academic paper decoding

make sense of and present within 10 minutes. I need your help to extract the essence of the paper, with explanation examples to drive these core ideas in succinct way. I need some graphical representation of data derived from this paper, to put into my presentation slides… especially Section III of this paper with the equations, etc… The point of this presentation is to explain complex concepts in lament terms to audience that is learning about international economics at MBA level (but not majoring in Economics… this is the only economics class). Goal here is for me to clearly explain these complex concepts in easy to understand terms to students that focuses in the consumer goods sector. If this can take on a “fun” tone that would be great for speeches at this dry level.
Immigration and Domestic Wage:
An Empirical Study on Competition among Immigrants
DRAFT – DO NOT QUOTE
Chong-Uk Kim
Department of Economics
Sonoma State University
1801 E. Cotati Avenue
Rohnert Park, CA 94928-3609 USA
Email: [email protected]
I. Introduction
The dropping of the Mayflower’s anchor at Cape Cod in 1620 marks the
beginning of the United States’ long history of immigration. Following this event,
immigration, through its massive and successive waves, has carved the image of the
United States. Due to its concordance with domestic economic issues, immigration policy
continues to be a center of social debate. This is evident in the frequent headlining of the
President’s actions on immigration in newspapers all over the country.
The main concerns regarding immigration reform center around the expectation,
or fear, that new immigrants entering a domestic labor market will replace and take
opportunity away from native workers while decreasing domestic wages. These concerns
have spurred many research efforts that provide a wide range of empirical findings. For
example, some studies such as Borjas (2003) and Mishra (2007) find that immigration
has a significant impact on wages in both receiving and sending countries while other
studies such as Card (2005) and Ottaviano and Peri (2012) show that there is no
meaningful impact of immigration on domestic labor markets.
While studies on the wage effects of immigration focus on native workers, there is
significantly less information on the wage effects of immigration on domestic foreignborn
workers. In addition to analyzing the impact of immigration on wages of native
workers, in this paper, we estimate the internal competition amongst foreign-born
workers in the United States. Firstly, using data from the Current
Population
Survey
(CPS),
we
find
no
empirical
evidence
supporting the substitutability of native workers
immigrants. Secondly, there is no statistical difference between skilled and unskilled
immigrants on the influence of the domestic labor market outcomes. Lastly, there is no
internal competition among immigrants. The income of non-citizen workers mainly
depends on state and national levels of economic situations, not the number of noncitizen
workers available in the labor market.
The paper proceeds as follows. Section II describes the literature on the effects of
immigration on domestic wages. Section III provides the empirical methodology and the
data used in this analysis. Section IV discusses empirical results and Section V
concludes.
II. Literature Review
The main focus of the majority of immigration literature is to empirically prove if
immigration suppresses domestic wages. In other words, primary to these studies is the
degree of substitutability of immigrants for domestic native workers. The general
conclusion has been that a 10 percent increase in the population due to immigration
decreases wages of native workers by 1 to 4 percent.1 Using US census data, Altonji and
Card (1991) measure the impact of immigration on wages of unskilled native workers;
their findings indicate that a 10 percent increase in the population due to immigration
lowers wages of unskilled native workers by 1.2 percent even though the effects of
immigration on wages heavily depend on native subgroup.
Immigration supporters most frequently cite Card’s famous 1990 paper on Cuban
boat people. In 1980, more than 125,000 unskilled Cubans immigrated to Miami. Card
finds no empirical evidence to support the theory that these immigrants lower domestic
wages or increase unemployment rates of either Cubans or non-Cubans. Butcher and
Card (1991), Card and DiNardo (2000), Card (2001, 2005) and more recently Henrickson
and Kim (2012) also find no empirical evidence supporting a strong degree of
substitutability of immigrants for native workers.
In contrast, a series of publications by Borjas (1995, 2003, and 2006), report
empirical evidence that an influx of immigrants decreases wages of native workers and
affects domestic labor market outcomes. Immigration critics frequently cite Borjas’s
famous 2003 paper that treats immigration as an increase in national labor supply. This
study finds that a 10 percent increase in the population due to immigration suppresses
wages of native workers by 3 to 4 percent.
While the majority of studies on immigration have focused on the supply side of
labor market, Lach (2007) suggests that immigrants not only increase domestic labor
1
Examples
of
such
results
are
numerous,
but
include:
Friedberg
and
Hunt
(1995),
Borjas
(2003),
Borjas,
Grogger,
and
Hanson
(2010).
supply, but also increase the demand for domestic output. Companies must hire more
labor to supply more output; therefore, immigrants eventually shift both labor supply and
demand curves outward to the right. More recently, without perfect substitutability, Peri
(2012) and Ottaviano and Peri (2012) show that immigrants do not replace native
workers and there is no short-run effect on wages of native workers.
III. Empirical Model and Data
1. Empirical Model
Our empirical model which we use to test the main hypothesis is based on the
standard Mincer earning equation. Since Jacob Mincer published his seminal book
Schooling, Experience, and Earnings in 1974, many labor economists have used his
earning equation as a key empirical framework. In the standard Mincer equation, your
earnings depend on your years of schooling and working experience.
(1) ln(Y) = a + ß1S + ß2E + ß3E2 + u
where
Y
is
earnings,
S
is
years
of
schooling,
and
E
is
years
of
working
experience.2
This
equation
(1)
has
been
used
as
a
starting
point
of
many
empirical
economic
papers
on
income
determination.
Based
on
the
individual
research
purposes,
this
standard
equation
has
been
modified
and
tested
in
many
academic
papers.3
To
test
our
main
hypothesis,
we
modify
Equation
(1)
to
implement
our
model.
Since
the
Current
Population
Survey
(CPS)
does
not
report
information
on
the
working
experience,
first,
we
use
an
age
variable
as
a
proxy
for
the
working
experience.
Second,
instead
of
using
a
years
of
schooling
variable,
we
use
a
ratio
of
high-­-skilled
and
low-­-skilled
workers
to
capture
the
impact
of
education
on
income.4
Third,
we
2
a
is
a
constant
term.
It
is
the
logarithm
of
the
income
level
with
no
years
of
schooling
and
no
working
experience.
u
is
a
Gaussian
white
noise
error
term.
3
Lemieux (2006) provides a nice summary of works using Mincer equations.
4
Details
on
this
variable
will
be
given
in
the
next
section.
include
a
state-­-level
unemployment
rate
variable
to
reflect
each
of
US
states
economic
situation.
Similarly,
we
add
a
national-­-level
real
Gross
Domestic
Product
(GDP)
variable
to
consider
the
US
national
economy.
Finally,
we
include
the
total
number
of
foreign
workers
which
is
our
key
variable
to
see
the
effects
of
immigration
on
domestic
wages.
(2)
Yit
=
a(FOREIGNit)ß(GDPt)?exp(dRATIOit+?UNEMit+?AGEit+?AGE2it+uit)
where
Yit
denotes
the
real
income
of
US
citizens
of
state
i
in
period
t;
FOREIGNit
denotes
the
total
number
of
foreign
workers
of
state
i
in
period
t;
GDPt
denotes
the
real
GDP
of
US
in
period
t;
RATIOit
denotes
the
ratio
of
high-­-skilled
and
low-­-skilled
US
citizens
of
state
i
in
period
t;
UNEMit
denotes
the
unemployment
rate
of
state
i
in
period
t;
AGEit
denotes
the
median
age
of
US
citizens
of
state
i
in
period
t;
AGE2it
denotes
the
square
of
AGEit
variable;
and
finally
uit
is
a
Gaussian
white
noise
error
term.
We take natural logs on both sides of equation (2) to implement our income equation and
it provides us the following estimation equation:
(3) ln(Yit)
=
C
+
ß
ln(FOREIGNit)
+
?
ln(GDPt)
+
d
(RATIOit)
+
?
(UNEMit)
+
?
(AGEit)
(+/-­-)
(+)
(+)
(-­-)
(+)
+
?
(AGE2it)
+
uit
5
(-)
5
C
=
ln(a)
which
is
a
constant
term.
Based on findings from previous research efforts, we expect that the coefficients of GDPt,
RATIOit,
and
AGEit
variables
are
positive
while
the
coefficients
of
UNEMit
and
AGE2it
variables
are
negative.
A
positive
coefficient
of
our
key
variable,
FOREIGNit,
will
support
the
complementarity
idea
between
immigrants
and
citizen.
Similarly,
a
negative
coefficient
of
FOREIGNit
will
uphold
the
idea
that
immigrants
and
citizens
are
substitutes.
2. Data
This research uses data mainly drawn from the Current Population Survey.6 The
CPS provides a variety of labor force statistics for the population of the United States.
We extract our main variables such as income, educational attainment, and citizenship
status from CPS. Information on the consumer price index (CPI) and gross domestic
product (GDP) come from the World Bank.7 We also use the StateData.info website to
collect data on U.S. state level unemployment rate.8
Since
we
are
testing
the
impacts
of
immigration
on
the
US
state-­-level
income,
the
final
sample
for
our
empirical
tests
consists
of
data
on
51
US
states
including
the
District
of
Columbia
(DC).
Our
data
set
is
strongly
balanced
and
comprises
16
years
from
1995
to
2010.
Table
I
presents
descriptive
statistics
for
our
data
set.
Since
we
want
to
investigate
the
impact
of
immigration
on
the
US
citizens’
income
at
the
state
level,
we
modify
the
CPS
data
set
to
fit
our
research
purposes.
To
attain
the
data
set
including
information
on
US
citizens
only,
first,
we
separate
the
CPS
data
set
into
two
groups,
US
citizens
and
non-­-citizens.
From
the
individual
observations,
second,
we
get
the
average
income
per
each
state
including
DC
for
both
US
citizens
and
non-­-citizens.9
Similarly,
the
average
age
variable
per
each
state
is
obtained
in
the
same
way.
6
http://www.census.gov/cps/
7
http://data.worldbank.org/
8
http://www.statedata.info/
9
The
average
incomes
are
measured
in
constant
2005
dollars.
Table
I.
Descriptive
Statistics
Variable
Obs
Mean
Std.
Dev.
Min
Max
Y
(Real
Income)
816
41,845.56
6376.732
28393.09
64327.52
FOREIGN
816
218.8517
410.5372
5
3,320
GDP
(M)
816
223,094
270,254.3
14,211
1,768,607
RATIO
816
.4360832
.1544683
.1972789
1.683784
UNEM
816
5.363603
1.857371
2.2
14.9
AGE
816
40.53957
.9966217
37.29797
43.22313
AGE2
816
1644.449
80.60823
1391.139
1868.239
Like
our
income
variable
(Y),
The
total
number
of
foreign
workers
per
state
also
comes
from
the
CPS
data
set.
We
count
the
number
of
non-­-citizen
individuals
who
are
working
in
the
US
by
states
and
year.
Table
II
shows
our
sample
number
of
non-­-citizen
workers
at
the
state
level.
Table
II.
The
Number
of
Foreign
Worker
in
2010
State
Foreign
State
Foreign
State
Foreign
State
Foreign
AK
131
ID
75
MT
23
RI
271
AL
50
IL
509
NC
202
SC
68
AR
57
IN
71
ND
21
SD
58
AZ
204
KS
103
NE
141
TN
89
CA
3,320
KY
93
NH
149
TX
1,185
CO
259
LA
46
NJ
645
UT
131
CT
384
MA
234
NM
106
VA
292
DC
278
MD
488
NV
386
VT
56
DE
169
ME
66
NY
1,092
WA
293
FL
938
MI
174
OH
107
WI
124
GA
268
MN
221
OK
84
WV
16
HI
353
MO
64
OR
173
WY
47
IA
166
MS
25
PA
157
The
CPS
data
set
contains
information
on
personal
educational
attainment.
However
it
does
not
include
information
on
the
years
of
schooling.
Educational
attainments
are
originally
classified
by
academic
degrees
such
as
high
school
diploma,
associate
degree,
professional
degree,
and
etc.
Therefore,
to
quantify
this
information,
we
generate
the
RATIO
variable
which
represents
the
quality
of
labor
force
for
each
states.
Equation
4
displays
how
we
generate
the
RATIO
variable.
(4)
?????????? = !h! !”#$%& !” !”#$%” !h! !”# ! !”#h!” !”#$”” !h!” !”#h !”h!!” !”#$%&’
(!h! !”#$%& !” !”#$%” !h! !”# ! !”#h !”h!!” !”#$”” !” !”#$%)
According
to
our
calculation,
the
average
value
of
the
RATIO
variable
is
around
.44
over
the
sample
time
period.
It
means
that
there
are
44
high-­-skilled
workers
available
per
each
100
low-­-skilled
workers.
Table
III
shows
the
ratio
in
2010.
Table
III.
The
Ratio
of
High-­-Skilled
and
Low-­-Skilled
US
Citizens
in
2010
State
Ratio
State
Ratio
State
Ratio
State
Ratio
AK
.4
ID
.40
MT
.51
RI
.57
AL
.34
IL
.59
NC
.49
SC
.36
AR
.29
IN
.35
ND
.47
SD
.44
AZ
.44
KS
.54
NE
.56
TN
.43
CA
.61
KY
.39
NH
.65
TX
.43
CO
.80
LA
.42
NJ
.68
UT
.45
CT
.70
MA
.89
NM
.45
VA
.71
DC
1.68
MD
.75
NV
.33
VT
.60
DE
.46
ME
.52
NY
.65
WA
.56
FL
.51
MI
.47
OH
.37
WI
.50
GA
.55
MN
.54
OK
.38
WV
.38
HI
.46
MO
.36
OR
.54
WY
.34
IA
.39
MS
.28
PA
.47
IV. Empirical Results
The main purpose of this empirical study is to quantify the impact of non-­-citizen
workers
on
US
citizens’
income.
Our
empirical
results
satisfy
standard
Mincer
equation
expectation.
The
empirical
results
are
shown
in
Table
IV.
First
three
results
are
from
the
ordinary
least
squares
(OLS)
method.
Regression
IV.4
shows
the
results
of
our
model
including
state
fixed
effects
and
regression
IV.5
presents
the
results
from
random
effects
estimation.
Regression
IV.1
contains
only
basic
Mincer
variables
such
as
educational
attainment
of
population
(RATIO)
and
working
experiences
(AGE
and
AGE2).
In
addition
to
these
variables,
regression
IV.2
includes
information
on
each
states
economy
(UNEM)
and
national
economy
(GDP).
Regression
IV.3,
IV.4,
and
IV.5
contain
all
six
independent
variables,
which
accords
with
our
equation
(3).
All
coefficients
on
the
RATIO
variable
are
positive
and
statistically
significant,
which
confirms
that
educational
attainment
is
positively
correlated
with
US
citizens’
income.
Coefficients
on
AGE
and
AGE2
variables
also
satisfy
the
theoretical
prediction
and
coefficients
on
UNEM
and
the
logarithm
of
the
GDP
variables
empirically
suggest
that
US
citizens’
income
depends
on
the
state
and
national
economic
situation.
The
coefficients
on
our
key
variable,
the
logarithm
of
the
FOREIGN
variable,
turn
out
positive
and
statistically
significant.
It
suggests
that
the
number
of
non-­-citizen
workers
is
positively
correlated
with
US
citizens’
income.
Since
the
coefficients
on
the
logarithm
of
the
FOREIGN
variable
are
elasticities,
we
need
to
calculate
the
marginal
effects
of
non-­-citizen
workers
on
citizens’
income
to
see
the
economic
significance.
In
2010,
there
are
approximately
31.3
million
non-­-
citizens
living
in
the
US
and
the
nominal
GDP
per
capita
in
the
US
is
$48,387.10
The
coefficients
on
the
logarithm
of
the
FOREIGN
variable
have
a
range
from
0.011
to
0.046.
Therefore
if
we
increase
the
number
of
non-­-citizen
workers
by
313,000
which
is
1%
point
of
the
total
number
of
non-­-citizen
workers
in
2010,
then,
according
to
our
model
prediction,
the
nominal
GDP
per
capita
will
increase
by
from
10
According
to
the
Department
of
Homeland
Security,
there
are
19.7
million
legal
residents
and
approximately
11.6
million
illegal
immigrants
in
2010.
$5.32
to
$22.26.
Since
the
number
of
foreign
born
living
in
the
US
increases
approximately
500,000
annually,
these
additional
non-­-citizen
workers
raise
the
nominal
income
of
US
citizens
by
from
$8.5
to
$35.6
in
2010,
which
is
negligible.
Table
IV.
Immigration
and
Real
Income
ln(Real Income) IV.1 IV.2 IV.3 IV.4 IV.5
RATIO 0.675 0.619 0.51 0.304 0.414
(29.06)*** (31.05)*** (26.47)*** (6.10)*** (14.58)***
AGE 0.416 0.085 0.155 0.304 0.547
(2.06)** (0.49) (1.01) (2.34)** (4.95)***
AGE2 -0.005 -0.001 -0.002 -0.004 -0.006
(-1.94)* (-0.34) (-0.85) (-2.25)** (-4.72)***
UNEM -0.006 -0.008 -0.007 -0.009
(-3.66)*** (-5.51)*** (-6.91)*** (-7.84)***
ln(GDP) 0.052 0.017 0.21 0.062
(18.07)*** (4.94)*** (10.31)*** (7.60)***
ln(FOREIGN) 0.046 0.011 0.026
(14.65)** (2.06)** (6.41)***
Constant 1.443 7.532 6.407 1.617 -1.925
(0.35) (2.16)** (2.07)** (0.64) (-0.87)
Observations 816 816 816 816 816
R-squared 0.57 0.69 0.76 0.68
F-statistic 358.77 366.62 421.83 271.10
Wald ?2 1525.14
Note: Figures in parentheses are t-statistics.
***Significant at 1%; **significant at 5%; *significant at 10%.
To see the role of educational attainments of non-citizen workers in affecting
citizens’ income, we separate non-citizen workers by two different groups: Skilled and
Unskilled. Table V and VI display our empirical results. The L-FOREIGN variable
represents the number of non-citizen workers who has a high school diploma or less
while the H-FOREIGN variable indicates the number of non-citizen workers who has
higher educational attainments than a high school diploma. The empirical results from
OLS do not differ. Compare to regression V.1 which is the same regression as regression
IV.3 in Table IV, the coefficients
on
the
logarithm
of
the
L-­-FOREIGN
and
H-FOREIGN
variables do not show any significant differences. These results support that both skilled
and unskilled non-citizen workers are complements to US citizens even though the
economic impact is negligible. Regression V.5 contains F-RATIO variable representing
the educational attainments of non-citizen workers. The mean value of the F-RATIO
variable is .30, which means that there
are
30
high-­-skilled
workers
available
per
each
100
low-­-skilled
workers.
The
coefficient
on
the
F-RATIO variable is positive but
statistically not different from zero. In short, our empirical results from OLS suggest that
there is no heterogeneity between skilled and unskilled non-citizen workers in affecting
US citizens’ income.
Table
V.
Non-­-Citizen
Workers’
Educational
Attainments
(OLS)
ln(Real
Income)
V.1
V.2
V.3
V.4
V.5
RATIO
0.51
0.531
0.489
0.497
0.502
(26.47)***
(28.02)***
(23.47)***
(24.50)***
(24.71)***
AGE
0.155
0.128
0.192
0.165
0.166
(1.01)
(0.83)
(1.22)
(1.08)
(1.08)
AGE2
-­-0.002
-­-0.001
-­-0.002
-­-0.002
-­-0.002
(-­-0.85)
(-­-0.67)
(-­-1.08)
(-­-0.92)
-­-(0.93)
UNEM
-­-0.008
-­-0.008
-­-0.008
-­-0.008
-­-0.008
(-­-5.51)***
(-­-5.33)***
(-­-5.09)***
(-­-5.51)***
(-­-5.53)***
ln(GDP)
0.017
0.021
0.02
0.016
0.016
(4.94)***
(6.22)***
(5.48)***
(4.34)***
(4.53)***
ln(FOREIGN)
0.046
0.047
(14.65)**
(13.90)***
ln(L-­-FOREIGN)
0.04
0.027
(14.07)***
(6.94)***
ln(H-­-FOREIGN)
0.043
0.02
(12.79)***
(4.33)***
F-­-RATIO
0.028
(1.22)
Constant 6.407 6.922 5.767 6.283 6.197
(2.07)** (2.22)** (1.81)* (2.03)** (2.00)**
Observations 816 816 816 816 816
R-squared 0.76 0.75 0.75 0.76 0.76
Note: Figures in parentheses are t-statistics.
***Significant at 1%; **significant at 5%; *significant at 10%.
Table VI contains results from the fixed effects estimator. Regression VI.1 is our
benchmark in Table VI. Regression VI.1 is the same regression as IV.4 in Table IV,
which is empirically most preferred result.11 Our empirical results from the fixed effects
estimator do not support the heterogeneity between skilled and unskilled non-citizen
workers neither. The
coefficients
on
the
logarithm
of
the
L-­-FOREIGN
variable
are
positive
and
statistically
significant.
Once
again,
however,
the
magnitude
is
way
to
small
to
consider.
Meanwhile,
The
coefficients
on
the
logarithm
of
the
H-­-FOREIGN
variable
are
also
positive
but,
surprisingly,
statistically
not
significant.
In
addition,
the
coefficient
on
the
F-­-RATIO
variable
is
statistically
insignificant,
too.
These
empirical
results
provide
evidence,
in
contrast
to
previous
findings,
that
it
is
not
skilled
immigrants
but
unskilled
immigrants
who
are
complements
to
US
citizens
even
though
the
economic
magnitude
is
small.
Table
VI.
Non-­-Citizen
Workers’
Educational
Attainments
(Fixed
Effects)
ln(Real
Income)
VI.1
VI.2
VI.3
VI.4
VI.5
RATIO
0.304
0.308
0.307
0.307
0.304
(6.10)***
(6.12)***
(6.21)***
(6.10)***
(6.09)***
11
The
Lagrange
Multiplier
(LM)
and
the
Hausman
tests
suggest
that
results
from
the
fixed
effects
estimator
are
the
most
preferred
results
over
OLS
and
the
random
effects
estimator.
AGE
0.304
0.294
0.295
0.296
0.303
(2.34)**
(2.27)**
(2.27)**
(2.27)**
(2.34)**
AGE2
-­-0.004
-­-0.003
-­-0.003
-­-0.004
-­-0.004
(-­-2.25)**
(-­-2.17)**
(-­-2.18)**
(-­-2.18)**
(-­-2.24)**
UNEM
-­-0.007
-­-0.007
-­-0.006
-­-0.007
-­-0.007
(-­-6.91)***
(-­-6.84)***
(-­-6.62)***
(-­-6.91)***
(-­-6.91)***
ln(GDP)
0.21
0.212
0.223
0.212
0.21
(10.31)***
(10.34)***
(11.14)***
(10.38)***
(10.27)***
ln(FOREIGN)
0.011
0.011
(2.06)**
(2.06)**
ln(L-­-FOREIGN)
0.01
0.009
(1.94)*
(1.73)*
ln(H-­-FOREIGN)
0.004
0.001
(1.07)
(0.17)
F-­-RATIO
-­-0.001
(-­-0.07)
Constant
1.617
1.788
1.657
1.762
1.629
(0.64)
(0.71)
(0.66)
(0.7)
(0.65)
Observations
816
816
816
816
816
R-­-squared
0.68
0.68
0.68
0.68
0.68
Finally, we also test for the internal competition among non-citizen workers in the
US. The dependent variable is now the state-level real income of non-citizen workers and
we include the US-­-WORKERS
variable
indicating
the
number
of
US
citizen
workers
in
the
model. The empirical results are shown in Table VII. Regression VII.1 and VII.2
from OLS and regression VII.3, VII.4, and VII.5 are from the fixed effects estimator. The
empirical results suggest that, first, there is no empirical evidence of the internal
competition among non-citizen workers.12 Second, the coefficients on the logarithm
of
the
US-­-WORKERS
variable are positive and statistically significant, which supports,
again, the complementarity between non-citizen workers and US citizens. Third, the real
income of non-citizen workers mainly depends on state and national level of economic
situations not their personal attributes.
Table
VII.
Non-­-Citizen
Workers’
Income
ln(F-­-Real
Income)
VII.1
VII.2
VII.3
VII.4
VII.5
F-­-RATIO
0.071
0.067
-­-0.173
-­-0.168
-­-0.173
(1.26)
(1.2)
(-­-1.23)
(-­-1.2)
(-­-1.23)
F-­-AGE
0.107
0.097
0.028
0.028
0.028
(4.99)***
(4.51)***
(0.6)
(0.58)
(0.59)
F-­-AGE2
-­-0.001
-­-0.001
0
0
0
(-­-4.13)***
(-­-3.72)***
(0.44)
(0.41)
(0.42)
UNEM
-­-0.018
-­-0.019
-­-0.019
-­-0.019
-­-0.019
(-­-4.68)***
(-­-4.91)***
(-­-3.96)***
(-­-4.03)***
(-­-3.95)***
ln(GDP)
0.032
0.009
0.202
0.191
0.196
(4.68)***
(0.97)
(4.23)***
(3.76)***
(3.86)***
ln(US-­-WORKERS)
0.072
0.051
0.046
(3.80)***
(2.29)**
(1.09)
ln(FOREIGN)
0.027
0.005
(1.26)
(0.13)
Constant
7.349
7.348
6.825
7.18
6.911
(17.72)***
(17.87)***
(7.64)***
(7.15)***
(6.66)***
Observations
816
816
816
816
816
R-­-squared
0.15
0.17
0.1
0.1
0.1
12
The
coefficients
on
the
logarithm
of
the
FOREIGN
variable
are
statistically
significant.
V. Conclusion
Using the CPS data set, in this paper, we investigate the impact of non-citizen
workers on US citizens’ state-level income. Our empirical findings suggest that first, noncitizen
workers, immigrants, are complements to US citizens but their contribution on US
citizens’ income is too small to consider. Second, there is no heterogeneity between
skilled and unskilled non-citizen workers in affecting US citizens’ income. Third, there is
no empirical evidence that there is internal competition among non-citizen workers in the
US.
References
Altonji, J. G., and D. Card. (1991) “The Effects of Immigration on the Labor Market
Outcomes of Less-Skilled Natives.” in J.M. Abowd and R.B. Freeman (ed.)
Immigration, Trade, and the Labor Market, University of Chicago Press: Chicago.
Borjas, G. (1995) “The Economic Benefits from Immigration.” Journal of Economic
Perspectives 9 (2): 3-22.
Borjas, G. (2003) “The Labor Demand Curve Is Downward Sloping: Reexamining the
Impact of Immigration on the Labor Market.” Quarterly Journal of Economics 118
(4): 1335-1374.
Borjas, G. (2006) “Native Internal Migration and the Labor Market Impact of
Immigration.” Journal of Human Resources 41: 221-58.
Borjas, G., J. Grogger, and G. Hanson. (2010) “Immigration and the Economic Status of
African-American Men.” Economica 33: 255-82.
Butcher, K., and D. Card. (1991) “Immigration and Wages: Evidence from the 1980s.”
American Economic Review Papers and Proceedings 81: 292-296.
Card, D. (1990) “The Impact of the Mariel Boatlift on the Miami Labor Market.”
Industrial and Labor Relations Review 43 (2): 245-257.
Card, D. (2001) “Immigrant Inflows, Native Outflows, and the Local Labor Market
Impacts of Higher Immigration.” Journal of Labor Economics 19: 22-64.
Card, D. (2005) “Is the New Immigration Really So Bad?” Economic Journal 115 (4):
F300-F323.
Card