Low-Skilled Immigration and the Labor Supply of Highly
By Patricia Cort´ es and Jos´ e Tessada∗
Low-skilled immigrants represent a significant fraction of employment
in services that are close substitutes of household production. This pa-
per studies whether the increased supply of low-skilled immigrants has
led high-skilled women, who have the highest opportunity cost of time,
to change their time-use decisions. Exploiting cross-city variation in
immigrant concentration, we find that low-skilled immigration increases
average hours of market work and the probability of working long hours
of women at the top quartile of the wage distribution. Consistently, we
find that women in this group decrease the time they spend in household
work and increase expenditures on housekeeping services.
JEL: J61, F22, J22
Keywords: Immigration, Household production, Female labor supply
Low-skilled immigrants work disproportionately in service sectors that are close sub-
stitutes for household production. For example, whereas low-skilled immigrant women
represent 1.9 percent of the labor force, they represent more than 25 percent of the work-
ers in private household occupations and 12 percent of the workers in laundry and dry
cleaning services. Low-skilled immigrant men account for 29 percent of all gardeners in
America’s largest cities although they represent only 3.3 percent of the labor force.
The importance of low-skilled immigrants in certain economic activities has been raised
as part of the discussion on immigration policies.
immigration reform in the United States, The Economist argues that:
For example, in an article about
“... in the smarter neighborhoods of Los Angeles, white toddlers occasionally
shout at each other in Spanish. They learn their first words from Mexican
nannies who are often working illegally, just like the maids who scrub Ange-
lenos’ floors and the gardeners who cut their lawns. ...Californians... depend
on immigrants for even such intimate tasks as bringing up their children.”
(The Economist, “Debate meets reality,” May 17th, 2007.)
If, as found by Patricia Cort´ es (2008), the recent waves of low-skilled immigration have
led to lower prices of services that are close substitutes for household production, we
∗Cort´ es: Boston University School of Management, 595 Commonwealth Avenue, Boston, MA 02215
(email: firstname.lastname@example.org). Tessada: Escuela de Administraci´ on, Pontificia Universidad Cat´ olica de Chile,
Avda. Vicu˜ na Mackenna 4860, Macul, Santiago 7820436, Chile (email: email@example.com). We thank
George-Marios Angeletos, Josh Angrist, David Autor, Marianne Bertrand, Olivier Blanchard, Esther
Duflo, Alexander Gelber, Carol Graham, Michael Kremer, Jeanne Lafortune, Alan Manning, and Stan-
ley Watt for their useful comments and suggestions. We also acknowledge seminar participants at MIT’s
Development and Labor lunches, NBER Labor Studies Meeting, CEA–Univ. de Chile, PUC–Chile, Mary-
land, ZEW’s Workshop on Gender and Labor Markets, the Seventh IZA/SOLE Transatlantic Meeting,
the 2008 IFN Stockholm Conference on Family, Children and Work, and Brookings for helpful comments.
Tessada thanks the Chilean Scholarship Program (MIDEPLAN) for financial support.
2AMERICAN ECONOMIC JOURNAL MONTH YEAR
should expect natives to substitute their own time invested in the production of house-
hold goods with the purchase of the now cheaper services available in the market. The
link between immigration and changes in the prices of these market-provided household
services indicates that even without a direct effect on wages, low-skilled immigration has
the potential to generate effects on natives’ decisions related to time-use.
This paper studies this unexplored channel, focusing particularly on the impact that
low-skilled immigration has on female labor supply. We first develop a simple model to
investigate which groups of the female population are more likely to change their time-
use decisions as prices for household related services decrease; we then test the model’s
predictions using Census data on immigration and labor supply, and information on time
devoted to household work and reported expenditures on housekeeping services.
Our empirical strategy is to exploit the cross-city variation in the concentration of
low-skilled immigrants. To address the potential endogeneity of the location choices of
immigrants, we instrument for low-skilled immigrant concentration using the historical
(1970) distribution of immigrants of a country to predict the location choices of recent
There are two main concerns with the validity of our instrumental variables strategy.
First, cities that attracted more immigrants in 1970 might be systematically different from
other cities, violating the identification assumption. To address this concern we include
specifications that allow cities to experience different decade shocks based on their 1970
value of key variables such as female educational attainment distribution, female labor
force participation, and industry composition. The second concern is that low-skilled
immigration might have an impact on the labor supply of women through other channels
besides lowering prices of household services; in particular, through interactions in market
production. To tackle this issue we present specifications that use men of similar skill
level as a control group and we test that the estimated relative increase in the labor
supply of women as a result of low-skilled immigration is not driven by an increase in
their wages relative to men. The relevance of the household services/time-use decision
mechanism is also tested by comparing the estimated pattern of the immigration effect
by skill level with the pattern predicted by our simple time-use model and by looking at
the mirror decisions of work at home and consumption of household services.
We first estimate reduced form regressions of female labor supply outcomes by wage
groups as a function of the supply of low-skilled workers, as our model predicts that
only women with high wages are the ones who will be affected by the reduction in prices
of household services resulting from a low-skilled immigration influx. We find a large
positive and statistically significant effect of low-skilled immigration on the hours worked
per week of working women at the top quartile of the female wage distribution. Consistent
with our framework, much smaller, but still statistically significant, effects are found for
women above the median, and no effects are found for women with wages below the
median. Looking at women grouped by the median male wage of their occupation and at
women at the top of the educational distribution, we confirm significant positive effects
on the intensive margin of the labor supply of highly skilled women, but find no similar
effects on labor force participation (that is already high in our group of interest).
VOL. VOL NO. ISSUEIMMIGRATION AND FEMALE LABOR SUPPLY3
Occupations with the highest wage levels (such as physicians and lawyers for example)
are also characterized by people having to work long hours in order to have a successful
career.1Low-skilled immigrants, on the other hand, are regarded as providing cheaper
and more flexible household services than those provided by native workers and compa-
nies.2Thus, part of the effects we estimate can come from a match between the services
demanded by women in occupations that require long hours and the more flexible ser-
vices provided by low-skilled immigrants. We test this hypothesis estimating regressions
of indicator variables for working more than 50 and 60 hours on our immigration vari-
able. Focusing on women working in occupations where men have long hours of work,
we find large positive and statistically significant effects of low-skilled immigration in the
probability that women also work long hours.
The sign and statistical significance of all our results are robust to specifications that
use men as a control group and that include city*decade fixed effects. However, the
magnitudes are smaller, between a fourth and a fifth of those that include only women.
This difference likely reflects that men’s time-use decisions might also be affected by
the lower prices of household services, and that part of the effects estimated using the
women sample were coming through other channels, for example, complementarities in
Overall, our estimates suggest that the low-skilled immigration wave of the period
1980-2000 increased by 20 minutes a week the time women in the top quartile of the
wage distribution devote to market work. Our more conservative estimates suggest that
at the very least, 4 of those minutes can be attributed to low-skilled immigrants reducing
prices of household services according to our more conservative estimates. The low-skilled
immigration wave has also increased the probability that women working in occupations
that demand long hours work more than 50 and 60 hours a week by 1.8 and 0.7 percentage
More hours of market work resulting from lower prices of household services should
be reflected in less time devoted to household production. Using data from the recently
released 2003-05 American Time Use Survey and from the 1980 PSID, we find that the
immigration wave of the 1980s and 1990s reduced by close to 7 minutes a week the time
women at the top of the wage distribution spent weekly on household chores.
Finally, using the Consumer Expenditure Survey (CEX) we find that low-skilled immi-
gration have increased the likelihood that highly-skilled women consume market provided
1For example, whereas the cross-occupation average of usual hours worked per week for men is 35.5,
and the share working more than 50 hours is 7.4 percent and more than 60 hours is 2.6 percent, the same
numbers for physicians are 47 hours, 44 percent, and 28 percent, and for lawyers 42 hours, 31 percent,
and 10 percent.
2For example, a study by Domestic Workers United (2006) in New York City reports that nearly half
of domestic workers (most of which are immigrants) work overtime, often more than 50 and 60 hours per
week, and that even when they are working a five-day week, the days extend to 10-12 hours. Zoe Baird,
who lost her chance to be attorney general for hiring illegal immigrants, supposedly placed the following
ad in three local newspapers: “Child Care Nanny. Live-in Nanny for 7 Mo. old Boy in warm family
setting. Light housekeeping, cook dinners. Long term position with appreciative family in beautiful
home. Non-smoker. Driver. Citizen or green card only.” She and her husband received not one response
(Anna Crittenden 2001).
4AMERICAN ECONOMIC JOURNAL MONTH YEAR
household services and their expenditures in these services.
Our findings with respect to highly skilled women have important implications. First,
we provide evidence of a specific channel, different from wages, through which low-skilled
immigration might be affecting the labor supply of highly skilled native workers; in partic-
ular, we find that women with high wages (and potentially their families) are benefitting
from low-skilled immigration because of the reduction in the prices of services that are
close substitutes for household production.3Furthermore, our results suggest that look-
ing purely at the effect on wages might not show all the various effects immigration has
on the different skill groups. This paper, therefore, provides a new perspective on the
literature of the labor market effects of low-skilled immigration, particularly to our un-
derstanding of the effects of immigration across the wage and educational attainment
Second, the results suggest that the availability of flexible housekeeping, including child
care services among others, at low prices might help women in occupations demanding
long hours or irregular work schedules to advance in their careers.5Conflicting demands
of the profession and of the household have been linked to the relative lack of women
in positions of leadership (such as partners in law firms) and in prestigious medical
specializations, such as surgery.6On the other hand, it provides some evidence against
recent theories that highly skilled women are opting out of demanding careers because
they place a higher value on staying home with their children.7Overall, it suggests that
differences in preferences are not the only reason that highly educated women are not
more actively involved in the labor market.
Outline. — The rest of the paper is organized as follows. The next section presents
the theoretical framework. Section II describes the data and the descriptive statistics.
Section III presents the empirical strategy and discusses the main results, and in Section
IV we conclude.
3Phanwadee Khananusapkul (2004) is, to the best of our knowledge, the only previous study that
relates low-skilled immigration with the labor supply of high skilled women. The author finds that an
increase in the proportion of low-skilled female immigrants in a metropolitan area raises the proportion
of private household workers and lowers their wages. She does not, however, find a significant effect on
the labor supply of college educated women.
4See Gordon Hanson (2009) for a recent survey of the literature on the effects of migration.
5Jonah Gelbach (2002) and Michael Baker, Jonathan Gruber and Kevin Milligan (2008) show some
results regarding the labor supply effects of differences in the cost of childcare driven by government
subsidies or the admission rules to public schools. Daniele Coen-Pirani, Alexis Le´ on, and Steven Lu-
gauer (2010) look at the same household production-labor supply connection, but focus in the increased
availability household appliances in the US during the 1960s.
6For example, Mona Harrington and Helen Hsi (2007) say that “While many women with children
negotiate a part-time schedule for family care... they are still less likely to be promoted to partner than
women who stay in firms but do not use part time options” ... “The expectation that an attorney needs
to be available practically 24/7 is a huge impediment to a balanced work/family life.”
7The headline for the October 26, 2003, edition of the New York Times Magazine was “Why don’t
more women get to the top? They choose not to.”
VOL. VOL NO. ISSUEIMMIGRATION AND FEMALE LABOR SUPPLY5
In this section we present a simple time-use model that illustrates the interactions
between wage levels, the decision to purchase household services, the market price of
household services, and labor supply. Its purpose is to derive implications about which
groups of the population are more likely to change their time-use decisions as prices for
household related services decrease. At the end of the section we also investigate if women
who face larger household demands (for example, women with young children) display a
differential sensitivity to prices, and discuss how career concerns might interact with the
price of household services.
An agent allocates her time between leisure, household production, and market work.
She receives a wage w per unit of time devoted to market work.
The agent consumes two goods. First, there is a homogeneous consumption aggregate
that can only be bought in the market; we normalize its price to 1. Second, the agent’s
household requires a certain number of units of a household service to function; this
service can be produced at home or bought in the market at a price p. The household
needs exactly R units of this service; the marginal benefit of units beyond R is 0.
Denote by y the amount of the consumption good, l the hours of leisure, h the hours
of household work, n the hours of market work, x the units of the household service
purchased on the market, and I the non-wage income of the household. Assume that
there is only one working agent per household and normalize total time available to the
agent to 1.
Utility is given by
(1)u(y) + ψ (l),
where u(·) and ψ (·) are concave and satisfy u?(y) → ∞ as y → 0 and ψ?(l) → ∞ as
l → 0. Household production is described by the function f (h), which we assume to have
decreasing marginal returns to time spent at working at home and to satisfy f?(h) → ∞
as h → 0. This condition implies that a person will never outsource all of her household
work. The agent also faces two constraints: a budget constraint and a time constraint
(which can be reduced into a full income constraint).
Four important results arise from the solution of the model. First, people with higher
wages (for a given level of I and p) supply labor in the market. Second, for a given w and
I, a decrease in p might induce a person to purchase market-provided household services,
or to purchase even more. Third, for a given p and I, people with higher wages are more
likely to buy household services. Finally, only those who purchase services will change
their decisions at the margin when p changes.8
8The details of the solution of the model are presented in the Theory Appendix.
6AMERICAN ECONOMIC JOURNALMONTH YEAR
B.The effects of an inflow of low-skilled immigrants
Based on Cort´ es (2008) we model an inflow of low-skilled immigrants as a decrease in
p. Furthermore, we assume the immigration inflow has no effect on wage levels, at least
for the group that purchases household services in the market. It follows according to our
simple model that women with higher wages will be more likely to respond to immigrant
induced changes in p. The model also suggests that if we observe time-use effects of
immigration in other groups, especially those characterized by low wages, they are likely
to come through other channels besides changes in p.
Effect on household work (h). — For agents with high enough productivity outside
the household such that it is optimal for them to outsource part of the household pro-
duction, a decrease in p will reduce the number of hours worked at home. We should not
see changes in hours spent in household production for households with lower wages (but
not low enough that we expect them to compete with immigrants in the labor market).
Two additional points are worth mentioning. First, a decrease in p might induce some
agents who were previously not buying household services –but who had high enough
wages to be close to the threshold– to start doing so. Second, under a fairly simple
household production function (for example f(h) = ln(h) ), within high salaried agents
that already work and purchase household services, the ones with lower salaries will
decrease their household work by more than those with higher salaries if p falls.9This
means that conditional on initially purchasing household services, the effect of a fall in p
might be decreasing in the wage. Therefore we expect the effect of a fall in p on household
work to be stronger for the high salaried group as long as the much lower share of women
purchasing household services in other groups dominates the intensive margin effect.
Effect on labor supply (n). — As with the effect on h, only certain agents’ labor
supply decisions will be affected by a drop in p. Only agents that are both working in the
market and purchasing household services will show any change on their labor supply in
response to a drop of p; as we mentioned before these agents are characterized by high
The effect on n will depend on how hours worked in household production and leisure
change after a decrease in p. From the previous subsection it is clear that∂h∗
that changes in p keep the relative price of leisure versus consumption good unchanged,
the effect on leisure happens through a change in disposable income only. Its direction
will depend on whether leisure is a normal or inferior good. If leisure is an inferior good
or if it doesn’t respond to income changes, then hours worked in the market is going to
unambiguously increase when p goes down. If leisure is a normal good (as in our case
∂p> 0. Given
9Intuitively, agents with very high wages are already spending very little time working at home;
therefore, compared to an agent with a high but relatively lower wage, her marginal productivity in
household production is relatively large. The shape of the production function, given by our assumptions
about f??(·) and f?(0), imply further decreases in h require larger reductions in p.
VOL. VOL NO. ISSUEIMMIGRATION AND FEMALE LABOR SUPPLY7
because the utility function is separable in y and l), then the direction of the effect will
depend on the relative magnitudes of∂h
increases or decreases after a change in p can only be determined empirically.
In our particular case, we can show that the total effect can be decomposed as
∂Income. Therefore, whether labor supply
w2u??(·) + ψ??(·).
Note that if the income effect is fairly small we have that
From equation (2) we can also conclude that all else being equal, agents with higher
unearned income (and therefore higher use of market-provided household services, x) will
react less to changes in p.
Summarizing, the model predicts that (only) women with high wages will be affected
by the reduction in prices of household services resulting from a low-skilled immigration
influx. This is true because for given household characteristics and preferences, women
with higher wages buy market services and supply labor in the market. For this group
of women a decrease in prices will likely reduce the hours spent in household production,
and might increase the hours worked in the market if leisure is not very sensitive to
income. Within the group of women affected by the change in p, those with the higher
wages, the ones with lower wages will react more. Finally, higher unearned income is
associated with a smaller labor supply response.
Household Composition: Children at home. — As we have argued before, in order
for p to have an effect on a woman’s labor supply and time spent at home production,
she must be purchasing household services in the market. An important characteristic
affecting the demand for household services is the presence of a child at home, and
although some services are provided for free by institutions such as public schools, when
children are not old enough to attend school the burden of the care provision lies on the
family. In this section we explore how time-use decisions, participation in the market of
household services, and the response to immigrant-induced reduction in prices of market
substitutes are affected by the presence of a child at home (age less than 6 so that they
cannot receive most of the services available from schools). For simplicity, we model
this case as a household requiring a higher number of units of household services, R, to
10Although a simplification, we think that this way to model the presence of children at home captures
the fact that more tasks must be performed at home. We could also assume that a woman must perform
a certain number of tasks when she has a child at home; when she performs these tasks, then we can
argue that f?(·) with children is less or equal to without children at home. We can think of this reflecting
that the person is more tired after performing the child-related work or that more of the same chores
8AMERICAN ECONOMIC JOURNALMONTH YEAR
Labor Force Participation and the Decision to Purchase Household Ser-
vices. — In our simple model, “participation” in the market for household services,
i.e., the decision of whether and how much household services to buy from the market is
directly linked to the total amount of services needed for the home. Labor supply is also
affected by the level of R but the direction of the effect will depend on other parameters
of the model, with the intensive and extensive margins showing different responses in
For example, consider the case of an agent that does not purchase market services.
In this case, a larger R has two effects: it increases the cost of time because the agent
spends more time doing household work, thus reducing leisure and/or labor supply, and
it lowers the marginal productivity of time devoted to household work. Eventually, for
a sufficiently large increase in R this agent might start purchasing services. A different
situation arises if we take the case of an agent who supplies labor and purchases household
services. In this case an increase in R translates into a higher demand for market services
x, and a higher labor supply to cover part of the bill of the additional market services.
All in all, compared to otherwise identical women, mothers of young children are more
likely to buy market provided household services. Also, conditional on purchasing house-
hold services, mothers buy more units of the household service than non-mothers.
Sensitivity to Changes in Prices of Household Services. — How is the time-use
response to an immigration-induced reduction in p affected by having young children?
While the model delivers clear predictions about level differences in time-use and ex-
penditures in household services by motherhood status, predictions about differences in
sensitivity to prices are less clear. Consider for example the case of a woman with a
sufficiently high wage such that she outsources part of her household work, x∗> 0, and
also supplies labor in the market, n∗> 0. From our discussion in section I.B, we know
the labor supply response to a change in p depends on how hours devoted to household
work react and on the magnitude and sign of the income effect on leisure. For these
women, the derivative (∂h∗/∂p) is independent of R, so differences in the sensitivity of
labor supply to prices between mothers and non-mothers depend only on differences in
the income effect on leisure. If leisure is a normal good, and given that mothers spend a
larger fraction of their total income purchasing market-provided household services, the
income effect of a reduction in p will be larger. Consequently, mothers would increase
their leisure (and thus decrease their labor supply) relatively more than non-mothers. An
opposite effect will be observed if leisure is an inferior good or if there is a strong degree
of complementarity between household work and consumption or if certain household
activities have an intrinsic utility value.12
must be done (i.e., the house cleaning is more demanding with children and so it takes more hours to
perform ”house cleaning”). A change in f?(·) in this direction produces qualitatively similar results, and
if added, it reinforces the channels we explain in this subsection.
11A more detailed analysis can be found in Theory Appendix, where we also make use of the solution
to the model to expand the discussion.
12There are other potential channels that can generate a link between these two elements. Consider
for example the case of R being a variable of choice, in which case it would respond to p, or if household
VOL. VOL NO. ISSUE IMMIGRATION AND FEMALE LABOR SUPPLY9
Taking the model’s predictions to the data complicate matters even more, given that we
will not be able to separate women according to their consumption of market-provided
Therefore, the observed differences between mother’s and non-
mother’s response to price changes will come from three sources. First, given that mothers
are more likely to purchase household goods, they are more likely to be affected by a
price change. Second, however, conditional on purchasing household services, their labor
supply react less relative to non-mothers if leisure is a normal good. Finally, differences
in R can also affect changes in the extensive margin of labor supply in response to the
changes in p, thus changing the composition of the groups constructed for the empirical
work. One plausible situation would be a mother that was working reduced hours before
and not purchasing household services, and that with the reduction in p it shifts into a
situation where she works more hours and uses market-provided services.
In sum, we cannot use our empirical estimation of the relative sensitivity of the labor
supply of women with (young) children as a test of the validity of our model. However,
given that women with young children are a natural group of interest, we explore their
differential labor supply response to prices in section III.C.
Career concerns. — The simple model outlined in this section focuses on a static
labor supply/household production decision. However, it is reasonable to think that
the labor supply decisions of women with high wages (who are the most affected by
changes in p in our model) are also likely to reflect choices about career paths, and thus,
include intertemporal issues not captured by our model. For example, recent empirical
evidence by Marianne Bertrand, Claudia Goldin and Lawrence Katz (2010) suggests
that in high wage occupations, in particular those in the corporate and financial sectors,
shorter work hours and career interruptions carry a huge penalty in terms of future
earnings growth, explaining to a large degree the significant gender gap observed a few
years after graduation. Given that a reduction in the price and an increase in the flexibility
of services that are close substitutes for household production allow women to extend their
working hours, in our empirical work we will study whether more highly skilled women
choose to work very long hours in response to low-skilled immigration flows. The effects
at the top of the distribution of hours worked might be more marked than the responses
of average hours of work and labor force participation if, as expected, career concerns are
important for a significant fraction of highly skilled women.14
II.Data and Descriptive Statistics
We now describe the basic details of the data we use to measure immigration, labor
market outcomes, and household production outcomes.
production actually contributes to utility (for example by turning the “market” good y into a consumable
13As it is explained later in section III.C, our empirical work separates women according to wages and
other labor market characteristics, but we cannot classify them at the same time according to (potential)
wages and use of market-provided household services.
14For a more technical discussion see the Appendix.
10AMERICAN ECONOMIC JOURNAL MONTH YEAR
Immigration Data. — This paper uses the 5 percent sample of the 1980, 1990, and
2000 Census Integrated Public Use Microdata Samples to measure the concentration of
low-skilled immigrants among cities. Low-skilled workers are defined as those who have
not completed high school and an immigrant is defined as someone who reports being
a naturalized citizen or not being a citizen. We restrict the sample to people age 16-64
who report being in the labor force and not enrolled in school.
Table 1 shows the evolution of the share of low-skilled immigrants in the labor force
for the 30 largest cities in the United States. As observed there is significant variation
in immigrant concentration both across cities and through time. This is the variation we
will exploit in our empirical strategy.
[TABLE 1 ABOUT HERE]
Market Work Data. — We also use the Census to quantify the labor supply of native
women, and restrict the sample to individuals who were between age 20 and 64. We start
by describing the labor market behavior of working women by wage percentile– as they
will be our main focus in the empirical analysis (Table 2). We also present descriptive
statistics by education level (Table 3), because, even if education is only a proxy for
market wage, we observe labor supply variables for every woman, including those not
[TABLE 2 ABOUT HERE]
As Table 3 shows, women at the bottom of the hourly wage percentile work fewer hours
a week, are younger, and are less likely to be married. On the other hand women above
the median tend to work more hours a week, and women in the top quartile and top
decile, in particular, are much more likely to work more than 50 or 60 hours.
[TABLE 3 ABOUT HERE]
Labor force participation and the number of hours worked a week increase system-
atically with the education level of the woman (see Table 3). Women with a graduate
degree, a college degree, and some college present a significant increase in their labor
force participation between 1980 and 1990. During the past decade, participation of all
education groups has stabilized, and if anything it has gone down. It is important to
note that close to a third of professional women (mostly doctors, lawyers and Ph.D.s)
reported working 50 hours or more a week in 2000, a double-fold increase from 1980 and
at least two times as large as the share for women from any other group. Highly educated
women are also at least three times as likely, compared to any other educational group,
to work 60 hours or more a week.
Household Work Data. — We combine information from the 2003-2005 American
Time Use Survey (ATUS) with the 1980 Panel Study of Income Dynamics (PSID) to
measure time devoted to household work.
Since 2003, the Bureau of Labor Statistics (BLS) has been running the ATUS, a
monthly survey, whose sample is drawn from CPS two months after households com-
plete their eight CPS interviews. An eligible person from each household is randomly
selected to participate, and there are no substitutions. The week of the month and the
VOL. VOL NO. ISSUEIMMIGRATION AND FEMALE LABOR SUPPLY 11
day of the week on which the survey is conducted are randomly assigned; weekends are
oversampled, they represent 50 percent of the sample. The overall response rate is 58
percent and the aggregated sample for 2003 to 2005 consists of approximately 38,000
Until the ATUS, only scattered time use surveys were available for the U.S. –all of
them with too few observations to provide reliable information about city-averages of
time allocation. Though not a time use survey, the PSID included between 1970 and
1986 a question about average hours a week spent by the wife and head of household on
household chores. We construct a similar variable using the ATUS data. Specifically, we
aggregate daily time spent on food preparation, food cleanup, cleaning house, clothes care,
car repair, plant care, animal care, shopping for food, and shopping for clothes/household
items, multiply this aggregate by 7 and divide it by 60. We hope to capture any difference
in the definition of household work using decade dummies. For both surveys, our sample
consists of women ages 21-64 who have completed the survey.
[TABLE 4 ABOUT HERE]
Table 4 presents the descriptive statistics of our time-use data. For all women, hours
spent on household work decreased significantly between 1980-2000. In both years, time
spent on household chores is significantly smaller for women above the median wage. The
time men spent doing household work is between a half and a third of the amount women
with similar wages spent. Note that PSID’s and ATUS’s statistics on usual hours worked
and general demographic characteristics are not very different from the Census.
Consumption Data. — We use the Consumer Expenditure Survey (CEX) to construct
two measures of consumption of market supplied household services. First, in order to
capture the extensive margin, we consider a dummy variable for positive reported ex-
penditures in housekeeping services. Second, we also consider the amount spent on each
of these services, a measure we identify as capturing mostly the intensive margin.15As
observed in Table 5, the probability of consuming household services increases signifi-
cantly with the wage percentile of the wife/female head of the household. Whereas in
2000 only 3 percent of households where such a female had a wage below the median,
that fraction rises to 8 percent, 18 percent, and 26 percent when considering females at
the third quartile, top quartile, and top decile, respectively. Note that this pattern is
consistent with the predictions of the model, where only women with high wages or high
unearned income will purchase household services. Expenditures on household services
tend to increase with the wage percentile of the main adult female in the household:
households with a wife or female household head at the top quartile of the wage distri-
bution (conditional on reporting positive expenditures) spent close to 30 percent more in
housekeeping services than other households.
[TABLE 5 ABOUT HERE]
15We do not include child-care at home because the variable in the CEX was redefined between 1990
12AMERICAN ECONOMIC JOURNALMONTH YEAR
A. Identification Strategy
We exploit the intercity variation in the (change of the) concentration of low-skilled
immigrants to identify their effect on the time-use decisions of American women and
purchases of household services in American households. There are two concerns with
this strategy. First, immigrants are not randomly distributed across labor markets. To
deal with this potential bias, we instrument for immigrant location using the historical
city-distribution of immigrants of a given country. The instrument will be discussed
thoroughly in section III.B.
The second concern is that local labor markets are not closed and therefore natives
may respond to the immigrant supply shock by moving their labor or capital to other
cities, thereby re-equilibrating the national economy. Most of the papers that have em-
pirically tested natives’ migration response to immigration have not found evidence of
large displacement effects. David Card and John DiNardo (2000), Card (2001), and Card
and Ethan Lewis (2005), using different samples and specifications, have all found that
native mobility has virtually no offsetting effect on the relative supply shocks created by
immigration. Larger, but still not perfectly off-setting displacement effects are found by
George Borjas (2006). He estimates that 6.1 fewer native workers choose to reside in a
city for every ten new immigrants that arrive in the city. In any case, if factor mobility
dissipates the effects of immigration flows to cities, our estimates should provide a lower
bound for the total effect of low-skilled immigration on the time-use of natives.
The instrument exploits the tendency of immigrants to settle in a city with a large
enclave of immigrants from the same country. Immigrant networks are an important
consideration in the location choices of prospective immigrants because these networks
facilitate the job search process and assimilation to the new culture, see Kaivan Munshi
(2003). The instrument uses the 1970 distribution of immigrants from a given country
across U.S. cities to allocate the new waves of immigrants from that country. For example,
if a third of Mexican immigrants in 1970 were living in Los Angeles, the instrument
allocates one third of all Mexicans in the 1990s to Los Angeles.
Formally, the instrument for the number of low-skilled immigrants in city i and decade
t can be written as
where j are all countries of origin included in the 1970 Census,
the percentage of all immigrants from country j included in the 1970 Census who were
living in city i, and LS Immigrantsjtstands for the total number low-skilled immigrants
from country j to the United States in decade t.
VOL. VOL NO. ISSUEIMMIGRATION AND FEMALE LABOR SUPPLY 13
All of the econometric specifications in the paper include city and region*decade fixed
effects (we use the 9 Census divisions). Therefore, the instrument will help in identifying
the causal effect of immigration concentration on time use of native women as long as
the following conditions hold:
1) The unobserved factors determining that more immigrants decided to locate in city
i vs. city i? (both cities in the same region) in 1970 are not correlated with changes
in the relative economic opportunities for skilled women offered by the two cities
during the 1980s and 1990s. To ameliorate the concern that cities that attracted im-
migrants in or before 1970 are systematically different from other cities we present
specifications that allow for cities within a region to experience different demand
shocks based on 1970 values of key variables, such as female labor force participa-
tion, education composition of women, industry composition of employment, and
2) The total (national) flow of low-skilled immigrants in a given decade (second term in
the interaction) is exogenous to differential shocks to cities within a given region.16
Estimation of the first stage and a few robustness tests are presented in Table 6. The
magnitudes of the coefficient suggest that, at current United States immigration levels,
an increase of 10 percent in the predicted number of low-skilled immigrants increases the
share of low-skilled workers by around 2 percent. The inclusion of the additional controls
and exclusion of California and of the top migrant cities do not change the magnitude or
statistical significance of the coefficient.17
[TABLE 6 ABOUT HERE]
Even if the identification assumption holds, an additional concern for the interpretation
of the IV estimations is the violation of the exclusion restriction, i.e., that changes in the
prices or the availability of household related services are not the only channels through
which low-skilled immigration might be affecting the time use of American women. A
natural candidate is the effect that low-skilled immigration might have on the wages of
natives. To partial out the confounding channels we present specifications that include
both men and women, allowing us to use men of identical skill as controls and to incor-
porate in the regressions city*decade fixed effects. These fixed effects also help address
even further potential violations of condition 1 above.
16One might be concerned that this condition is violated if city specific pull factors are the driving force
in the decision of low-skilled foreigners’ migration decisions. Leah Boustan (2007) notes this problem
and assesses its quantitative importance. She compares results from instruments that assign either the
actual or the predicted migrant flows, where the predictions are based on push factors from source areas,
and finds little difference between the two.
17First-stage coefficients estimated with the individual data are presented in Table 6. Once the error
terms are properly clustered, the magnitude and statistical significance is very similar to the coefficient
estimated with data at the city level.
14AMERICAN ECONOMIC JOURNALMONTH YEAR
C. Econometric Specifications and Results
Our theoretical framework suggests that price indexes (in particular, the price index
of household services in a city) should be the explanatory variable in our analysis of
time-use and consumption. However, there are no price indexes available that cover the
universe of activities we consider and the few that are available cover only a subset of the
sample (they are available only for 30 cities).18Therefore, we present basic reduced-form
specifications using as explanatory variable the log of the share of low-skilled workers
in the labor force (henceforth denoted by Lit), which is a simplified version of Cort´ es’s
(2008)’s price equations’ main explanatory variable.
Labor Supply of Highly Skilled Women. — We start our empirical exercise by
investigating the labor supply effects of changes in low-skilled immigrant concentration
by wage percentile. We use the following specification, where the dependent variables of
interest are usual hours a week worked, the probability of working at least 50 hours a
week, and the probability of working at least 60 hours a week:
(3)LSnit= δw× Lit+ X?
t+ τw× Additional Controlsit+ φw
?LS Immigrants + LS Natives
and w corresponds to the wage percentile of the individual, i is city, t is decade, and j is
region.19The variable LSnitrepresents the labor supply variable of choice of a woman n
in city i and decade t. The vector Xnitincludes individual level characteristics, namely
age, age squared, race, marital status, and the presence of children in several age brackets.
Henceforth, φiand ψjtrepresent city and region*decade fixed effects, respectively.
To account for the fact that the main predictive variable Litvaries only at the city*decade
level and, moreover, that labor supply is not independent among workers in a given city,
the standard errors are clustered at the city*decade level. In our robustness checks we also
show the standard errors using city clusters to address the possibility of serial correlation
within cities across decades (see Appendix Table A1).
Based on our theoretical model, our hypothesis is that δw?= 0 for women with very high
wages (i.e. for high values of w). The direction of the effect is theoretically ambiguous;
however, if the income elasticity of leisure is negative or not very large, then we should
expect to find a positive effect of low-skilled immigration on the labor supply of highly
educated women. Additionally, and as explained in our theoretical framework, career
concerns actually reinforce this effect as women will take advantage of more flexible
18In separate specifications we have run some specifications for labor supply outcomes, similar to those
in section III.C, using the prices of household services as the explanatory variable; results are qualitatively
similar to those we obtain here (they are available upon request from the authors).
19See Cort´ es (2008) pages 389-393 for a derivation of ln
explanatory variable in the determination of the prices of nontraded goods.
?LS Immigrants+LS Natives
as the main
VOL. VOL NO. ISSUEIMMIGRATION AND FEMALE LABOR SUPPLY 15
and/or cheaper services to take up positions that require longer (and maybe irregular)
hours of work.
Table 7 presents the estimates of equation (3). We divide the working women popula-
tion into quartiles (and also study the top decile), and present OLS and IV estimations.20
Our model predicts that only women in the highest percentiles of the wage distribution
should change their labor supply as a result of low-skilled immigration lowering the prices
of household services. As observed in the table, our IV coefficients exhibit a clear de-
creasing pattern as we move down to groups with lower wages. For women in the top 25
percent of the female wage distribution, a 10-percent increase in low-skilled immigration
from current levels increases the time women in this group work by 5 to 6 minutes per
week.21Note that the estimate is not driven entirely by the highest 10th percentile. The
effect is reduced to between a third and a half for women earning hourly wages above
the median, but below the top quartile. For women with wages below the median we
observe no significant effect of low-skilled immigration on their hours worked. Specifica-
tions that include as additional controls the 1970 values of key variables –such as labor
force participation of women, education composition of women, industry composition of
employment, and wage levels– interacted with decade dummies show a very similar pat-
tern. Other robustness tests that address concerns about the importance of outliers and
of endogenous internal migration of highly skilled women are presented in the Appendix
[TABLE 7 ABOUT HERE]
Table 7 also shows that OLS coefficients are smaller than their IV counterparts. Ex-
ante it is difficult to anticipate in which direction the bias will go: If low-skilled workers
tend to move to thriving economies, we would have expected OLS to have an upward bias.
However, if on the contrary, low-skilled workers stay away from cities with a high cost
of living (where highly skilled women are likely to work longer hours), OLS coefficients
should be smaller. Measurement error will also push OLS estimates towards zero.
Because group classification by wage percentile does not allow us to explore the effects
of immigration on the extensive margin, in Table 8 we present alternative classifications
20To construct the quartiles we use the wage distribution of women in each one of the 9 Census
regions, women are then assigned to the corresponding quartiles and these groupings are used for all the
metropolitan areas within each region.
21Given that we are ultimately interested in the magnitude of the effect of immigration flows on
consumption and time use, we use the chain rule for its estimation:
the coefficient that measures the impact of L on outcome LS.
The last equality is based on the assumption that
Note that the share of immigrants in the low-skilled labor supply varies significantly by city. We use its
value for each city from the 1990 Census to calculate the city-specific immigration effect on consumption
and time use of the low-skilled immigration flow of the 1990s. We report the weighted average across
cities of these effects unless explicitly noted.
d(ln LSImmigrants )= θ ×
LS Immigrants + LS Natives
LS Immigrants + LS Natives
is the share of immigrants in the low-skilled labor supply and θ is
d(ln I)= 0, i.e. there are no displacement effects.