and Job Turnover.
1
Ph.D. in Economics, Lehigh University. Contact Author. Address: Department of Economics, 621 Taylor Street, Bethlehem, PA
18015, USA. Email: [email protected]. Tel: 484-767-1682
2
Professor, Department of Economics, Lehigh University. Email: [email protected]
3
Ph.D. Candidate in Economics, Lehigh University. Email: [email protected]
I. Introduction
Periodic debates over legislative proposals to increase the statutory minimum wage have led
to a voluminous empirical literature presenting estimates of the employment effects of minimum
wages. While it might be argued that these efforts generally support the conclusion that a 10%
higher minimum wage would likely result in 1% to 3% fewer jobs for affected workers, the
estimated effects differ widely for various groups of workers and there is still an active debate
over the existence of a negative employment effect. Most recently, this debate has centered on
whether inadequate control for spatial heterogeneity leads to spurious estimates of a negative
employment effect in studies relying on cross-state variation to identify the effect of higher
minimum wages.
A much smaller literature has begun to use new data sources that permit empirical analysis of
the effect of minimum wage levels on labor market flows: worker turnover via separations and
accessions and job turnover via job creation at new- and growing establishments and job losses
at declining or failing business units. Labor market flows are of interest in part because they
dwarf net employment changes in magnitude and because they can be used to test hypotheses
from models of firm and worker search with heterogeneous jobs and imperfect information. A
simple example shows how an understanding of flows may be relevant for the employment
effects debate. The level of employment observed in a local labor market at a point in time can
be described as the difference between the quantity of labor demanded and the number of jobs
that are vacant. If a higher minimum wage lowers the quantity of labor demanded and the
number of vacancies (by affecting worker turnover rates) then the net effect on observed
employment is uncertain. It is also clear that estimates of the effect of the minimum wage on job
creation and destruction at the establishment level would be highly relevant for attempts to assess
1
the effectiveness of this policy.
This paper extends the examination of the impact of the minimum wage beyond the earnings
and employment effects to an examination of the impact of the minimum wage on dynamic
aspects of youth labor markets, such as hiring, separation through quits and layoffs, and
establishment level job creation and destruction. We use cross-state variation to identify
minimum wage effects and attempt to control for spatial heterogeneity by including separate time
effects for each of the 179 Economic Areas (EA) identified by Bureau of Economic Analysis in
our panel regressions. We also employ a mixed/fixed effect model to use lagged dependent
variables for each county to control for unmeasured heterogeneity and idiosyncratic trends.
The level of the minimum wage in a county has important effects on labor market outcomes
for the youngest workers – resulting in higher monthly earning, lower employment and lower
worker turnover. For slightly older workers in the 19 to 21 and 22 to 24 age groups we find
limited impact on earnings and employment but similar significant declines in worker
separations from employment and hires into new positions. Our results add to a small but
growing number of studies in suggesting that the most consistent impact of the minimum wage
may well be its negative influence on worker turnover rates.
The rest of the paper proceeds as follows. In section II we briefly review the literature, with
a focus on labor market flows. Section III describes the data. Section IV presents the empirical
model. Section V contains results; Section VI offers an alternative specification. Section VII
concludes.
II. Background
A large empirical literature, mainly looking at teenage workers or those employed in
restaurants or other establishments who are likely to be affected by minimum wage laws, has
2
examined the effect of such laws on the levels of wages and employment. There is considerable
support for the competitive market hypothesis that an effective minimum wage would result in
lower employment. For example, Neumark and Wascher (2008) conclude that panel data studies
of cross-state variation in minimum wage levels with time and state fixed effects provide
renewed support for the consensus employment elasticity estimate of -0.1 to -0.3 identified in
Brown’s (1982) influential survey of time series studies. Recent papers reporting similar results
include Sabia (2009), Thompson (2009), Kalenkoski and Lacombe (2011) and Sabia (2012).
However, the results in this literature are not uniform and a substantial number of papers
report zero or positive employment responses to higher minimum wages. Card and Krueger
(1994 and 1995) explain positive employment responses by arguing that job and worker
heterogeneity means that employers are wage setters acting as monopsonistic buyers of labor
(see Burdett and Mortensen, 1998, and Bhaskar, Manning and To, 2002). In addition, several
recent papers argue that effective controls for spatial heterogeneity in employment trends results
in a zero estimated response of employment to cross-state variation in the minimum wage. This
is the finding in Dube, Lester, and Reich’s (2010) study of a panel of contiguous border countypairs in the United States over the years from 1990 to 2006 and in papers by Allegretto, Dube
and Reich (2011) and Addison, Blackburn, and Cotti (2009).
Neumark et al. (2014) provide a detailed critique of these approaches to controlling for
spatial heterogeneity. They argue that the strategy of limiting identification of the minimum wage
effect to within-area or relative-to-area-trend variation leads to neglect of valid information. Still
this cluster of minimum wage studies raises an important question about the method of testing
for minimum wage effects in local labor markets. We compare results using the panel data
approach with estimates that incorporate controls for spatial heterogeneity to add additional
3
information on this issue.
Far less attention has been paid to the impact of minimum wage levels on labor market flows
so this is the primary focus of the rest of this review and of our empirical analysis below. Our
data allow us to study two related sets of flows in local labor markets for young people: worker
reallocation flows and job reallocation flows. Worker reallocation flows are separations from
employment through quits, layoffs or firings and accessions of individuals to jobs through hiring
or recalls from temporary layoffs. The worker reallocation or turnover rate, defined as the
number of people in a given period who are changing their job situation, is defined as the number
of accessions plus the number of separations divided by the number employed.
While worker reallocation flows take a worker perspective, job reallocation flows count
employment adjustments at the establishment level. Job reallocations stem from job gains due to
added employment at expanding or new establishments in the local labor market and job losses
due to establishment that are contracting or going out of business. The job reallocation rate is
defined as the fraction of workers who would have to change jobs in a given market and period
because of the decline and growth of positions at different establishments and is equal to the
number of jobs created plus the number of jobs lost divided by the level of employment. The
worker reallocation rate generally exceeds the job reallocation rate because some individuals
make multiple job changes in a given period. The difference between these rates is the rate of job
“churning” in the particular labor market.
Worker reallocation
In considering the effect of the level of the minimum wage on separations, accessions and
turnover rates it is useful to consider the matching model of the labor market. In labor markets
with workers searching across heterogeneous jobs and firms searching across heterogeneous
4
workers and where all searchers have imperfect, asymmetric information, labor market
equilibrium can be described by an aggregate matching function such as equation (1) below:
M = m(V,S) ,
(1)
where M is the number of matches between workers and firms, or hires, V is the number of
vacant jobs in the labor market and S is the number of people searching for jobs in the labor
market. This function indicates that a greater number of hires will be observed in a given labor
market if the number of vacancies and/or job searchers is higher because the probability of
finding the right employee for the right job increases with a greater pool of vacant jobs and
searchers. The ratio M/S corresponds to the job offer arrival rate and the probability of
transitioning from unemployment to employment, if all of the job searchers are unemployed.
If a higher minimum wage reduces voluntary labor turnover and also reduces the incentives to
shirk on the job and thus the number who are discharged4, the number of separations will decline.
A decrease in the number of separations will also mean a lower number of job vacancies and
fewer job searchers, other things being equal. Since efforts to estimate matching functions such
as equation (1) generally find evidence of constant returns to scale (Petrongolo and Pissarides
2001), we would expect a proportionate decrease in the number of hires (matches) for a given
decrease in vacancies and searchers associated with the effect of a higher minimum wage on
separations. The few studies of the relationship of minimum wages to employee turnover
generally find evidence of negative effects of higher minimum wages on both separations and
accessions.
4
Hirsch, Kaufman and Zelenska (2011), in their study of the adjustments by quick service restaurants to the 2007 and 2009
increases in the national minimum wage, find evidence of significantly lower turnover and higher productivity.
5
Portugal and Cardoso (2006) examine the effect of a substantial increase in the youth
minimum wage in Portugal in 1987, using a difference-in-difference approach to compare
outcomes in 1987 and 1988 to those observed in the years from 1979 to 1986. They find that the
minimum wage increase was associated with a significant drop in the share of teens separated
from continuing firms, a decrease in the share of teens hired by continuing firms5 and an increase
in the probability of teens staying with the same employer. The negative separation effect of the
minimum wage appears to be substantially higher than the negative hiring effect in this study, but
the fact that the dependent variables are expressed as teen shares of all separations and all hires
makes it difficult to compare the minimum wage effects.
Brochu and Green (2013) use cross-province differences in the minimum wage in Canada to
examine the effect of higher minimums on turnover among low skilled workers. They find that
the minimum wage is negatively correlated with both the probability of separation and the
probability of finding a job. For workers of all ages with just a high school education and less
than a year of job tenure, a higher minimum wage is estimated to reduce both probabilities by a
similar quantitative amount. For teens, the estimated negative effect on hiring is about twice the
size of the effect on separations. Of particular interest is their finding that the negative impact of
a higher minimum on the probability of layoff accounts for fully 70% of the estimated effect of
the minimum wage on separations.
Dube, Lester and Reich (2013) use 2001 to 2007 US data from the Quarterly Workforce
Indicators (QWI) program to examine the effect of cross-state differences in minimum wages on
worker flows for teens aged 14-18 and restaurant employees in a sample of cross-border county
pairs designed to control for spatial heterogeneity. For teenage workers, minimum wage
5
Thompson (2009) also reports a significant negative effect of the minimum wage on the teen share of total hires. He does not
report estimates of the separation response.
6
elasticities are estimated at -.22 for hires, -.23 for separations and -.20 for the turnover rate while
for restaurant workers the corresponding estimates are -.26 for hires, -.23 for separations and -.21
for the turnover rate. Thus three studies on three different countries yield a similar empirical
conclusion: a higher minimum wage is associated with a decline in job separations and a decline
in job accessions that are often quite similar in magnitude, as suggested by the matching model
framework.
An interesting issue concerns the possibility that reduced labor market turnover, or “fluidity”,
is associated with lower levels of employment because fewer turnovers lead to a slower arrival
rate of new job opportunities that, in turn, extends the length of unemployment spells and
reduces the extent of job-to-job transitions (Davis and Haltiwanger, 2014). Brochu and Green
(2013) find some evidence relevant to this question that minimum wage levels are associated
with reduced flows from out of the labor force into unemployment and increased flows from
unemployment to labor market inactivity.
Job Reallocation
Search theories suggest potentially offsetting influences of a higher minimum wage on firm
level job creation (Acemoglu, 2001 and Flinn, 2006). On the one hand, a higher minimum wage
has a direct negative effect on the returns from filling new marginal jobs which would be
expected to reduce the net number of jobs created. However, a higher minimum wage could also
result in increased search intensity by all searchers, most importantly higher productivity
workers. Increased search in response to a higher minimum wage could lower the cost of filling a
vacancy and potentially boost the returns from creating and filling new jobs.
These potentially offsetting effects of the minimum wage can be seen in the interesting study
by Giuliano (2013). She examines establishment level employment responses to the percentage
7
change in average store wages required to meet higher federal minimum wage levels using
personnel data for 700 stores of a national retailer for the period from February 1996 to July
1998. The change in overall FTE employment and teenage FTE employment was negatively
correlated with the impact of higher minimum wages for adult workers but positively related to
the effect of higher minimum wages for teenage workers. This evidence, that a higher minimum
wage shifted employment away from adults to teens in these stores, seems to be due to an
increase in search activity by high productivity teens. The fraction of teen employees residing in
high socio-economic zip codes rose significantly, suggesting that the minimum wage drew more
productive teenage workers into the local labor markets in which these stores were located. In
terms of job reallocation, these results imply that stores in her sample would be counted as
creating jobs for teens and destroying jobs for adults with uncertain effects on overall net job
growth depending on how much the minimum wage increase affected adult wages in a given
locality.
Even if one were convinced that the minimum wage effect on the cost of filling a job would
dominate employment decisions or that the negative employment effect drawn from the
competitive labor market model was most likely in a given situation, it still not clear how that
might affect aggregate job creation and job loss rates. Does the high minimum wage lead to
lower employment by a slope shift reducing the rate of net job growth? Or would a high
minimum wage level result in an intercept shift with lower employment at any point in time but
no change in the rate of net job growth, except that observed in the transition from a higher to a
lower growth path?
The working paper by Meer and West (2013) provides some empirical information on the
effect of different minimum wage regimes on state level net job growth, using administrative
8
data from Business Dynamics Statistics (BDS), the Quarterly Census of Employment and Wages
(QCEW) and the Quarterly Workforce Indicators (QWI). In all three data bases, with varying
controls for spatial heterogeneity, they find evidence of a statistically significant negative
relationship between the state minimum wage level and the rate of net job growth but,
interestingly, little evidence of a negative effect on the level of employment. The minimum wage
effect on net job growth is highest for the youngest workers. The BDS and QWI data allow Meer
and West (2013) to separate net job growth into changes due to job creation and job destruction.
They find a statistically significant negative effect of the minimum wage level on job creation in
the BDS data, with elasticities ranging from -.23 to -.36, but no effect on job destruction. With
the QWI data, statistically significant negative effects of the minimum wage level on both job
creation and job destruction are attenuated by added controls, including state level time trends.
Finally, a recent paper by Gittings and Schmutte (2014) finds that state level variation in the
minimum wage was negatively correlated with worker turnover rates for teen workers using state
level QWI data. However, they find no evidence for a statistically significant effect of the
minimum wage on job reallocation. Clearly additional study of the minimum wage effect on job
reallocation is called for. We use county level QWI data to examine the effect of cross-state
differences in the minimum wage level on the overall rates of job creation, job destruction and
job reallocation for workers in different age categories.
III. Data
We use the nationwide data from the Quarterly Workforce Indicators (QWI) from the
Longitudinal Employer-Household Dynamics (LEHD) program at the U.S. Census Bureau. 6 The
6
The QWI variables, including employment, job creation, job loss, net job flows, accessions, separations, and average monthly
9
QWI are built on wage records in the Unemployment Insurance (UI) system and information
from the Bureau of Labor Statistic’s Quarterly Census of Employment and Wages (QCEW,
formerly known as ES-202). 7 The QWI data provide employment levels, employment flows
(accessions, separations and turnover rates), job creations and destructions, and average earnings
for demographic subgroups (age and gender) by different levels of geography: state, county,
metro, and workforce investment area, as well as by detailed industry. We use the QWI data for
46 states, from 2000 through the first quarter of 2009.8 We measure the minimum wage at the
state level as the higher of the federal or state minimum and include the total population each
year at the county level9 in the QWI panel dataset for all counties in these 46 states.
The QWI has two concepts of a job, measured as in full-quarter and point-in-time (not lasting
the full quarter) levels. For example, employment for a full quarter assumes that the individual
has been continuously employed throughout the quarter with the same employer, i.e., he has
valid UI wage records in the current quarter, the preceding quarter, and the subsequent quarter.
Employment at a point in time can be divided into two types in the QWI — at the beginning of
the quarter and at the end of the quarter. When the individual has valid UI wage records for the
current and the preceding (subsequent) quarter she is defined as employed at the beginning (end)
of the quarter (Abowd et al., 2005). In our paper, we use the end-of-quarter employment count as
the measurement of point-in-time employment.
earnings for both full-quarter and point-in-time jobs can be accessed at the county level in each state separately, which are
published by the LEHD Program at the U.S. Census Bureau. http://lehd.did.census.gov/led/
For our study, the full public-use QWI data were accessed through the Cornell Institute for Social and Economic Research, using
the VirtualRDC @ Cornell. http://www2.vrdc.cornell.edu/news/data/qwi-public-use-data/
7
The ES-202 program, also known as the Covered Employment and Wages (CEW) program, includes the employer reports based
on information from each state’s Department of Employment Security.
8
State participation is important for the QWI system to construct the national-wide data. Connecticut, Massachusetts, North
Carolina, New Hampshire, and District of Columbia had not joined the Local Employment Dynamics (LED) Federal/State
Partnership that provides the data, by 2009. The QWI data have been available for Alabama since 2001, Arizona since 2004,
Arkansas since 2002, Kentucky since 2001, Mississippi since 2003, Wyoming since 2001. The QWI data are available for Illinois
from 2000 through 2008.
9
Intercensal Estimates of the Resident Population for Counties, United States Census Bureau.
http://www.census.gov/popest/data/intercensal/county/CO-EST00INT-01.html
10
The QWI system also provides information about accessions, separations, job creations, job
destructions, and net job flows, at both full-quarter and point-in-time levels, respectively.
Accessions and separations are defined by the QWI at the job level (an individual - employer
pair). Point-in-time accessions are defined as the number of new employees added to the payroll
of the employer during the current quarter. That is, the individuals were not employed by the
employer during the previous quarter but received UI-covered earnings during the current
quarter.10 Point-in-time separations are the number of workers who left the employer during the
current quarter. These individuals received UI-covered earnings during the current quarters but
did not receive any earnings in the following quarter. We define the point-in-time turnover rate as
the ratio of the average of point-in-time accessions and separations over the end-of-quarter
employment: Turnover =
.11
To accurately estimate the minimum wage effects on teenage workers, we prefer to employ
the concept of point-in-time jobs and job changes. Thompson (2009) finds that transitory jobs
account for nearly half of all teen employment, compared to just over one-fourth of non-teen
employment; and he defines transitory employment as the difference between total employment
and stable employment (full-quarter employment). The point-in-time data are also more
comparable to other federal statistics that measure labor market situations as of a given day or
week each month.
In the QWI system, dynamic job flows — job creations and destructions — are defined at the
employer level for single-unit employers, or the establishment level for multi-unit employers.
Jobs are created (destroyed) at the establishment if end-of-quarter employment is greater (less)
10
In the QWI system, accessions are divided into two subcategories — new hires and recalls. If there are no valid wage records
for this job within the last four quarters, then an accession into a job during the current quarter is called a new hire; otherwise, it
is a recall.
11
An arbitrary aggregate k = county × age group.
11
than beginning-of-quarter employment. We calculate the net job reallocation rate for point-intime jobs at the county level as the ratio of net point-in-time job changes across the
establishments within the county divided by the end-of-quarter employment.
The QWI data allow us to examine the separate effects of the minimum wage on young
people at the county level for workers in three age categories: 14-18, 19-21 and 22-24. Workers
in these age groups are most likely to be affected by minimum wage legislation. In 2007 about a
fifth of hourly wage earners earning the minimum wage were 16 to 19 years old and nearly half
were under 25.12 And younger workers are also more likely to be constrained in their choice of
residence and so limited to employment opportunities within a geographic area (Ihlanfeldt, 1990
and Stoll, 1999).
We control for spatial heterogeneity in the national sample by allowing for separate time
effects for BEA Economic Areas (EAs). BEA Economic Areas consist of one or more economic
“nodes” — metropolitan or micropolitan statistical areas — and the surrounding counties that are
economically related to these nodes.13 There are 179 Economic Areas centered in the U.S. And,
98 of these Economic Areas are defined to include counties in two or more neighboring states.
So we can take advantage of potential differences in the minimum wage within these Economic
Areas as well as time variation within each Economic Area to help identify minimum wage
effects in local labor markets.
State minimum wages from 2000 to 2009 are reported in Table 1. The federal minimum of
$5.15 per hour prevailed in most states in 2000 but legislation adopting higher state minimum
wages was enacted in all states at several points during the sample period. Thus differences in
federal and state legislation define different minimum wage regimes over time within each state
12
13
U.S Department of Labor, Bureau of Labor Statistics. http://www.bls.gov/cps/minwage2007.pdf
Bureau of Economic Analysis, U.S. Department of Commerce. www.bea.gov/regional/docs/econlist.cfm
12
and across states within the region at various intervals in the sample period.
IV. Empirical Model
Using county-level quarterly data, we examine the effect of the level of the minimum wage
prevailing in each state from 2000 to the first quarter of 2009 (the minimum wage is the higher
of the federal or state minimum wage level) on labor outcomes for youth in three age groups —
teenagers between the ages of 14 and 18, youth between the ages of 19 and 21, and older youth
between the ages of 22 and 24.
We start with what Dube et al. (2010) refer to as the “canonical model” for panel studies of
spatial differences in the minimum wage, which is written as equation (2) below:
where i, s, and t respectively indicate county, state, and quarterly time for all observations. The
dependent variables in our paper can be divided into three sets; all of which are measured for
three age groups at the point-in-time level: first, the static level of employment and earnings —
the natural log of total employment and average monthly earnings; second, employment flows —
the natural log of accessions and separations as well as turnover rates, and third, the dynamic job
changes — the natural log of job creation, destruction, and reallocation rates.
All independent variables have been transformed into natural log form.
refers to
the natural log of the minimum wage, which is the same for all counties within each state in each
quarter. To control for aggregate labor market conditions and the relative size of the local labor
market, two control variables are added — the natural log of total employment of persons
between the ages of 25 to 65 years old at the county level
total population at the county level
and the natural log of
. The model also includes county fixed effects
13
and time effects
common to all of the counties in the sample.
To further address unmeasured spatial heterogeneity in the traditional panel data model, we
modify this model by adding the specific time effects,
, for each multi-county Economic
Area within the region. Our main focus then is comparing the estimates of β1 in regressions with
equation (2) with those obtained from the expanded model in equation (3).
.
(3)
V. Results
This section presents our findings about the effects of the minimum wage on outcomes in
local labor markets for the three groups of young workers in our sample. In all specifications, we
cluster the standard errors at the state level — these are reported in parentheses. The clustering is
to address the concern that unobserved factors or idiosyncratic errors across counties within a
state may be correlated since minimum wages are set at the state level.
Table 2 presents estimates of the employment level and earnings effects in our sample.
Panel A presents results from the two-way fixed effect model, equation (2), and Panel B reports
estimates of the model with EA-specific time effects, equation (3). We prefer equation (3) as
simple two-way fixed-effects results could be driven by spatial heterogeneity, instead of
capturing a true minimum wage effect. We find that higher minimum wage levels are negatively
correlated with the teenage (14 to 18-year-olds) employment level. By adding EA-specific time
effects in the fixed-effects model, we find a slightly lower negative teenage employment effect
with the elasticity of -0.17, compared with that of -0.23 based on the simple two-way fixed effect
model. Thus, our estimates for teenagers fall within the range of -.1 to -.3, as in many previous
14
studies. In contrast to Dube et al. (2010) and Allegretto et al. (2011) a negative teen employment
effect from higher minimum wage levels remains even after controlling for spatial heterogeneity.
We do not find any significant effect on the level of employment for the 19 to 21-year-olds
or 22 to 24-year-olds in either specification in Table 2, although the minimum wage is positively
related to average monthly earnings for those 19 to 21 in the preferred specification in Panel B of
Table 2. Again this result -- that estimated employment and wage effects vary across different
groups of workers -- is common in the literature.
In Tables 3 and 4 we turn to the main focus of our paper, presenting results about the effect
of the minimum wage on employment flows, estimated from equations (2) and (3), respectively.
We consistently find negative effects of the minimum wage on accessions and separations that
are similar in magnitude for all three age groups. This is consistent with the emerging evidence
in the small number of papers addressing this issue and suggests that higher minimum wages
may induce a lower quit rate and greater effort leading to fewer dismissals. With a given level of
employment and a lower separation rate, we would also expect a lower hiring rate since there
would be fewer vacancies to fill. Note that in the regressions controlling for spatial heterogeneity
(Table 4) the absolute value of the elasticity of separations with respect to the minimum wage is
substantially higher than the elasticity of accessions for both 14-18 and 19-21 year olds, the two
groups for which we find positive earnings effects from the minimum wage in Table 2. Another
possible reason for a negative turnover rate elasticity is that a higher minimum wage may shift
the employment distribution away from high-turnover, low-wage firms to low-turnover, highwage ones (Dube, 2011).
In our preferred specification with EA time effects reported in Table 4, the minimum wage
effects on establishment level job creation, job destruction and job reallocation rates are mostly
15
insignificant. We find a negative effect on job loss for 19 to 21 year olds and on job creation for
those 22to 24 but these effects are significant at just the 10 percent level. Note that worker flows
measured as separations, accessions (hiring), and turnover in this paper are affected differently
from job flows such as job creations and job destructions, which are mainly caused by expansion
and contraction in economy. Our overall finding, that there is a negative effect of the minimum
wage on worker reallocation through turnover but no evident impact on job reallocation is also
consistent with the previous studies summarized above.
Our full sample is an unbalanced panel with one state, Arizona, included only after 2004. In
addition, the full sample covers two recessions, the 2001 downturn and the jobless recovery that
followed (Groshen and Potter, 2003) and the very serious recession of 2008-2009. As a
robustness check, we re-estimate equation (3) for the shorter period from the first quarter of 2004
to the fourth quarter of 2007. The results reported in Table 5 are largely similar to the full sample
results reported in Tables 2 and 4 for those in the 14 to 18 age group. For those 19 to 21 there is
evidence of a small but significant negative employment effect and a smaller, statistically
insignificant accession effect. Some interesting results are seen for the age group of 22 to 24
year olds. We find evidence for a marginally significant positive effect on the level of
employment for these workers; the minimum wage appears to have a larger effect on accessions
than on separations for this group, and there is evidence of a negative effect on the creation of
new jobs.
Davis and Haltiwanger (2014) argue that reduced worker turnover, or fluidity in their terms,
might lead to lower employment as the job offer rate falls along with accessions and job
searchers respond by reducing search activity. As a robustness check we re-estimate the effect of
the minimum wage on youth employment levels, controlling for the labor market fluidity by
16
including the worker turnover rate and the job reallocation rate as independent variables in the
regressions. Our results are very close to those from the baseline specification14 reported in Table
4.
In another robustness check, we examine the minimum wage effect for those in the 35 to 44
age group, expecting that the minimum wage would have little impact on these prime working
age individuals. We do not find any statistically significant effects on earning, the employment
level, accessions, job turnover, job destruction, and job creation for this group. The results did
show a very weak negative effect on separations that is significant at just the 10 percent level.15
VI. Alternative Specification
We turn to an alternative specification to control for heterogeneous trends in labor market
outcomes at the county level and estimate the dynamic effects of the minimum wage. The
outcome measures such as employment level or turnover in the current period consist of both
their previous level plus new changes between the previous and current periods. In other words,
current measures are not only affected by the minimum wage, but also by their lagged values.
For this purpose, one can consider a generic dynamic panel data model in the following form:
.
(4)
However, since the lagged value is obviously correlated with the error term, we cannot
consistently estimate equation (4). One difficult-to-implement option for dealing with this
inconsistency problem is to find a valid instrument variable. Instead, we use the “mixed-fixed”
14
15
We do not report results to save space, and they are available upon request.
Results are available upon request.
17
effects model, in the spirit of Hsiao (1986, 2003). This model is in the following form (see
Greene, 2012, pp. 424-427):
(5)
We are treating
and
as random.
is a fixed-effect. Equation (5) can be estimated
consistently by ordinary least squares (OLS), but the estimates will be inefficient. To see why,
let’s write equation (5) in the following form:
),
where
=
(6)
. This equation has a heteroskedastic variance, i.e., var(
but generalized least squares (GLS) can produce efficient estimates. We use the
following three-step procedure:
1) Run the log
dummies*log(
) on county dummies, log minimum wage and log population, and county
.
2) Get the residuals and regress the squared values on a constant and the log minimum
wage-squared and log population-squared to estimate
and
.
3) Finally, re-estimate the model described in the step (1) using weighted least squares, with
the inverse of
as a weight.
This procedure produces consistent and efficient estimates (Greene, 2014).
The results from this exercise, using the full sample period, are reported in Table 6. We find
stronger earning and employment effects of the minimum wage here than we see in the
18
specifications reported in previous tables for all three groups. Earnings and employment for 14 to
18 year olds are much more responsive to the level of the minimum wage than is the case for the
other age groups. However, for all three groups there seems to be quite similar worker turnover
responses to the minimum wage level with estimates of the accession elasticity, the separation
elasticity and the turnover rate elasticity in the neighborhood of -.30. Finally, we also find
statistically significant negative responses of job creation by expanding and new establishments
to the minimum wage level, with this effect increasing in strength as we move from the youngest
to the oldest age group in our sample.
We also estimate the mixed/fixed model over the subsample from the first quarter of 2004 to
the fourth quarter 2007 for reasons explained above. Table 7 contains these results, which are
qualitatively similar to the full sample’s results reported in Table 6. Removing the recessions that
bracket our full sample from the data reduces the estimated minimum wage responses of
earnings and employment for the youngest workers and the turnover responses for the older
groups.
VII. Conclusions
With myriad empirical studies producing wide ranging results, varying by time, place and
worker group, there appears to be no consensus empirical answer to the question of the impact of
the minimum wage on young and other affected workers. Yet an understanding of the magnitude
of these empirical is critical for the evaluation of legislative proposals for a higher minimum
wage. Our study adds two main conclusions to this analytical mix. First, even with controls for
spatial heterogeneity we find that the youngest workers in local labor markets with higher
minimum wage levels experience the most pronounced positive earnings and negative
19
employment effects. As is often seen in the literature, these effects are not always found for older
workers in the youth labor force we study.
The second result is that higher minimum wage levels are associated with substantially lower
worker turnover. Separations from quits and layoffs are smaller and so is the number of hires in a
given period. We have interpreted these results in the context of a matching model framework
whereby reduced separations create proportionately fewer job searchers and job vacancies, which
in turn helps explain the lower number of accessions. This worker turnover effect has been found
by the other studies that have examined labor market flows so for the moment we might argue
that there is a consensus empirical result that a higher minimum wage is related to lower worker
turnover. Reduced turnover would be a positive outcome for many employers of minimum wage
workers who are forced to deal with very high turnover rates. The benefits would be even greater
if lower turnover were accompanied by improved quality of job seekers (Giuliano, 2013) and
greater productivity (Hirsch et al., 2011). The impacts on the target population of minimum wage
policy are more nuanced and depend on the extent of such composition effects on the incidence
of earnings enhancement and job displacement.
20
References
Abowd, John M., Bryce E. Stephens, Lars Vilhuber, Fredrik Andersson, Kevin L.
McKinney, Marc Roemer, and Simon Woodcock. 2005. The LEHD infrastructure files and the
creation of the quarterly workforce indicators. Technical Paper No. TP-2006-01. U.S. Census
Bureau.
Acemoglu, Daron. 2001. Good jobs versus bad jobs. Journal of Labor Economics. 19(1):
1-22.
Addison , John T., McKinley L. Blackburn, and Chad D. Cotti. 2009. Do minimum
wages raise employment? Evidence from the U.S. retail-trade sector. Labor Economics, 16(4):
397-408.
Allegretto, Sylvia A., Arindrajit Dube, and Michael Reich. 2011. Do minimum wages
really reduce teen employment? Accounting for heterogeneity and selectivity in state panel data.
Industrial Relations, 50(2): 205-240.
Bhaskar, Venkatarman, Alan Manning and Ted To. 2002. Oligopsony and monopsonistic
competition in labor markets. The Journal of Economic Perspectives. 16(2): 155-174.
Brochu, Pierre and David A. Green. 2013. The impact of minimum wages on labour
market transitions. The Economic Journal, 123: 1203-1235.
Brown, Charles, Curtis Gilroy, and Andrew Kohen. 1982. The effect of the minimum
wage on employment and unemployment. Journal of Economic Literature, 20: 487-528.
Burdett, Kenneth, and Dale T. Mortensen. 1998. Wage differentials, employer size, and
unemployment. International Economic Review. 39(2): 257-273.
Burkhauser, Richard V., Kenneth A. Couch, and David C. Wittenburg. 2000. A
reassessment of the new economics of the minimum wage literature with monthly data from the
21
Current Population Survey. Journal of Labor Economic, 18(4): 653-680.
Card, David and Alan B. Krueger. 1994. Minimum wages and employment: A case study
of the fast-food industry in New Jersey and Pennsylvania. American Economic Review, 84(4):
772-793.
Card, David and Alan B. Krueger. 1995. Myth and Measurement: The New Economics of
the Minimum Wage. Princeton, NJ: Princeton University Press.
Davis, Stephen J., and John Haltiwanger. 2014. Labor market fluidity and economic
performance, Working Paper 20479. National Bureau of Economic research.
Dube, Arindrajit, T. William Lester, and Michael Reich. 2010. Minimum wage effects
across state borders: Estimates using contiguous counties. Review of Economics and Statistics,
92(4): 945-64.
Dube, Arindrajit, T. William Lester, and Michael Reich. 2011. Do frictions matter in the
labor market? Accessions, separations and minimum wage effects. Discussion Paper No. 5811.
Institute for the Study of Labor.
Dube, Arindrajit, T. William Lester, and Michael Reich. 2013. Minimum Wage Shocks,
Employment Flows and Labor Market Frictions. Working Paper 149-13. Institute for Research
on Labor and Employment. UC Berkeley
Flinn, Christopher J. 2006. Minimum wage effects on labor market outcomes under
search, matching, and endogenous contact rates. Econometrica.4:1013-1062.
Gittings, R. Kaj, and Ian M. Schmutte. 2014. Getting handcuffs on an octopus: minimum
wages, employment and turnover. Available at SSRN: http://ssrn.com/abstract=2234118.
Giuliano, Laura. 2013. Minimum wage effects on employment, substitution, and the
teenage labor supply: evidence from personnel data. Journal of Labor Economics. 31(1): 155-
22
194.
Groshen, Erica L., and Simon Potter. 2003. Has structural change contributed to a jobless
recovery? Current Issues in Economics and Finance. 9(8): 1-7.
Greene, William. 2014. Dynamic Models, Time Series, Panels and Nonstationary Data
[PowerPoint slides]. Accessed October 20, 2014, from http://people.stern.nyu.edu/wgreene
/Econometrics/PanelDataNotes.htm
Greene, William. H. 2012. Econometric Analysis -7th Edition (pp. 424-426). Prentice
Hall, Upper Saddle River, NJ .
Hsiao, C. 1986. Analysis of Panel Data, Econometric Society monographs No. 11, New
York: Cambridge University Press.
Hsiao, C. 2003. Analysis of Panel Data, 2nd edition, Cambridge: Cambridge University
Press (Econometric Society monograph no. 34).
Hirsch, Barry, Bruce E. Kaufman, and Tetyana Zelenska. 2011. Minimum Wage Channels
of Adjustment. Forthcoming in: Industrial Relations, 2015.Ihlanfeldt, Keith R., and David L.
Sjoquist. 1990. Job accessibility and racial differences in youth employment rates. American
Economic Review, 80(1): 267-76.
Kalenkoski, Charlene M., and Donald J. Lacombe. 2011. Minimum wages and teen
employment: A spatial panel approach. Discussion Paper No. 5933. Institute for the Study of
Labor.
Meer, Jonathan , and Jeremy West. 2013. Effects of the minimum wage on employment
dynamics. Working Paper 19262. National Bureau of Economic Research.
Neumark, David and William L. Wascher. 2008. Minimum Wages. Cambridge, MA: MIT
Press.
23
Neumark, David; Ian J. M. Salas, and William L. Wascher. 2014. Revisiting the minimum
wage-employment debate: Throwing out the baby with the bathwater? Industrial and Labor
Relations Review. 64:608-648.
Petrongolo, Barbara, and Christopher A. Pissarides. 2001. Looking into the black box: a
survey of the matching function. Journal of Economic Literature. 39(2): 390-431.
Portugal, Pedro and Ana Rute Cardoso. 2006. Disentangling the minimum wage puzzle:
an analysis of wage puzzle: An analysis of worker accessions and separations. Journal of the
European Economic Association, 4(5): 9988-1013.
Sabia, Joseph. 2009. The effects of minimum wage increases on retail employment: New
evidence from monthly CPS data. Journal of Labor Research, 30(1): 75-97.
Sabia, Joseph J., Richard V. Burkhauser, and Benjamin Hansen. 2012. Are the effects of
minimum wage increases always small? New evidence from a case study of New York State.
Industrial and Labor Relations Review, 65(2): 350-376.
Stoll, Michael A. 1999. Spatial mismatch, discrimination, and male youth in the
Washington, DC area: Implications for residential mobility policies. Journal of Policy Analysis
and Management, 18(1): 77-98.
Thompson, Jeffrey P. 2009. Using local labor market data to re-examine the employment
effects of the minimum wage. Industrial and Labor Relations Review, 62(3): 343-66.
24
Table 1
State Minimum Wages
State or Other
Jurisdiction
Region
Federal
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
$5.15
$5.15
$5.15
$5.15
$5.15
$5.15
$5.85
$6.55
$5.15
$5.15
Maine
New England
5.15
5.15
New England
5.65
6.15
6.25
6.15
6.75
6.35
6.75
6.50
6.75
6.75
Rhode Island
5.75
6.15
6.25
7.40
7.00
7.40
7.25
7.40
Vermont
New England
5.75
6.25
6.25
6.25
5.65
6.15
6.15
6.65
Mideast
5.15
5.15
5.15
5.15
5.15
5.15
6.15
7.15
6.15
8.06
7.15
Maryland
6.15
5.15
7.25
6.15
7.68
Mideast
7.00
6.15
7.53
Delaware
6.75
6.15
New Jersey
Mideast
5.15
5.15
5.15
5.15
5.15
5.15
6.15
7.15
7.15
7.15
New York
Mideast
5.15
5.15
5.15
5.15
7.15
5.15
5.15
5.15
5.15
6.75
5.15
7.15
Mideast
6.00
5.15
7.15
Pennsylvania
5.15
5.15
7.15
7.15
Illinois
Great Lakes
5.15
5.15
5.15
5.15
5.15
5.15
5.15
5.15
6.50
5.15
7.75
Great Lakes
6.50
5.15
7.50
Indiana
5.50
5.15
6.25
6.50
5.15
5.85
6.55
Michigan
Great Lakes
5.15
5.15
5.15
5.15
5.15
5.15
5.15
6.95
7.15
7.40
Ohio
Great Lakes
5.15
5.15
5.15
5.15
5.15
5.15
5.15
6.85
7.30
Wisconsin
Great Lakes
5.15
5.15
5.15
5.15
5.15
5.15
Iowa
Plains
5.15
5.15
5.15
5.15
5.15
5.15
5.70
5.15
6.50
5.15
7.00
6.50
7.25
6.55
7.25
Kansas
Plains
5.15
5.15
5.15
5.15
5.15
5.15
5.15
5.15
5.85
6.55
Minnesota
Plains
5.15
5.15
5.15
5.15
5.15
5.15
5.25
5.85
6.55
Missouri
Plains
5.15
5.15
5.15
5.15
5.15
5.15
5.25
5.15
6.65
7.05
Nebraska
Plains
5.15
5.15
5.15
5.15
5.15
5.15
5.15
6.50
5.15
5.85
6.55
North Dakota
Plains
5.15
5.15
5.15
5.15
5.15
5.15
5.15
5.15
5.85
6.55
South Dakota
Plains
5.15
5.15
5.15
5.15
5.15
5.15
5.15
5.15
5.85
6.55
Alabama
Southeast
5.15
5.15
5.15
5.15
5.15
5.15
5.15
5.15
6.55
Arkansas
Southeast
5.15
5.15
5.15
5.15
5.15
5.15
5.15
6.25
5.85
6.25
Florida
Southeast
5.15
5.15
5.15
5.15
5.15
5.15
5.15
5.15
5.15
5.15
5.15
5.15
6.67
5.15
7.21
Southeast
6.40
5.15
6.79
Georgia
5.85
6.55
6.55
6.55
25
Kentucky
Southeast
5.15
5.15
5.15
5.15
5.15
5.15
5.15
5.15
5.85
6.55
Louisiana
Southeast
5.15
5.15
5.15
5.15
5.15
5.15
5.15
5.15
5.85
6.55
Mississippi
Southeast
5.15
5.15
5.15
5.15
5.15
5.15
5.15
5.15
5.85
6.55
South Carolina
Southeast
5.15
5.15
5.15
5.15
5.15
5.15
5.15
5.15
5.85
6.55
Tennessee
Southeast
5.15
5.15
5.15
5.15
5.15
5.15
5.15
5.15
5.85
6.55
Virginia
Southeast
5.15
5.15
5.15
5.15
5.15
5.15
5.15
5.15
5.85
6.55
West Virginia
Southeast
5.15
5.15
5.15
5.15
5.15
5.15
5.15
5.85
6.55
7.25
Arizona
Southwest
5.15
5.15
5.15
5.15
5.15
5.15
5.15
6.90
7.25
New Mexico
Southwest
5.15
5.15
5.15
5.15
5.15
5.15
6.50
7.50
Oklahoma
Southwest
5.15
5.15
5.15
5.15
5.15
5.15
6.75
5.15
5.15
5.15
5.15
5.85
6.55
Texas
Southwest
5.15
5.15
5.15
5.15
5.15
5.15
5.15
5.85
6.55
Colorado
Rocky Mountain
5.15
5.15
5.15
5.15
5.15
5.15
5.15
5.15
7.02
7.28
Idaho
Rocky Mountain
5.15
5.15
5.15
5.15
5.15
5.15
5.15
6.85
5.15
5.85
6.55
Montana
Rocky Mountain
5.15
5.15
5.15
5.15
5.15
5.15
5.15
5.15
5.85
6.55
Utah
Rocky Mountain
5.15
5.15
5.15
5.15
5.15
5.15
5.15
5.15
5.85
6.55
Wyoming
Rocky Mountain
5.15
5.15
5.15
5.15
5.15
5.15
5.15
Alaska
Far West
5.65
5.65
5.15
5.65
7.15
7.15
7.15
7.15
6.55
7.15
California
Far West
5.75
6.75
6.75
6.75
6.75
7.50
Far West
5.25
6.25
5.15
5.15
5.15
5.15
6.75
5.15
7.25
Far West
6.25
5.15
6.25
Nevada
5.75
5.15
8.00
7.25
8.00
Hawaii
6.25
5.25
7.15
6.75
5.85
7.15
6.15
6.33
6.55
Oregon
Far West
6.50
6.50
6.50
6.90
7.05
7.25
7.50
7.80
7.95
8.40
Washington
Far West
6.50
6.72
6.90
7.01
7.16
7.35
7.63
7.93
8.07
8.55
7.25
Note: The minimum wage data are reported as the higher of the federal or state level for 46 states, whose data on labor market outcome are also available
in the QWI.
Source: Changes in Basic Minimum Wages in Non-farm Employment under State Law: Selected Years 1968 to 2010, U.S. Department of Labor, Office of
State Standards Programs Wage and Hour Division web site Minimum Wage and Overtime Pay Standards Applicable to Nonsupervisory NONFARM
Private Sector Employment under State and Federal Laws. http://www.dol.gov/whd/state/stateMinWageHis.htm
26
Age Group
Panel A Fixed-Effect Model
ln (Minimum Wage)
Table 2
Minimum Wage Effects on Employment and Average Monthly Earnings: 2000-2009
14-18
19-21
22-24
(1)
(2)
(1)
(2)
(1)
(2)
ln(Employment)
ln(Eearnings)
ln(Employment)
ln(Eearnings)
ln(Employment)
ln(Eearnings)
0.127**
[0.057]
-0.147**
[0.066]
0.232***
[0.025]
102,924
0.417
0.027
[0.056]
0.198***
[0.069]
0.965***
[0.053]
102,697
0.426
-0.081
[0.069]
-0.334***
[0.060]
0.246***
[0.025]
102,968
0.341
0.030
[0.054]
0.042
[0.059]
1.078***
[0.043]
102,844
0.464
-0.108*
[0.058]
-0.284***
[0.056]
0.236***
[0.020]
103,010
0.402
Panel B Fixed Effect Model with EA × Time Fixed Effect
ln (Minimum Wage)
-0.173***
0.209***
[0.047]
[0.030]
ln (Population)
0.354***
-0.185***
[0.080]
[0.041]
-0.046
[0.059]
0.180**
[0.075]
0.083***
[0.028]
-0.281***
[0.035]
0.073
[0.054]
0.125**
[0.055]
0.028
[0.029]
-0.253***
[0.037]
0.916***
[0.050]
102,697
0.994
0.204***
[0.023]
102,968
0.710
1.033***
[0.040]
102,844
0.996
0.205***
[0.020]
103,010
0.761
ln (Population)
ln(Employ. Ages 25-65)
Observations
R-squared
ln(Employ. Ages 25-65)
Observations
Adj. R-squared
-0.230***
[0.067]
0.314***
[0.057]
0.672***
[0.048]
102,630
0.494
0.589***
[0.040]
102,630
0.992
0.194***
[0.024]
102,924
0.755
Notes: The estimates in Panel A are calculated using equation (2) and that in Panel B are using the two-way fixed effects model with economic-areaspecific time effects, equation (3). Standard errors are reported in parentheses.
* Significant at the 10 percent level.
** Significant at the 5 percent level.
*** Significant at the 1 level.
27
Table 3
Minimum Wage Effects on Labor Market Flows: 2000-2009
(1)
(2)
(3)
(4)
ln(Accessions)
ln(Separations)
ln(Turnover)
ln(Job Creation)
Panel A Age 14-18
ln(Minimum Wage)
Observations
Adj. R-squared
Panel B Age 19-21
ln(Minimum Wage)
Observations
Adj. R-squared
Panel C Age 22-24
ln(Minimum Wage)
Observations
Adj. R-squared
(5)
ln(Job Loss)
(6)
ln(Job Relocation)
-0.138**
[0.058]
101,503
0.550
-0.199***
[0.062]
102,006
0.599
-0.161***
[0.057]
101,196
0.598
0.020
[0.053]
101,361
0.547
-0.017
[0.052]
100,944
0.577
0.009
[0.042]
101,687
0.495
-0.150**
[0.061]
101,546
0.359
-0.178***
[0.063]
102,200
0.448
-0.159**
[0.061]
101,317
0.425
-0.093
[0.061]
101,440
0.411
-0.141**
[0.061]
101,423
0.455
-0.091*
[0.054]
101,811
0.340
-0.155**
[0.064]
101,482
0.269
-0.148**
[0.064]
102,130
0.252
-0.149**
[0.063]
101,223
0.293
-0.130**
[0.063]
101,628
0.138
-0.112*
[0.058]
101,557
0.175
-0.109*
[0.057]
101,969
0.176
Notes: The estimates are calculated using the two-way fixed effects model, equation (2). Standard errors are reported in parentheses.
* Significant at the 10 percent level.
** Significant at the 5 percent level.
*** Significant at the 1 level.
Table 4
Minimum Wage Effects on Labor Market Flows Controlling for Economic Area Time Effects: 2000-2009
(1)
(2)
(3)
(4)
(5)
ln(Accessions)
ln(Separations)
ln(Turnover)
ln(Job Creation)
ln(Job Loss)
Panel A Age 14-18
ln(Minimum Wage)
Observations
Adj. R-squared
Panel B Age 19-21
ln(Minimum Wage)
Observations
Adj. R-squared
Panel C Age 22-24
ln(Minimum Wage)
Observations
Adj. R-squared
(6)
ln(Job Relocation)
-0.115*
[0.062]
101,503
0.755
-0.163**
[0.069]
102,006
0.744
-0.139**
[0.064]
101,196
0.777
0.004
[0.052]
101,361
0.652
0.039
[0.078]
100,944
0.667
0.004
[0.057]
101,687
0.695
-0.096*
[0.051]
101,546
0.642
-0.149***
[0.047]
102,200
0.646
-0.122**
[0.049]
101,317
0.677
-0.015
[0.068]
101,440
0.601
-0.102*
[0.060]
101,423
0.622
-0.058
[0.059]
101,811
0.691
-0.138**
[0.054]
101,482
0.600
-0.136**
[0.053]
102,130
0.542
-0.135**
[0.054]
101,223
0.627
-0.122*
[0.072]
101,628
0.461
-0.086
[0.069]
101,557
0.409
-0.096
[0.067]
101,969
0.592
Notes: The estimates are calculated a fixed-effect model with BEA-Economic Areas (EA)-specific time effects model, equation (3). Standard errors are
reported in parentheses.
* Significant at the 10 percent level.
** Significant at the 5 percent level.
*** Significant at the 1 level.
29
(1)
ln(Empl.)
Panel A Age 14-18
ln(min. wage)
Observations
Adj. R-squared
Panel B Age 19-21
ln(min. wage)
Observations
Adj. R-squared
Panel C Age 22-24
ln(min. wage)
Observations
Adj. R-squared
Table 5
Minimum Wage Effects Controlling for Economic Area Time Effects: 2004-2007
(2)
(3)
(4)
(5)
(6)
ln(Earnings)
ln(Acces.)
ln(Separ.)
ln(Turnovers)
ln(Job Creat.)
(7)
ln(Job Loss)
(8)
ln(Job Realoc.)
-0.139***
[0.030]
47,059
0.994
0.180***
[0.026]
47,191
0.785
-0.112***
[0.034]
46,853
0.761
-0.151***
[0.042]
46,755
0.750
-0.125***
[0.034]
46,701
0.785
0.017
[0.052]
46,795
0.650
0.082
[0.069]
46,571
0.667
0.047
[0.046]
46,951
0.702
-0.059*
[0.033]
47,081
0.996
0.019
[0.026]
47,199
0.743
-0.038
[0.043]
46,872
0.653
-0.124**
[0.049]
46,849
0.653
-0.079*
[0.041]
46,765
0.688
0.079
[0.065]
46,839
0.608
-0.070
[0.076]
46,829
0.629
-0.002
[0.043]
47,001
0.704
0.052*
[0.028]
47,163
0.997
-0.017
[0.022]
47,227
0.788
-0.103***
[0.033]
46,848
0.613
-0.077**
[0.035]
46,843
0.555
-0.091***
[0.033]
46,737
0.645
-0.111**
[0.049]
46,917
0.467
-0.007
[0.061]
46,895
0.420
-0.057
[0.048]
47,085
0.612
Notes: The estimates are calculated a fixed-effect model with BEA-Economic Areas (EA)-specific time effects model, equation (3), using the sample
from 2004 to 2007. Standard errors are reported in parentheses.
* Significant at the 10 percent level.
** Significant at the 5 percent level.
*** Significant at the 1 level.
30
(1)
ln(Empl.)
Panel A Age 14-18
ln(min. wage)
Observations
Adj. R-squared
Panel B Age 19-21
ln(min. wage)
Observations
Adj. R-squared
Panel C Age 22-24
ln(min. wage)
Observations
Adj. R-squared
Table 6
Minimum Wage Effects in the Mixed/Fixed Effect Model: 2000-2009
(2)
(3)
(4)
(5)
(6)
ln(Earnings)
ln(Acces.)
ln(Separ.)
ln(Turnovers)
ln(Job Creat.)
(7)
ln(Job Loss)
(8)
ln(Job Realoc.)
-0.421***
[0.033]
99,545
0.985
0.513***
[0.036]
99,914
0.606
-0.319***
[0.042]
98,205
0.298
-0.327***
[0.041]
98,677
0.189
-0.301***
[0.038]
97,794
0.273
-0.140**
[0.061]
98,051
0.124
0.043
[0.066]
97,354
0.0834
With th-0.031
[0.031]
98,583
0.267
-0.156***
[0.032]
99,659
0.992
0.280***
[0.033]
99,976
0.664
-0.292***
[0.042]
98,325
0.371
-0.291***
[0.045]
98,977
0.199
-0.270***
[0.040]
98,031
0.337
-0.232***
[0.055]
98,124
0.226
-0.087
[0.061]
98,107
0.159
-0.091**
[0.035]
98,755
0.456
-0.045**
[0.021]
99,829
0.996
0.260***
[0.029]
100,039
0.749
-0.353***
[0.042]
98,211
0.422
-0.302***
[0.041]
98,834
0.328
-0.310***
[0.039]
97,887
0.441
-0.264***
[0.045]
98,348
0.323
-0.079
[0.048]
98,236
0.226
-0.144***
[0.036]
98,929
0.491
Notes: The estimates are calculated using the mixed/fixed effect model in the specification of equation (4). Standard errors are reported in parentheses.
* Significant at the 10 percent level.
** Significant at the 5 percent level.
*** Significant at the 1 level.
31
(1)
ln(Empl.)
Panel A Age 14-18
ln(min. wage)
Observations
Adj. R-squared
Panel B Age 19-21
ln(min. wage)
Observations
Adj. R-squared
Panel C Age 22-24
ln(min. wage)
Observations
Adj. R-squared
Table 7
Minimum Wage Effects in the Mixed/Fixed Effect Model: 2004-2007
(2)
(3)
(4)
(5)
(6)
ln(Earnings)
ln(Acces.)
ln(Separ.)
ln(Turnovers)
ln(Job Creat.)
(7)
ln(Job Loss)
(8)
ln(Job Realoc.)
-0.243***
[0.044]
44,052
0.986
0.304***
[0.026]
44,217
0.615
-0.302***
[0.061]
43,760
0.288
-0.281***
[0.069]
43,648
0.152
-0.287***
[0.064]
43,562
0.255
-0.215***
[0.056]
43,688
0.0913
0.013
[0.095]
43,348
0.0354
-0.123**
[0.058]
43,938
0.245
-0.134***
[0.046]
44,096
0.992
0.155***
[0.029]
44,228
0.663
-0.191***
[0.055]
43,820
0.361
-0.188***
[0.052]
43,796
0.171
-0.190***
[0.051]
43,680
0.324
-0.137**
[0.065]
43,746
0.198
-0.090
[0.075]
43,737
0.124
-0.110**
[0.052]
44,012
0.452
-0.059**
[0.028]
44,193
0.997
0.143***
[0.023]
44,269
0.743
-0.215***
[0.057]
43,774
0.420
-0.192***
[0.061]
43,768
0.323
-0.199***
[0.058]
43,649
0.448
-0.178***
[0.055]
43,836
0.310
-0.101
[0.069]
43,788
0.221
-0.135**
[0.057]
44,102
0.516
Notes: The estimates are calculated using the mixed/fixed effect model in the specification of equation (4), using the sample from 2004-2007. Standard
errors are reported in parentheses.
* Significant at the 10 percent level.
** Significant at the 5 percent level.
*** Significant at the 1 level.
32