€¦ · web viewcultural attitudes may impact economic decisions, especially those concerning...
TRANSCRIPT
Mahmoud Salari
PhD CandidateGraduate Part-time Instructor
Department of Economics, Texas Tech University, PO Box 41014, Lubbock, TX 79409-1014, United States
Tel. +1 806 474 9549
E-mail address: [email protected]
Culture and Economic Decision Making
Cultural attitudes may impact economic decisions, especially those concerning female labor force
participation. To distinguish the economic/institutional effects from the cultural effects on female
labor force participation, an epidemiological approach is used studying second-generation immigrant
women. Cultural proxies are found to be statistically significant for women who kept their heritage
languages when controlling for characteristics of the women, their husbands and their families.
Additionally, the results show that the cultural proxies are positively significant explanatory
variables for those women who kept their heritage languages when controlling origin characteristics.
Economic outcomes have experienced large variation across nations since World War II. Female
Labor Force Participation (FLFP) as one of the main economic outcomes varies remarkably across
time and countries [1,2]. Many theories have been proposed as an explanation for this deviation. The
majority of these theories highlight the role of market prices, technological factors, policies and
institutions. Also, few current studies suggest that cultural factors can explain this deviation [1,3–6].
Culture is defined as those traditional beliefs, values and ethics of social groups that transmit from
one generation to the next generation without substantial changes [5,7,8]. Most economists believe
that culture plays an important role in economic decisions; however, designing testable hypotheses
for measuring the cultural effect is quite challenging. The key argument of any cultural analysis asks
how to distinguish between the cultural effects and the impact of other economic factors. One of the
best ways to measure the cultural effect is assessing second-generation immigrants’ behaviors in a
host country. Each immigrant has a specific culture that he/she brings to the host country and may
transmit to his/her own second generation [9,10]. Immigrant families may present their origin culture
to own children through speaking their Heritage Language (HL) [11,12]. HL is a primary instrument
that transmits parents’ origin cultural beliefs to the second-generation. When parents speak HL to
their children, they can easily convey their own values, beliefs and experiences [11,13].
There are two main reasons why immigrant parents desire to transmit their heritage culture through
speaking HL to their own children. First, immigrants who have a strong ethnic orientation like to be
involved in activities that present their culture to their family; therefore, their children need to speak
HL if they want to participate in these cultural events. Second, speaking HL is an important factor
for second-generation children to make sense of themselves as a part of the heritage group, and HL
links them to their parents’ heritage country [11,14]. The purpose of this study is twofold. 1) To
show that cultural proxy is a kind of dynamic interaction that moves slowly over time and is not the
same for all second-generation immigrant women. 2) To show that speaking HL in the host country
allows parents to fully transmit their heritage cultures’ beliefs about FLFP to their own children.
This study examines the labor supply of second-generation American women who grow up in the
U.S. and have the same economic, institutional environment, and English language competency as
other native women in the U.S. labor market. The main difference for second-generation immigrant
women compared to other native women is being exposed to their ancestors’ different cultures. This
study results indicate that there is not a same cultural impact on the married women decisions to
work. Additionally, there is strong evidence of the cultural impact for the second-generation
immigrant women who kept their HLs. The main finding shows female second-generation
immigrants who kept their HLs and whose parents immigrated from higher/lower FLFP countries
work more/less hours in the U.S., respectively.
This section is followed by literature review, theoretical frameworks, and motivations, which is then
succeeded by dataset descriptions and sample selection. Epidemiological approach and results are
presented in subsequent sections. Finally, the last section presents the conclusion.
1. Literature Review, Theoretical Framework and Motivation
FLFP varies notably across countries due to existing different beliefs about women working outside
the home. In some cultures, women who work are appreciated and actively participate in the job
market outside the home. However, in other cultures, women who stay at home and pay attention to
family duties are preferred to those who have jobs outside the home. These different views about a
woman’s role in society is observed significantly across countries [15]. Figure 1 presents the FLFP
for OECD countries in 2013 [16]. This variety of FLFP among countries implies that some countries
strongly depend on the women working, while others do not rely on women’s contributions to their
economics.
**Figure 1**
Female labor supply is a broad topic, and economists explain it using various models from the
sophisticated structural econometric models to more qualitative ones [17–19]. Most of these models
focused on investigating how structural changes in the economy resulted in increased FLFP.
Few studies in the field of Economics have shown that culture impacts women’s decisions to work
outside the home. Reimers (1985) shows ethnic groups have different views about men and women’s
roles in the family and working outside the home. These differences are based on various cultures
and therefore may result in dissimilar utility functions for the various groups [20]. Bisin and Verdier
(1998) find that children born without specific beliefs and attitudes and they achieve preferences
through their observations and experiences they received from their parents and their own
environments [21]. Similarly, Antecol (2000) suggests that the cultural influence is not an
observable variable that impacts FLFP. He finds that the country of ancestry plays a significant role
in the existing inter-ethnic gap in Labor Force Participation (LFP) [22].
Fernández et al. (2004) find a man’s mother working significantly increased the probability of his
wife working in the U.S. [23]. Moreover, Morrill and Morrill (2013) provide evidence that show a
married woman’s LFP is related to both her mother and her mother-in-law’s working experiences
[24]. Additionally, Fernández and Fogli (2009) and Blau et al. (2013) show that the married FLFP is
a kind of economic behavior that transmits from parents to their children; they indicate that second-
generation FLFP depends on FLFP in their parents’ heritage countries [25].
This study examines cultural impacts on women’s decisions to work in the U.S. The results suggest
that the cultural attitudes about women working exist and pass from one generation to the next while
they kept their HLs. Therefore, the role of HL in intergenerational transmission of culture regarding
the women’s decisions to work and the number of hours women worked are examined in several
models to reach accurate and precise results.
2. Datasets and sample selection
The main data source is the 1% 1970 Form 2 Metro sample from the Integrated Public Use
Microdata Series (IPUMS) [26]. The year 1970 is the last year when individuals specifically
answered the question about their parents’ birthplaces and also reported the language spoken in their
home as a child (mother tongue)1. However, IPUMS does not fully specify the country of origin of
the individual’s birth mother when both parents were born outside of the U.S. Therefore, this study
uses father’s birthplace as a cultural proxy for the second-generation immigrant women due to the
data limitation. Consequently, the women include those fathers who were born outside the U.S.,
while their mothers were born in the U.S. Furthermore, this study refines the sample data to the U.S.
women between 25-54 years old. This age range which is chosen due to three main reasons: women
are old enough to seek a full-time job; their education generally has been completed 2; and they have
their own families (as in Staubli 2011; Blau and Kahn 2013).
In this study, heritage country is defined as the father’s birthplace, and HL is identified as the main
languages that are spoken in the father’s birthplace3. Since second-generation immigrant women
were born and raised in the U.S., they certainly are able to speak the English language the same as
other native women4.
This study limits the countries in the main sample. The Soviet Union was a multinational federal
1 Questions about child mother tongue and parents’ birthplaces directly connect to the goals of this study.2 This paper excludes women who are attending to the school.3 This paper focuses on the father’s characteristics due to the lack of data ability for second-generation immigrant mothers.4 Therefore, these women do not have any difficulty in finding a job because of linguistic barriers.
union that was included 16 countries that were governed as a single-party state that existed from
1922 to 1991. Thus, this study eliminates those countries that were parts of the Soviet Union (as in
Fernández and Fogli 2009; Fernández 2007). During the post war years, some countries economic
systems developed in the way of other centrally planned economies; therefore, this study excludes
this group of countries1.
This paper restricts the main sample to the heritage countries2 that have at least 20 observations. The
observations for women whose fathers were born in China are only 36, and this might not represent
China’s population properly; therefore, this paper eliminates China from the main sample. Finally,
the women who do not specify their own mother tongues are excluded from the main sample. The
final sample contains 7526 American women whose fathers were born outside the U.S. from 23
different heritage countries.
This study identifies a woman’s decision about working status based on whether the woman
participated in the labor force. Additionally, to measure the number of hours a woman worked, this
study uses the number of hours a woman was at work during the previous week of her interview 3. To
determine cultural proxies for the number of hours a woman worked, this study employs ancestors’
FLFPs data that are available on the International Labour Organization Statistics Database
(ILOSTAT) [29]. This study suggests that the cultural effect is a kind of dynamic phenomenon that
shifts slowly. Therefore, FLFP in heritage countries in 1945 to 1995 is used to examine culture
transmission through generations during this period. Therefore, this study uses the average annual
FLFP of heritage countries for two quarter-centuries (25 year periods) that starts from 1945. ILO
provides detailed information about all kinds of LFP, and this study uses the FLFP of heritage
countries where the women are between the ages of 25 to 544.
Women also report their mother tongues in IPUMS, and this study links the second-generation
immigrant women to their own HLs. The cross-country data for HL is based on Pearson Education
[30]. There are some countries where people speak more than one specific language; therefore, this
study defines HL as the language spoken by at least 1%5 of a country’s population. Also, there are
some countries where people speak English, and it is not easy to find whether or not second-
generation immigrant women from those countries whose mother tongues are English keep their
1 Albania, Bulgaria, Czechoslovakia, Estonia, Hungary, Poland, Romania, and Yugoslavia.2 Luxembourg, Estonia, Korea, India, Iran, Israel, Palestine, Jordan, Lebanon, Africa, Australia, Turkey, and New Zealand.3 The woman reports the number of hours worked in intervals instead of a certain number of hours, and this study uses the midpoint of each interval as hours they worked in the previous week.4 Total (5-year age bands).5 There are countries where some populations speak English. This paper considers the English language as a HL when more than 5% of population speaks English.
HLs. Therefore, this study assumes that women from United Kingdom, Canada, Ireland, and
Philippines (UCIP) whose mother tongues are English kept their HLs.
Table 1 provides descriptive statistics for the main variables of the second-generation immigrant
women in the sample. The observations contain women whose ancestors were from European,
American, Asian, and Middle Eastern countries. The majority of the observations consist of the
women whose fathers’ heritage countries were Italy, Canada, Germany, and the United Kingdom1.
FLFP for the first quarter (1945-1970) is 22.30% on average and this varies from heritage countries.
The highest value for FLFP of first quarter belongs to the Finland by 36.15%, and Syria has the
lowest value of FLFP by 5.45%. The average of FLFP in the second quarter (1970-1995) is almost
25% more compared to the first quarter.
**Table 1**
3. The Epidemiological Approach and Results
The most important part of cultural analysis in economic decisions is to find a way to examine an
individual’s culture’s impact on economic attitudes. Designing a suitable model to capture the effect
of culture regarding economic outcomes is the most important part of economic analysis. The
epidemiological approach relies on observing second-generation immigrant women’s attitudes in a
host country to isolate the effect of culture from economic conditions and institutions. This approach
suggests that women who were born and live in the same country and face the same law, markets,
and institutions may have different economic decisions regarding their cultures. The main idea of
this approach implies that native women from various cultures that have the same opportunity might
be treated differently regarding the job market. Therefore, examining the work outcome of women
from an identical country, while their fathers were born in different countries has motivated recent
studies to pick out dissimilarities in heritage cultures [4,6,31].
3.1. Empirical Model for FLFP
This study evaluates the cultural effects on a woman’s decision to work outside the home. This study
uses a probit model to investigate on women’s decisions to participate in the labor supply. This
paper uses the cultural proxy as one of the main control variables for the women labor supply, while
considering other key variables in the model as follows:
P (Work ihs=1∨Zihs ,W ihs , S ihs ,F ihs) (1)
1 It contains England, Scotland and Wales- Ireland
Where the variable work shows female participation, which takes the value 1 if the female
participates in the labor supply and 0 otherwise. Z defines the variable that measures the culture
effect by a woman’s heritage country (father’s birthplace country) FLFP, W shows a woman’s
characteristics including her age and educational level, S presents a woman’s spouse’s
characteristics such as his educational level and income, and F represents a woman’s family’s
characteristics such as her family size and the number of children younger than 5 years old.
Additionally, the following estimation model is used to measure precisely the impact of heritage
culture on the number of hours second-generation immigrant women work in the U.S.
Y ihs=α+ β1~ZhT+ β2(
~ZhT∗HL)+∑ a' W ihs+∑ b ' Sihs+∑ c ' F ihs+ƒs+CS+εihs(2)
Where Y ihs is the total number of hours a woman i worked during the previous week whose heritage
country is h and lives in region s. The variable that measures cultural impact is defined as~ZhT . The
value of ~ZhTis determined by a woman’s heritage country (father’s birthplace country) that is FLFP
during time T. Time T defines the average of FLFP for a woman’s ancestor country in the first
quarter (1945-1970) and the second quarter (1970-1995). W ihs shows a woman’s characteristics, Sihs
represents a spousal characteristics, and F ihs represents her family’s characteristics. Since the
working status varies by different regions, this study includes full fixed effects (full set of MSA
effects), ƒs, and full set of heritage country fixed effects, C s, in the model to capture unobservable
factors of heritage countries.
3.2. Women’s decisions to work and cultural proxies
Culture is a kind of dynamic interaction that moves slowly over time. Individuals’ cultures are
created mostly in their early ages, while they often communicate with the surrounding people who
mostly have similar heritage cultures. Therefore, individuals who might not be able to visit their
heritage countries are often able to keep their beliefs updated by direct or indirect contact with
people from their heritage countries.
This study supposes that FLFP of ancestors’ countries may fluctuate over time, while ranking of
countries regarding their FLFPs normally wouldn’t change. There is a positive correlation between
the first quarter of FLFPs and the second quarter of FLFPs that are statistically significant1. This
correlation implies that the heritage cultures for the women’s decisions to work outside the home
may shift over time, while these changes move simultaneously among countries. This study repeats
the cultural analysis with the second quarter data, and the findings about cultural effect are similar to
1 The correlation is 0.72 and it is statistically significant at 1% level.
the first quarter data. There is no evidence for transmitting cultural values regarding participation in
the job market for all women, but the results show that there are significantly positive cultural effects
for the women who kept their HLs during the focused 50 years of the time frame. Since it is difficult
to control all variables that may affect a woman’s decision to work, this study attempts to use main
control variables including a woman’s age and her educational level (High school 1, some college2,
College+3) as well as her husband’s educational level and income. Finally, this study controls family
characteristics including her family size and the number of children under age 5. These variables
helped to measure more precisely the effect of culture on women’s decisions to work.
Table 2 shows the effect of dependent variables on the second-generation immigrant women labor
force participation from 1945 to 1995.
**Table 2**
Column 1 to 3 show the probit results that use first quarter FLFP’s heritage country and other main
explanatory variables for women’s decisions to work in the U.S. The coefficient for the FLFP is
insignificant in column 1 and ancestors’ FLFP cannot explain women’s decisions to work, while
column 2 indicates that the cultural proxy for the women who kept their HLs is economically and
significantly positive. The findings show that a woman who kept her HL and whose father was born
in a heritage country with a higher level of FLFP, most likely worked more than another woman in
the identical country.
The results show that a woman’s educational level is associated with the women’s working statuses.
Adding the educational levels indicates that a woman with a higher educational level has more
opportunity to get a proper job outside the home. Women with higher educational levels prefer to
work more than women with lower educational levels since the opportunity cost for educated women
is higher than uneducated women. Furthermore, husbands’ characteristics may have an effect on the
decision of a woman to work outside the home. A woman who tends to stay at home is more likely
to marry a man with a higher income [32,33]. The results indicate that a woman with an higher
educated, richer husband potentially prefers to work less than other women. Thus, the women’s
decisions to work are negatively affected by their husbands’ educational levels and incomes. Finally,
family variables including the family size and the number of children less than five years old
negatively affected the women’s decisions to work. The findings indicate that cultural effect is
statistically significant for the women who kept their HLs when controlling all other explanatory
1 Below high school and above some college (more than 1 year) omitted.2 Below 1 year of college and above Bachelor’s degree omitted.3 Includes 5+ years of college and above.
variables. Column 4 to column 6 show results for second quarter FLFP’s heritage country for
women’s decisions to work regarding their cultural attitudes over the next 25 years. The results
indicate that culture would slowly change over time. Remarkably, the results show that the cultural
proxy is statistically significant for women who kept their HLs throughout this period.
3.3. Age of women and culture proxies
Data restrictions limit this study to find a year when a woman’s father left his heritage country and
immigrated to the U.S. Therefore, this study uses two main age cohorts of women to link them in the
first quarter of their ancestors’ countries. Accordingly, married women are studied by two age
cohorts: women in the age cohort 25-401 (to be denoted as “young” married women) and women in
the age cohort 40-54 (to be denoted as “middle-aged” women). The value of FLFP’s heritage
countries has been reported since 1945. Morover, fathers of women in the young cohort were
expected to be in the U.S. by 1945, while fathers of women in the middle-aged cohort were likely to
be in the U.S. before 1930. The young women may tend to keep the most recent culture more than
the middle-aged women. In particular, if the female labor supply is related to the period when their
parents left their heritage countries, then one would expect that the recent heritage culture mostly
influences the younger married women compared to the older. Accordingly, the cultural proxies for
the young cohort of married women should be more significant than the middle-aged cohort of
married women when the cultural proxies are not dynamic experiences. Most likely, if women kept
up to date with their heritage cultures, then cultural proxies would be economically and statistically
significant for both cohorts. Results for different age cohorts are reported in Table 3.
**Table 3**
Column 1 and column 3 show cultural proxies are not statistically significant for young and middle-
aged women without considering their HLs. On the other hand, column 2 and column 4 indicate that
cultural proxies can explain women’s decisions to work for both cohorts when they kept their HLs.
The findings indicate that culture is a kind of interaction that is dynamic, and women kept
themselves up to date regarding their heritage cultures.
Previous studies (such as Fernández 2007; Fernández and Fogli 2009; Tabellini 2010; Fernández
2011; Marcén 2014) assume that the effect of heritage culture is the same for the women from the
identical heritage country. However, this study shows that the effect of the heritage culture exists
among those women who kept their HLs. This study reports the women’s decision to work are
positively associated with FLFP’s in their heritage countries for both cohorts when the women kept
1 This study uses an age cutoff of 40 since it is the median age of the sample.
their HLs. Therefore, women’s behaviors regarding the job market are similar for younger and
middle-aged married women during a quarter before 1970 while the results are consistent for a
quarter after 1970. This result strengthens the dynamic effect of heritage culture on the women’s
decisions to work.
3.4. Regression and cultural proxies
This study uses cultural proxies to estimate the number of hours women worked regarding different
specifications. Table 4 reports the main results of cultural proxies for the women with different
specifications regression models.
**Table 4**
Column 1 and column 4 demonstrate regression results when using the cultural proxy for estimation
the number of hours women worked, while column 2 and column 5 adds women’s characteristics to
the primary model. Column 3 and column 6 includes their spousal characteristics for estimating the
number of hours women worked. Finally, column 4 and column 8 show the main model for
estimating the number of hours women worked. Column 1 to column 4 demonstrate the effect of
cultural proxies for the women without considering their HLs, while column 4 to column 8 show the
impact of heritage culture for the women who kept their HLs.
The coefficients for the cultural effect in column 1 to column 4 are not statistically significant
regarding different regression specifications. Previous studies (such as Fernández 2007; Fernández
and Fogli 2009; Tabellini 2010; Fernández 2011) find significant results for cultural proxies for all
women. There are two main possibilities that explain the findings of previous studies. Firstly, they
have not included all main variables for women’s characteristics in their models that may affect the
women’s decisions to work1. Secondly, they have not fixed the effect of unobservable variables for
the countries of origin in their studies. Therefore, countries of origin’ variables may correlate to
cultural proxies. These two biases most likely resulted in overestimating the coefficient for cultural
proxies.
The findings indicate that there is no economically and statistically significant cultural effect for all
women. Meanwhile, there is an economically and statistically significant cultural impact on the
number of hours women worked when they kept their HLs in different specification regression
models. In order to test if the results are driven by the cultural proxies are robust, this study
performed several tests. This study performs robustness checks on alternative new data sets.
Therefore, this study re-estimates the number of hours women worked while excluding some
1 This study uses their models and find similar results, while using all main variables and the fixed effect for heritage countries would change the results and cultural proxies are not statistically significant for all the women.
countries from the main sample2. The robustness analysis indicate that the cultural proxies are
economically and statistically significant to present the number of hours women worked under
different sub-set samples.
3.5. HL’s scenarios
There are some countries that have English language, as their HLs. Women from these countries
may tend to participate U.S. culture due to their potential willingness to join the American society.
The goal of this section is to approximate HL’s role on the transmission of heritage culture under
different scenarios. In this regard, this section exercises five different scenarios that are derived
based on a woman’s culture regarding her mother tongue. Table 5 shows the impact of HL on the
number of hours women worked in a range of different scenarios.
**Table 5**
Each row of Table 5 represents a different scenario that has a main assumption about the women’s
HLs. These scenarios are using the full set of the control variables in their models. Scenario I is
developed based on the following assumption: (1) English language can be a HL for the women
whose fathers were born in UCIP. In this scenario, UCIP’s second-generation immigrant women
whose mother tongues were English are assigned as women who kept their HLs. The scenario I is
the main model that are presented in Table 5. The cultural proxy for the women who kept their HLs
is economically and statistically significant in all levels.
Scenario II assumes that women whose mother tongues are English and whose fathers were born in
UCIP counties did not keep their HLs. These women learned the English language in the U.S. from
their parents and their society. The results demonstrate that cultural proxies about the number of
hours women worked are economically and statistically significant for the women who kept their
HLs. Scenario III suggests that assigning HL for the women originally from UCIP countries is a
controversial issue. Thus, this scenario excludes women who are originally from UCIP countries
from the main sample. Dropping all four English language countries leads to more accurate
estimation across remaining countries regarding transmitting heritage culture through speaking HL.
This scenario shows that the cultural proxy is economically and statistically significant in all levels.
Women whose parents immigrated to the U.S. may have attitudes more similar to the U.S. culture
compared to their heritage cultures. This paper argues that HL can transmit the culture from one
2 The baseline model that includes all the main variables for estimating the number of hours women worked. The second model estimated the number of hours women worked without those second-generation immigrant women whose fathers were born in Italy because Italy is the country with the highest frequently observations. The third model is estimating the model without women whose fathers are from Turkey because Turkey is the country with the highest level of FLFP. The fourth model excludes women whose fathers were born in the Syria because Syria is the country with the lowest level of FLFP. Finally, the fifth model eliminates women whose fathers born in Canada because Canada is the English language country that has highest frequency of observations.
generation to the next generation. If a woman did not keep her HL, she may assimilate to the host
country; thus, women whose mother tongues are English, may tend to the U.S. culture more than
their heritage cultures. Scenario IV assumes that women whose mother tongues are English tend to
the U.S. culture regarding the labor market. This scenario assigns FLFP of the U.S. to women whose
mother tongues are English and whose fathers were born in UCIP countries. The cultural proxy for
the women who kept their HLs is economically and statistically significant.
**Table 6**
Finally, Scenario V assumes that all women whose mother tongues are English more likely tend to
the U.S. culture. Thus, this paper uses FLFP of the U.S. for the women whose mother tongues are
English. This scenario supposes that these women’s culture may tend to the U.S. culture. The
cultural proxy coefficient is presented in the last row of Table 6.
3.6. Social Networks
The existing empirical studies indicate that people’s decisions are associated with ethnic groups that
those people are involved with. People who are exposed to geographic locations connected more
closely with their heritage cultures are more likely to maintain connections with their heritage
cultures. This section provides evidence for second-generation immigrant women who are contacted
by more people from their countries of origin. Figure 2 presents recent data on immigration densities
in various states.
**Figure 2**
Social networks strengthen cultural transmission in the female labor supply. This study uses the
“spoken at home” variable to show the existing social networks among people in the same region.
Second-generation immigrant women who are living in areas where they are exposed to more people
from their countries of origin and who kept their HLs will most likely keep their own heritage
cultures more than women who did not keep their HLs. Similar to the strategy used by Bertrand et al.
(2000) and Furtado et al. (2013), this study examines the cultural effect on the number of hours
women work while considering their social networks. Therefore, this study uses the following model
to measure the effect of social networks on the number of hours women worked.
Y ihs=α+ β1 NEhs+ β2~ZhT∗HL+β3 (~ZhT∗HL∗NEhs )+∑ a' W ihs+∑ b' S ihs+∑ c ' F ihs++CS+εihs
(3)
Where, NEhms shows the social networks that exist among people from the same country of origin (h)
that belong to language group (m) and live in the region (s). The other variables are defined same as
before in the model. The social networks are calculated as follows:
NEhms=ln ¿¿) (4)
Where Chms is the number of people whose heritage country is (h), whose mother tongue is (m), and
who live in region (s), As is the total number of people who live in region (s), Lhm is the total number
of people in the U.S. and their mother tongue is m, and T is the total number of people in the U.S.
This study uses the same datasets with more observation to prevent underestimating the social
networks. Thus, this study redefines the main sample (IPUMS) for social networks. The new sample
consists of people who are between the ages of 25 to 54 and have defined birthplace regions. The
new sample is restricted to the people whose heritage countries are known and defined. Also, the
states in which people were living are defined1. The new sample provides an index for social
networks among people for different geographical areas. This social network index is assigned to the
second-generation immigrant women from the main sample2.
Interaction of cultural effect and density of people from the same country of origin examines the
power of cultural transmission among women who are exposed differently to the people from the
same country of origin. Table 7 presents the results for social network and cultural transmission.
**Table 7**
The findings indicate that existing social network among people with the same culture is resulted to
have more chance to keep given heritage culture.
4. Country of Origin and Unobserved Heterogeneity
This study controls all main observable variables that may affect women’s decisions to work but
unobservable variables may also affect their decisions. The main results may also be accounted by
the characteristics of the country of origin. This section examines the impact of the characteristics of
the country of origin on the number of hours women worked. This study controls a variety of those
characteristics that might impact the second-generation immigrant women’s behaviors in the U.S.
labor market. This section includes characteristics of the country of origin: GDP per capita, human
capital, and women’s attitudes to work3. Table 8 reports the summary of these variables for each
country.1 This paper excludes the U.S. territories: Puerto Rico, State groupings, Military/Mil, Reservations, and District of Columbia.2
This study excludes UCIP countries in the final sample because the main language of these countries is English.3 The results also are robust under the physical distance between the country of origin and the U.S.
**Table 8**
4.1. GDP
The level of GDP among countries of origin may be biased against the findings rather than the
cultural effect. Since women tend to stay at home more to take care of their families, they might
become less involved in production for their own country; however, the second-generation
immigrant women that originally belong to richer countries (higher GDP) may tend to participate on
the production more than other women. Therefore, the results may be concluded that in difference
level of GDP at the country of origin, rather than cultural effect.
This study includes the level of GDP in the country of origin in the model to examine the possibility
that the results are mainly driven by an important aggregate variable of the country of origin. This
study uses three kinds of GDP in the main model: per capita GDP in 1945 and the average per capita
GDP from 1945 to 1970, and per capita GDP in 1970 for the countries of origin. Table 9 presents
results when including GDP per capita in the model.
**Table 9**
The results show that the per capita GDP does not impact the hours that women worked, while the
cultural proxies are statistically significant for those women who kept their HLs. The results are
robust to this specification.
4.2. Human Capital
Human capital is one of the important component in many of the immigrants’ economic outcomes.
The results might arise from omitting a human capital equality variable that could explain the FLFP
in the country of origin. This study uses the human capital index, which is from Barro and Lee
dataset [34] and provides information for educational achievement across countries1. This dataset
was created by Barro and Lee and used human capital as an estimate of the educational attainment
for the population over 15 and 25 years old2. Table 10 shows the results of including human capital
in the regression model.
**Table 10**
The results show that cultural proxies for women who kept their HLs are not changed in magnitude
1 This study uses “Education Attainment by 5-year Age Group” that includes average years of schooling. 2 This study uses average schooling years of the total population as the human capital for 25 years old.
and still are statistically significant, while the coefficients for human capital are not statistically
significant. Thus, cultural proxies for women who kept their HLs remain statistically significant
even after including this variable to the model.
4.3. Women’s attitudes
The World Value Survey (WVS) includes a number of questions that evaluate people’s beliefs and
values across almost 100 countries. The WVS is administered to a representative sample of the
world population from 1981 to 2014 and has been carried out six times (1981-84, 1990-1993, 1995-
97, 1999-2004, 2005-2009 and 2010-2014). The survey contains questions about demographics and
economic characteristics, and specific questions about people’s attitudes, including women’s
attitudes toward work. The main advantage of using WVS is that the same questions on attitudes
and beliefs are asked among countries.
This study uses the fifth wave (2005-2009) of the survey, which covered 58 countries. This wave
includes the main questions about women’s attitudes regarding the job market. This paper focuses on
two key questions that can particularly measure women’s attitudes to work. The first question asked
about prioritizing men in the job market compared to the women: “When jobs are scarce, men
should have more right to a job than women.”1 The second question asked about willingness to work
at home or outside the home: “Being a housewife is just as fulfilling as working for pay.” 2 This
study uses these two questions to measure the women’s attitudes regarding the job market in each
heritage country and provides the “women’s attitude index.”3 Table 11 presents the results of
including women’s attitudes in the model. The results indicate that cultural proxies for women who
kept their HLs are statistically significant, while the coefficients of WVS’s questions (women’s
attitude index) are not statistically significant.
**Table 11**
5. Conclusion
Female labor supply in the economy has been the most meaningful change in labor markets over the
last several decades, especially for married women. A few recent studies propose that culture plays a
significant role in women’s decisions regarding the job market. This study redefines the cultural
proxy and shows that its impact exists through economic decisions, particularly on women’s
1 The individuals have three choices to answer this question: (i) agree, (ii) neither (iii) disagree. This study dropped the individuals who responded with “neither”. 2 The Individuals have four choices to answer this statement (i) strongly disagree, (ii) disagree, (iii) agree, and (iv) strongly agree3 This study provides the women’s attitudes index calculated from the ratio of Individuals who “disagreed” or “strongly disagreed” for those statements.
decisions to work and the number of hours women worked.
This study shows the cultural proxies are economically and statistically significant in explaining the
women’s decisions to work whenever they kept their HLs. Even as the heritage culture changes its
values over time, women, who kept their HLs, clearly choose their participation according to their
culture’s trend. This study shows cultural effect is a dynamic interaction. The empirical results are
consistent with the expectations regarding HL’s role in cultural transmission. There is considerable
evidence of intergenerational transmission in the female labor supply regarding women’s attitudes
toward keeping their HLs. Different estimation models are employed to show that an ancestor’s
FLFP is more likely to have an impact on the number of hours a woman works when she is able to
make the connection to her own heritage culture. This study shows that cultural transmission is
statistically significant for women whose social networks are stronger than other women’s. Finally,
the results are consistent when controlling country of origin and unobserved heterogeneity among
those women. This study proposes the new model to quantitative impact of culture, which is the
main factor for economic decisions. This paper shows culture plays an important role in explaining
the large variation of FLFP across time and countries. There are other possible interpretations that
might exist for these empirical results, while they would not explain why the impact of the cultural
proxy is statistically significant in all models.
Moreover, policy makers and planners make different strategies regarding female labor supply and
immigration policies in the labor market. They can use the results of this study to better estimate
future LFP of women from different cultures with various approaches to keep women’s heritage
culture. Finally, this paper suggests that the second-generation immigrant who kept their HLs can
represent their heritage cultures.
References[1] Fernández R, Fogli A. Culture: An Empirical Investigation of Beliefs, Work, and Fertility. American
Economic Journal: Macroeconomics 2009;1:146–77.[2] Fogli A, Veldkamp L. Nature or Nurture ? Learning and the Geography of Female Labor Force
Participation. Econometrica 2011;79:1103–38.[3] Marcén M. The role of culture on self-employment. Economic Modelling 2014.[4] Fernández R. Does Culture Matter ? In: Benhabib J, Bisin A, Jackson MO, editors. Handbook of
Social Economics, 2011, p. 481–510.[5] Guiso L, Sapienza P, Zingales L. Does Culture Affect Economic Outcomes ? The Journal of
Economic Perspectives 2006;20:23–48.[6] Fernández R. Alfred Marshall lecture women, work, and culture. Journal of the European Economic
Association 2007;5:305–32.[7] Guiso L, Sapienza P, Zingales L. Cultural Biases in Economic Exchange? The Quarterly Journal of
Economics 2009;124:1095–131.[8] Luttmer EFP, Singhal M, Alesina A, Glaeser E. Culture, Context, and the Taste for Redistribution.
American Economic Journal: Economic Policy 2011;3:157–79.[9] Alesina A, Giuliano P. The power of the family. Journal of Economic Growth 2010;15:93–125.[10] Tabellini G. Culture and institutions: economic development in the regions of Europe. Journal of the
European Economic Association 2010;8:677–716.[11] Tsai KM, Park H, Liu LL, Lau AS. Distinct pathways from parental cultural orientation to young
children’s bilingual development. Journal of Applied Developmental Psychology 2012;33:219–26.[12] Hughes D, Rodriguez J, Smith EP, Johnson DJ, Stevenson HC, Spicer P. Parents’ ethnic-racial
socialization practices: a review of research and directions for future study. Developmental Psychology 2006;42:747–70.
[13] Fillmore LW. When learning a second language means losing the first. Early Childhood Research Quarterly 1991;6:323–46.
[14] Kim SY, Chao RK. Heritage language fluency, ethnic identity, and school effort of immigrant Chinese and Mexico adolescents. Cultural Diversity & Ethnic Minority Psychology 2009;15:27–37.
[15] Alesina AF, Giuliano P, Nunn N. On the Origins of Gender Roles: Women and the Plough. The Quarterly Journal of Economics 2013;128:469–530.
[16] OECD. OECD Stat, (database) 2014.[17] Fernández R. Cultural Change as Learning: The Evolution of Female Labor Force Participation over a
Century. The American Economic Review 2013;103:472–500.[18] Euwals R, Knoef M, Vuuren D van. The Trend in Female Labour Force Participation: What Can Be
Expected for the Future? Empirical Economics 2011;40:729–53.[19] Goldin C. The Quiet Revolution That Transformed Women ’ s Employment , Education , and Family.
The American Economic Review 2006;96:1–21.[20] Reimers CW. Cultural Differences in Labor Force Participation Among Married Women. The
American Economic Review 1985;75:251–5.[21] Bisin A, Verdier T. On the cultural transmission of preferences for social status. Journal of Public
Economics 1998;70:75–97.[22] Antecol H. An examination of cross-country differences in the gender gap in labor force participation
rates. Labour Economics 2000;7:409–26.[23] Fernández R, Fogli A, Olivetti C. Mothers and Sons : Preference Formation and Female Labor Force
Dynamics. The Quarterly Journal of Economics 2004;4:1249–99.[24] Morrill MS, Morrill T. Intergenerational links in female labor force participation. Labour Economics
2013;20:38–47.[25] Blau FD, Kahn LM, Liu AY-H, Papps KL. The transmission of women’s fertility, human capital, and
work orientation across immigrant generations. Journal of Population Economics 2013;26:405–35.[26] Ruggles S, Alexander JT, Genadek K, Goeken R, Schroeder MB, Sobek. M. No Title. Integrated
Public Use Microdata Series: Version 50 [Machine-Readable Database] Minneapolis: University of Minnesota 2010.
[27] Blau FD, Kahn LM. Female labor supply: Why is the US falling behind? The American Economic Review 2013;103:251–6.
[28] Staubli S. The impact of stricter criteria for disability insurance on labor force participation. Journal of
Public Economics 2011;95:1223–35.[29] ILOSTAT Database (Geneva). International Labour Office (ILO) 2013.[30] Languages Spoken in Each Country of the World. Pearson Education, Publishing as Infoplease 2007.[31] Furtado D, Marcén M, Sevilla A. Does culture affect divorce? evidence from European immigrants in
the United States. Demography 2013;50:1013–38.[32] Lundberg S. Labor Supply of Husbands and Wives: A Simultaneous Equations Approach. The
Review of Economics and Statistics 1988;70:224–35.[33] Blau FD, Kahn LM. Changes in the Labor Supply Behavior of Married Women : 1980 – 2000. Journal
of Labor Economics 2007;25:393–438.[34] Barro R, Lee J-W. A New Data Set of Educational Attainment in the World, 1950-2010. Journal of
Development Economics 2010;104:184–98.
Table 1. Data descriptive by heritage countries.County Obs Hours
workFLFP
1945-1970
FLFP
1970-1995
Age Family size
Children HL High School
Some
CollegeCollege +
Canada 1,084 12.70 19.70 31.10 39.73 4.30 2.20 0.95 0.49 0.19 0.03
Mexico 460 10.40 10.90 15.90 35.95 5.49 3.25 0.93 0.27 0.09 0.00
Cuba 25 12.08 11.95 23.00 38.08 3.92 1.80 0.76 0.60 0.20 0.00
Denmark 164 14.04 31.50 46.42 40.78 4.22 2.13 0.24 0.54 0.20 0.04
Finland 82 15.77 36.15 43.22 42.13 4.15 2.00 0.62 0.63 0.20 0.00
Norway 270 14.05 19.90 42.85 42.63 3.96 1.88 0.39 0.47 0.26 0.02
Sweden 341 14.10 29.77 44.92 42.01 3.96 1.86 0.34 0.54 0.26 0.03
United Kingdom 814 13.14 30.95 34.35 41.02 3.98 1.88 0.95 0.51 0.25 0.04
Ireland 332 10.69 21.00 21.27 41.48 4.51 2.38 0.98 0.61 0.18 0.04
Belgium 42 10.44 20.27 28.50 39.31 4.36 2.26 0.21 0.57 0.10 0.05
France 108 11.68 27.95 24.37 39.81 4.33 2.22 0.37 0.56 0.23 0.02
Netherlands 155 11.19 17.80 19.00 40.88 4.36 2.31 0.28 0.48 0.24 0.02
Switzerland 81 13.40 28.57 37.85 41.64 4.12 2.00 0.37 0.48 0.20 0.05
Greece 145 10.27 20.45 21.07 38.92 4.28 2.17 0.46 0.54 0.19 0.03
Italy 1,814 11.30 20.60 23.15 39.45 4.31 2.17 0.59 0.54 0.14 0.02
Portugal 109 12.51 16.30 32.25 39.26 4.31 2.27 0.54 0.43 0.14 0.01
Spain 48 10.51 12.90 16.50 37.56 4.54 2.42 0.58 0.50 0.23 0.04
Austria 346 11.61 35.50 34.23 41.49 4.22 2.12 0.18 0.53 0.18 0.04
Germany 940 14.02 31.57 35.10 42.39 4.01 1.91 0.40 0.53 0.18 0.02
Japan 78 19.33 35.38 37.35 35.74 4.22 1.90 0.74 0.54 0.29 0.08
Philippines 47 13.04 18.80 14.10 33.94 4.66 2.53 0.85 0.53 0.17 0.02
Lebanon 32 12.08 9.50 9.50 37.50 4.75 2.59 0.63 0.56 0.28 0.03
Syria 25 10.82 5.45 4.10 39.48 3.84 1.80 0.36 0.52 0.16 0.08
Average 327.91 12.57 22.30 27.83 39.62 4.30 2.18 0.55 0.52 0.20 0.03
Std. Dev 441.74 2.10 8.98 11.76 2.31 0.35 0.33 0.26 0.07 0.05 0.02Source: IPUMS data U.S. Census, 1% 1970 from 2 Metro sample and ILO.
Table 2. The effects of the women’s cultures on the probability of female labor participation.(1) (2) (3) (4) (5) (6)
FLFP (1945-1970)0.016
(0.186)-0.011(0.019)
HL*FLFP (1945-1970)0.005
(0.001)***0.005
(0.001)***
FLFP (1970-1995)0.009
(0.010)0.004
(0.010)
HL*FLFP (1970-1995))0.004
(0.001)***
0.004(0.001)***
Age0.009
(0.022)0.008
(0.022)0.008
(0.022)0.009
(0.022)0.008
(0.022)0.008
(0.022)
Age square / 100-0.014
(-0.027)-0.014(0.027)
-0.014(0.027)
-0.014(0.027)
-0.013(0.027)
-0.013(0.027)
High school0.217
(0.039)***0.219
(0.039)***0.219
(0.039)***0.217
(0.039)***0.219
(0.039)***
0.219(0.039)***
Some college0.399
(0.055)***0.402
(0.055)***0.402
(0.055)***0.399
(0.054)***0.402
(0.054)***
0.402(0.054)***
College+1.092
(0.111)***1.094
(0.111)***1.094
(0.111)***1.092
(0.111)***1.093
(0.111)***
1.093(0.111)***
Husband high school-0.095
(0.039)**-0.095
(0.039)**-0.095
(0.039)**-0.095
(0.039)**-0.095
(0.039)**-0.095
(0.039)**
Husband some college-0.21
(0.048)***-0.212
(0.048)***-0.212
(0.048)***-0.21
(0.049)***-0.213
(0.048)***
-0.213(0.048)***
Husband college+-0.414
(0.073)***-0.416
(0.073)***-0.025
(0.003)***-0.414
(0.073)***-0.417
(0.073)***
-0.417(0.073)***
Husband’s income-0.025
(0.003)***-0.025
(0.003)***-0.025
(0.003)***-0.025
(0.003)***-0.025
(0.003)***
-0.025(0.003)***
Family size-0.064
(0.011)***-0.065
(0.011)***-0.065
(0.011)***-0.064
(0.011)***-0.065
(0.011)***
-0.065(0.011)***
Child < 5-0.553
(0.034)***-0.553
(0.034)***-0.553
(0.034)***-0.553
(0.034)***-0.553
(0.034)***
-0.553(0.034)***
Observations 7526 7526 7526 7526 7526 7526Pseudo R2 0.0937 0.0949 0.0949 0.0937 0.0948 0.0948
***Statistically significant at the 1% level**Statistically significant at the 5% level*Statistically significant at the 10% level
Table 3. Cultural proxies and the probability of women decision to work by different age cohorts.Women 25 to 40 years old women 40 to 54 years old
(1) (2) (3) (4) (5) (6)
FLFP 0.003(0.027)
0.009(0.028)
-0.024(0.028)
-0.019(0.028)
HL*FLFP 0.005(0.002)**
0.005(0.002)**
0.005(0.002)***
0.005(0.002)***
Age -0.303(0.095)***
-0.304(0.095)***
-0.304(0.095)***
0.147(0.119)
0.148(0.119)
0.148(0.119)
Age square / 100 0.466(0.146)***
0.467(0.146)***
0.467(0.146)***
-0.168(0.126)
-0.169(0.126)
-0.169(0.126)
High school 0.111(0.065)*
0.113(0.065)*
0.113(0.065)*
0.276(0.051)***
0.278(0.051)***
0.278(0.051)***
Some college 0.282(0.087)***
0.286(0.087)***
0.286(0.087)***
0.480(0.073)***
0.482(0.073)***
0.482(0.073)***
College+ 0.928(0.160)***
0.929(0.160)***
0.929(0.160)***
1.251(0.164)***
1.257(0.164)***
1.257(0.164)***
Husband high school -0.098(0.062)
-0.098(0.062)
-0.098(0.062)
-0.112(0.051)**
-0.112(0.051)**
-0.112(0.051)**
Husband some college -0.217(0.075)***
-0.218(0.075)***
-0.218(0.075)***
-0.187(0.066)***
-0.189(0.066)***
-0.189(0.066)***
Husband college+ -0.356(0.105)***
-0.357(0.105)***
-0.357(0.105)***
-0.426(0.106)***
-0.428(0.106)***
-0.428(0.106)***
Husband’s income -0.033(0.005)***
-0.033(0.005)***
-0.033(0.005)***
-0.024(0.003)***
-0.024(0.003)***
-0.024(0.003)***
Family size -0.084(0.018)***
-0.083(0.018)***
-0.083(0.018)***
-0.053(0.014)***
-0.054(0.014)***
-0.054(0.014)***
Child < 5 -0.525(0.038)***
-0.526(0.038)***
-0.526(0.038)***
-0.683(0.096)***
-0.677(0.096)***
-0.677(0.096)***
Observations 3433 3433 3433 4045 4045 4045Pseudo R2 0.1380 0.1389 0.1389 0.0717 0.0729 0.0729
***Statistically significant at the 1% level**Statistically significant at the 5% level*Statistically significant at the 10% level
Table 4. Culture and women hours worked. (1) (2) (3) (4) (5) (6) (7) (8)
FLFP0.087
(0.249)0.096
(0.247)0.105
(0.244)0.192
(0.239)0.028
(0.250)0.031
(0.248)0.040
(0.245)0.127
(0.240)
FLFP*HL0.050
(0.020)**0.054
(0.017)***0.054
(0.019)***0.054
(0.019)***
Age0.032
(0.250)0.363
(0.249)**0.079
(0.270)0.020
(0.250)0.351
(0.249)0.071
(0.270)
Age square / 1000.286
(0.313)-0.107(0.311)
-0.075(0.335)
0.303(0.313)
-0.091(0.313)
-0.063(0.335)
High school1.898
(0.489)***3.169
(0.508)***3.097
(0.497)***1.919
(0.489)***3.192
(0.508)***3.120
(0.682)***
Some college1.941
(0.627)***5.410
(0.698)***5.293
(0.683)***1.959
(0.627)***5.433
(0.698)***5.315
(0.683)***
College+8.160
(1.304)***12.897
(1.369)***12.583
(1.340)***8.169
(1.304)***12.909
(1.369)***12.593
(1.340)***
Husband high school-0.910
(0.5106)*-1.048
(0.494)**-0.911
(0.506)*-1.050
(0.494)**
Husband some college-2.413
(0.622)***-2.374
(0.608)***-2.427
(0.621)***-2.389
(0.608)***
Husband college+-5.663
(0.901)***-5.372
(0.881)***-5.666
(0.90)***-5.376
(0.881)***
Husband’s income-0.327
(0.031)***-0.285
(0.031)***-0.327
(0.031)***-0.285
(0.031)***
Family size-0.927
(0.131)***-0.931
(0.131)***
Child < 5-5.345
(0.372)***-5.338
(0.371)***
Observations 7542 7542 7542 7542 7542 7542 7542 7542Adjusted R2 0.0097 0.0270 0.0527 0.0963 0.0104 0.0278 0.0536 0.0973
***Statistically significant at the 1% level**Statistically significant at the 5% level*Statistically significant at the 10% level
Table 5. Scenario's description and type.Scenarios Implementation Description
Scenario I Baseline Assign English Language as the HL to women whose fathers were born in UCIP countries.
Scenario II English language mainly from the U.S Define HL is not English for women whose fathers were born in UCIP countries.
Scenario III Exclusion of UCIP countries Exclude the UCIP countries from the sample.
Scenario IV UCIP women’s who speak English consider as U.S. women
Assign FLFP of the U.S. to women whose mother tongues are English and whose fathers were born in UCIP countries.
Scenario V Women who speak English consider as the U.S. women
Replace FLFP of the U.S. for women whose mother tongues are English.
Table 6. Languages and cultures in alternative scenarios
Scenario I Scenario II Scenario III Scenario IV Scenario V
FLFP 0.127(0.240)
0.183(0.238)
0.475(0.645)
0.127(0.240)
0.181(0.239)
FLFP*HL0.055
(0.019)***0.054
(0.019)***0.055
(0.019)***0.055
(0.020)***0.026
(0.030)
Observations 7542 7558 5265 7542 7542Adjusted R2 0.0973 0.0972 0.0948 0.0973 0.0963***Statistically significant at the 1% level**Statistically significant at the 5% level*Statistically significant at the 10% level
Table 7. Social networks and culture(1) (2) (3) (4)
Density HL in States0.087
(0.035)**-0.053(0.104)
-0.001(0.142)
FLFP*HL0.083
(0.058)0.139
(0.119)
FLFP*HL* Density HL in States-0.006(0.011)
0.004(0.001)***
Age0.658
(0.327)**0.655
(0.327)**0.654
(0.327)**0.657
(0.327)**
Age square / 100-0.790
(0.407)*-0.787
(0.407)*-0.786
(0.407)*-0.788
(0.407)*
High school3.259
(0.586)***3.255
(0.586)***3.257
(0.586)***3.257
(0.586)***
Some college4.989
(0.823)***4.986
(0.823)***4.988
(0.823)***4.988
(0.823)***
College+10.839
(1.720)***10.887
(1.720)***10.873
(1.720)***10.867
(1.720)***
Husband high school-1.180
(0.585)**-1.181
(0.585)**-1.178
(0.585)**-1.182
(0.585)**
Husband some college-2.269
(0.732)***-2.285
(0.732)***-2.298
(0.733)***-2.268
(0.732)***
Husband college+-6.566
(1.099)***-6.616
(1.099)***-6.632
(1.10)***-6.576
(1.099)***
Husband’s income-0.257
(0.037)***-0.256
(0.037)***-0.256
(0.037)***-0.257
(0.037)***
Family size-0.964
(0.155)***-0.965
(0.155)***-0.969
(0.155)***-0.963
(0.155)***
Child < 5-4.981
(0.453)***-4.979
(0.453)***-4.973
(0.453)***-4.984
(0.453)***
Observations 5265 5265 5265 5265Adjusted R2 0.0945 0.0947 0.0946 0.0947
***Statistically significant at the 1% level**Statistically significant at the 5% level*Statistically significant at the 10% level
Table 8. Summary of variables for origin countries.
County GDP 1945
GDP 1945-1970
GDP 1970
H.Capital 1950
H.Capital 1950-1970
H.Capital 1970 WVS-Q1 WVS-Q2 WVS-Q1&2
Canada 7133 8872 12050 7.39 8.04 8.66 77.90 2.40 40.15
Mexico 2134 3023 4320 2.17 2.42 2.89 67.60 2.90 35.25
Finland 3450 5896 9577 3.82 4.38 5.4 81.50 0.80 41.15
Norway 3980 6968 10027 7.54 7.72 8.27 88.60 18.0 53.30
Sweden 5454 8660 13011 6.44 7.18 7.96 94.10 8.80 51.45
United Kingdom 7056 8336 10767 6.11 6.68 7.54 76.10 5.20 40.65
France 2573 7019 11410 4.31 4.46 4.96 73.80 17.60 45.70
Netherlands 2686 7714 11967 6.03 6.7 8.03 81.40 11.60 46.50
Switzerland 7752 11842 16904 8.66 9.08 9.76 62.90 7.20 35.05
Italy 1922 5598 9719 4.04 4.61 5.17 59.20 5.06 32.40
Spain 2102 3471 6319 3.47 3.81 4.34 76.00 22.10 49.05
Japan 1346 4044 9714 5.91 6.68 7.08 17.90 0.40 9.15
Cuba 1776 2052 1917 3.44 4 4.83
Denmark 5066 8588 12686 5.39 4.99 5.79
Ireland 3019 4246 6199 6.09 6.56 7.01
Belgium 4333 6898 10611 6.52 6.91 7.19
Greece 938 3154 6211 3.80 5.56 6.04
Portugal 1804 3030 5473 1.62 2.01 2.49
Austria 1725 5656 9747 6.42 6.62 7.63
Germany 4514 6564 10839 6.71 7.42 7.71
Philippines 646 1384 1764 1.88 3.03 4.03
Syria 2409 3246 3540 0.61 0.94 1.36
Lebanon 2429 2571 2917
Average 3315.09 5601.39 8595.17 4.93 5.45 6.1 16.1 66 41
Std. Dev 2023.85 2693.42 3993.52 2.15 2.14 2.18 7.56 14.66 9.47
Table 9. GDP of origin countries and women’s work(1) (2) (3) (4) (5) (6)
GDP 19452.53
(3.133)1.672
(3.146)
GDP1945-19702.727
(3.383)1.805
(3.396)
GDP 19702.238
(2.777)1.481
(2.788)
FLFP*HL0.055
(0.019)***0.055
(0.019)***0.055
(0.019)***
Age0.079
(0.270)0.079
(0.270)0.079
(0.270)0.071
(0.270)0.071
(0.270)0.071
(0.270)
Age square / 100-0.075(0.335)
-0.075(0.335)
-0.075(0.335)
-0.063(0.335)
-0.063(0.335)
-0.063(0.335)
High school3.097
(0.497)***3.097
(0.497)***3.097
(0.497)***3.120
(0.497)***3.120
(0.497)***3.120
(0.497)***
Some college5.293
(0.683)***5.293
(0.683)***5.293
(0.683)***5.315
(0.683)***5.315
(0.683)***5.315
(0.683)***
College+12.583
(1.341)***12.583
(1.341)***12.583
(1.341)***12.592
(1.340)***12.592
(1.340)***12.593
(1.340)***Husband high school
-1.048(0.494)**
-1.048(0.494)**
-1.048(0.494)**
-1.050(0.494)**
-1.050(0.494)**
-1.050(0.494)**
Husband some college
-2.374(0.608)***
-2.374(0.608)***
-2.374(0.608)***
-2.389(0.608)***
-2.389(0.608)***
-2.389(0.608)***
Husband college+-5.372
(0.881)***-5.372
(0.881)***-5.372
(0.881)***-5.376
(0.881)***-5.376
(0.881)***-5.376
(0.881)***
Husband’s income-0.285
(0.031)***-0.285
(0.031)***-0.285
(0.031)***-0.285
(0.031)***-0.285
(0.031)***-0.285
(0.031)***
Family size-0.927
(0.131)***-0.927
(0.131)***-0.927
(0.131)***-0.932
(0.131)***-0.932
(0.131)***-0.932
(0.131)***
Child < 5-5.345
(0.372)***-5.345
(0.372)***-5.345
(0.372)***-5.338
(0.371)***-5.338
(0.371)***-5.338
(0.371)***
Observations 7542 7542 7542 7542 7542 7542
Adjusted R2 0.0963 0.0963 0.0963 0.0973 0.0973 0.0973
***Statistically significant at the 1% level**Statistically significant at the 5% level*Statistically significant at the 10% level
Table 10. Human capital for countries of origin and women’s work.(1) (2) (3) (1) (2) (3)
H.Capital 19500.071
(0.116)0.041
(0.128)
H.Capital 1950-19700.085
(0.122)0.014
(0.117)
H.Capital 19700.071
(0.116)0.028
(0.123)
HL*FLFP0.046
(0.016)***0.046
(0.016)***0.046
(0.016)***
Age0.041
(0.270)0.044
(0.268)0.041
(0.268)0.049
(0.268)0.046
(0.268)0.048
(0.268)
Age square / 100-0.029(0.333)
-0.033(0.333)
-0.029(0.333)
-0.038(0.333)
-0.032(0.332)
-0.035(0.333)
High school3.067
(0.491)***3.064
(0.491)***3.067
(0.491)***3.062
(0.491)***3.069
(0.491)***3.065
(0.491)***
Some college5.220
(0.679)***5.214
(0.679)***5.218
(0.679)***5.206
(0.678)***5.217
(0.678)***5.212
(0.678)***
College+12.525
(1.336)***12.518
(1.336)***12.525
(1.336)***12.487
(1.336)***12.503
(1.336)***12.495
(1.336)***
Husband high school-1.082
(0.493)**-1.084
(0.493)**-1.083
(0.493)**-1.093
(0.492)**-1.088
(0.492)**-1.091
(0.492)**
Husband some college-2.394
(0.607)***-2.397
(0.607)***-2.394
(0.607)***-2.438
(0.606)***-2.429
(0.607)***-2.433
(0.606)***
Husband college+-5.390
(0.880)***-5.391
(0.880)***-5.390
(0.880)***-5.444
(0.880)***-5.438
(0.880)***-5.441
(0.880)***
Husband’s income-0.286
(0.031)***-0.287
(0.030)***-0.287
(0.030)***-0.286
(0.031)***-0.286
(0.031)***-0.286
(0.031)***
Family size-0.912
(0.130)***-0.911
(0.130)***-0.912
(0.130)***-0.915
(0.130)***-0.916
(0.130)***-0.915
(0.130)***
Child < 5-5.435
(0.372)***-5.434
(0.372)***-5.435
(0.372)***-5.422
(0.372)***-5.424
(0.372)***-5.422
(0.372)***
Observations 7510 7510 7510 7510 7510 7510
Adjusted R2 0.0989 0.0989 0.0989 0.0998 0.0997 0.0998
***Statistically significant at the 1% level**Statistically significant at the 5% level*Statistically significant at the 10% level
Table 11. Attitudes and women’s decisions to work.(1) (2) (3) (4) (5) (6)
WVS-Q1-0.004(0.003)
-0.003(0.003)
WVS-Q2-0.113(0.104)
-0.103(0.104)
WVS- Q1&2-0.007(0.007)
-0.007(0.007)
FLFP*HL0.004
(0.002)*0.004
(0.002)*0.004
(0.002)*
Age0.008
(0.026)0.008
(0.026)0.008
(0.026)0.008
(0.026)0.008
(0.026)0.008
(0.026)
Age square / 100-0.014(0.032)
-0.014(0.032)
-0.014(0.032)
-0.013(0.032)
-0.013(0.032)
-0.013(0.032)
High school0.179
(0.047)***0.179
(0.047)***0.179
(0.047)***0.180
(0.047)***0.180
(0.047)***0.180
(0.047)***
Some college0.401
(0.066)***0.401
(0.066)***0.401
(0.066)***0.401
(0.066)***0.401
(0.066)***0.401
(0.066)***
College+1.063
(0.135)***1.063
(0.135)***1.063
(0.135)***1.057
(0.135)***1.057
(0.135)***1.057
(0.135)***
Husband high school-0.117
(0.047)**-0.117
(0.047)**-0.117
(0.047)**-0.116
(0.047)**-0.116
(0.047)**-0.116
(0.047)**
Husband some college-0.235
(0.058)***-0.235
(0.058)***-0.235
(0.058)***-0.235
(0.058)***-0.235
(0.058)***-0.235
(0.058)***
Husband college+-0.490
(0.088)***-0.490
(0.088)***-0.490
(0.088)***-0.489
(0.088)***-0.489
(0.088)***-0.489
(0.088)***
Husband’s income-0.027
(0.003)***-0.027
(0.003)***-0.027
(0.003)***-0.027
(0.003)***-0.027
(0.003)***-0.027
(0.003)***
Family size-0.074
(0.013)***-0.074
(0.013)***-0.074
(0.013)***-0.074
(0.012)***-0.074
(0.012)***-0.074
(0.012)***
Child < 5-0.546
(0.040)***-0.546
(0.040)***-0.546
(0.040)***-0.545
(0.039)***-0.545
(0.039)***-0.545
(0.039)***
Observations 5298 5928 5928 5298 5298 5298Pseudo R2 0.1007 0.1007 0.1007 0.1012 0.1012 0.1012
***Statistically significant at the 1% level**Statistically significant at the 5% level*Statistically significant at the 10% level
Turkey
Mexico Ita
lyChile
KoreaGree
ce
Hungary
Polan
d
Belgium
Slovak
Repub
lic
Irelan
d
Luxe
mbourg
Japan
Czech R
epub
lic
Sloven
iaFra
nce
United St
atesIsr
aelSp
ain
Portu
gal
Australi
a
Austria
Eston
ia
United Kingd
om
German
y
New Ze
aland
Finlan
d
Canad
a
Netherl
ands
Denmark
Norway
Switz
erlan
d
Swed
en
Icelan
d20
30
40
50
60
70
80
90
Country
Fem
ale
labo
r fo
rce
part
icip
atio
n
Figure 1. Female labor force participation for OECD countries in 2013.
Figure 2. Immigration densities in the U.S.