Time use of youths by immigrant and native-born parents: ATUS results

Abstract

A study based on the American Time Use Survey finds that, although native-born and immigrant youths pass their days in similar ways, Latino and Asian immigrant youths spend more time studying and less time in paid employment than do native-born youths; more time devoted to study may be a mechanism by which immigrants achieve educational mobility

Classical and contemporary literature on immigration has been driven by questions concerning how and when immigrants assimilate to American society. Understanding the assimilation process is especially important with regard to adolescents, whose trajectory will drive the future incorporation of immigrant groups into U.S. society.

The literature presents two competing stories regarding the behaviors and life chances of immigrant youths. The dominant theory, called segmented assimilation, proposes that immigrant youths face a segmented path to assimilation, based on (1) the conditions of their parents’ departure from the home country; (2) their parents’ initial human capital; (3) the “mode of incorporation” that immigrants experience in the host country, including federal, state, and local immigration policies as well as reception by native groups; (4) cultural and economic barriers, including racial discrimination and, in the United States, the increasingly bifurcated labor market; and (5) the family and community resources that are available to confront such barriers.¹ Second-generation youths who have parents with high levels of initial human capital, who receive a positive reception by native groups, and who have access to strong co-ethnic communities (where resources developed by earlier immigrants are available) are poised for upward mobility. Many Asian immigrant groups fit this assimilation pathway. By contrast, a substantial portion of Latin American and Caribbean immigrants have parents with low human capital, are received less positively by the host country, have access to weaker co-ethnic communities, and often live in areas mired by poverty, crime, and negative peer influence; their children are at risk of falling behind.²

In contrast to segmented assimilation theory, classical and neoclassical assimilation theory highlights a more positive conclusion: that, for the most part, immigrant youths successfully assimilate to mainstream American society and experience upward mobility compared with their parents, despite different starting points among immigrant groups.³ Adherents of classical assimilation admit that some immigrant groups experience downward mobility, but they emphasize that there may also be advantages to the second generation that stem from membership in an immigrant group.⁴ A study of second-generation immigrants in New York, for example, found that Russians and Chinese were doing better economically than native-born Whites.⁵

Adolescence is a time of increased peer influence, and one central question about youths with an immigrant background is the extent to which they adopt behaviors similar to those of their peers in the host country. One measure of this adaptation is how similarly immigrant youths structure their daily lives relative to their native-born peers. This article draws on American Time Use Survey (ATUS) data to shed light on the immigration assimilation process by examining the time use of native-born youths and youths from different immigrant backgrounds in order to discern possible strategies undertaken in the teenage years that might help explain successful assimilation in the adult years.

Immigrant teens and the assimilation debate

Immigrant youths are the fastest growing segment of the U.S. population under age 18, representing almost a fourth of all U.S. youths.⁶ Immigrant children start out in vastly different socioeconomic positions: poverty rates range from 9.5 percent among non-Latino White immigrant children, to 14.7 percent for Asians, 18.7 percent for non-Latino Africans, and 32.9 percent for Mexican immigrant children.⁷ Given these different starting points, how well are immigrant children adapting to U.S. society?

On the whole, studies find that second-generation immigrants do much better than their parents in educational attainment, even when the educational attainment of the first generation is very low.⁸ For example, 4.2 percent of first-generation Mexican immigrants are college graduates, whereas almost 15 percent of Mexican-Americans graduate from college by the second generation.⁹ The second generation also is less concentrated in low-wage jobs than is the first generation. Among Mexican first-generation immigrants, a full 79 percent are in low-wage jobs, but the percentage drops quite substantially, to 37 percent, by the second generation.¹⁰ Because some immigrant groups have even better outcomes than native-born Whites and almost all immigrant groups do better than their own parents, some scholars in the classical or neoclassical assimilation camp posit that there is a pattern of “second-generation advantage.”¹¹ One mechanism by which immigrant youths may achieve second-generation advantage is education; for example, Andrew Fuligni found that, regardless of their country of origin, youths from immigrant families tend to place a higher value on doing well in school and tend to work harder in school than do youths from U.S.-born families.¹²

At the same time, other indicators suggest variation in the rates of successful incorporation into American society. For instance, a study by Joel Perlmann found that while 9 percent of White males and 16 percent of Black males dropped out of high school in 2000, the rate was 33 percent for second-generation Mexican-American males.¹³ The same study found that females of Mexican origin are especially likely to give birth at a young age—48 percent before age 24, a figure that exceeds the rate of Black females (41 percent).¹⁴ Similarly, Mexican and West Indian immigrant males have a higher incidence of arrests and incarceration than do other immigrant groups.¹⁵ Finally, in a study of the 10 largest ethnic groups of foreign parentage in the United States, Ruben Rumbaut and Golnaz Komaie found that, whereas almost all groups experienced upward educational and job mobility, second-generation Koreans did worse than the first generation on both measures:¹⁶ while 66.6 percent of first-generation Koreans have college degrees, the percentage goes down to 59 percent for the second generation, and although 15.1 percent of first-generation Koreans have low-wage jobs, the figure increases to 18.8 percent in the second generation (however, overall poverty rates decline for the second generation).

In addition, some qualitative works suggest that the generally positive relationship between educational achievement and immigrant parentage does not hold for some immigrant groups, such as native-born Mexican-Americans, who feel excluded from and exploited by American society; these youths see academic achievement as “acting White” and the educational system as a medium for continued exploitation.¹⁷ John Ogbu offers a general theory to explain different outcomes by linking the degree to which an immigrant group assimilates to the group's social status in the receiving society. He argues that voluntary minorities (those who immigrated voluntarily for better opportunities, as well as their descendants) are more successful partly because they view adapting to the dominant cultural practices as additive and nonthreatening to their cultural identity. By contrast, says Ogbu, involuntary minorities (those who were conquered, colonized, or enslaved, as well as their descendants) view adaptation as threatening to their collective cultural identity.¹⁸ Segmented-assimilation scholars argue that these quantitative and qualitative findings indicate a pattern of downward mobility and resistance to assimilation among some immigrant groups.

Although research on immigrant outcomes is well developed, there is less empirical analysis of what mechanisms may lead to outcomes consistent with either segmented assimilation or second-generation advantage (partly because of the paucity of nationally representative data that allow for sufficient disaggregation by immigration status and ethnicity). This article seeks to help fill that gap by focusing on the time use of a nationally representative sample of immigrant youths compared with their native-born counterparts.

The analysis that follows focuses on the time teenagers spend on activities in five major categories: paid work, unpaid work in the home (housework and caregiving), personal care (sleep, eating, and grooming activities), education activities (including studying), and free-time activities. Special attention is paid to the three “productivity-related” time use activities: education activities, unpaid work in the home, and paid work outside the home.

It is widely believed that attending school is a crucial first step toward successful adaptation to U.S. society for children with immigrant backgrounds, because schooling leads to the attainment of knowledge, skills, and credentials that can later be capitalized on in the labor market.¹⁹ Education, which is publicly available to all children, is traditionally seen as the key means of socioeconomic mobility; thus, schooling often occupies a key role in immigrant aspirations.²⁰

Although youths gain work experience through employment and, it could be argued, learn responsibility for others through work in the home, the time that teenagers spend in paid work and in unpaid work in the home may reduce the amount of time teenagers have to devote to schooling. Some children may be forced into adult roles, including taking on substantial work and caregiving responsibilities, too early and at the expense of academic achievement.²¹ In a 2007 article, Linda Burton argues that children in poor families take on some of the responsibility for managing their parents’ financial and health problems at relatively young ages whereas their more affluent peers are protected from such adult concerns.²² Welfare-to-work demonstration programs have found harmful effects of maternal employment on adolescents’ educational attainment, with one conjecture being that older children (especially daughters) in these families must assume the burden of caregiving for younger siblings. This caregiving responsibility interferes with schooling by increasing tardiness or absences.²³ Immigrant youths may have the added responsibility of serving as the family interpreter, particularly in families in which the parental generation has low English-language ability.

A number of researchers also have found that employment during high school diminishes teenagers’ school outcomes.²⁴ For example, in a 1995 article, Linda Worley found that grades decline as hours worked during the school year increase²⁵ (although another study²⁶ found little difference once background measures were controlled). Similarly, a 1991 study by Herbert Marsh found an inverse correlation between total hours worked during high school and 17 of 22 senior-year and postsecondary measures, including academic achievement, grade point average, academic track, amount of time devoted to homework, social and academic self-concept, and educational aspirations.²⁷ The detrimental consequences of hours of paid employment during high school also have been shown to last beyond high school. For example, two studies conducted 15 years apart found that hours worked during high school are inversely correlated with the probability of attending college and of completing college.²⁸ Some scholars argue that work intensity during high school is attributable to preemployment differences, such as being initially less engaged in school and having more autonomy from parents.²⁹ However, a recent study by Charlene Marie Kalenkoski and Sabrina Pabilonia, using ATUS data, found that employment decreases the time high school students spend on homework and extracurricular activities, two activities that build human capital.³⁰ The authors found that students who worked on the day about which they reported in their diary (their diary day) spent 49 minutes less on homework than students who did not work on their diary day. Combined with Julian Betts's finding that an additional half hour of homework per night in grades 7 to 11 increases math scores by two full grade levels,³¹ Kalenkoski and Pabilonia's results suggest that too much employment may in fact reduce academic outcomes.

Hypotheses

To the extent that teens with immigrant backgrounds come from families with lower socioeconomic status than do families of native-born teens, we expect to find higher levels of paid work and household work among immigrant teens. Similarly, the lesser resources of youths from immigrant households lead us to expect that they will spend less time in school and extracurricular activities. However, the high premium placed on intergenerational mobility may mitigate this expectation, and we might actually find higher investment in school-related activities—such as time spent studying—among immigrant than native-born youths. This phenomenon could occur even in the face of higher paid and unpaid workloads if youths in immigrant households trade off leisure with work and schooling activities to a greater extent than youths in non-immigrant households do.

The analysis also seeks evidence of cultural differences between immigrant and nonimmigrant youths—for example, whether youths with foreign-born parents spend more time with family than teens with native-born parents. The family is one of the most important institutions for the socialization of youths. A number of quantitative and qualitative studies have attributed the academic success of some immigrants to integration within the family, which places high values on education and a work ethic.³² In addition, Marcelo and Carola Suárez-Orozco found that European-American adolescents are more ambivalent toward authority and schooling, and are more peer oriented, than Latino-American adolescents, who are more respectful of authority and more family oriented; these differences lead to more intergenerational conflicts among European-American adolescents.³³ In turn, higher levels of intergenerational conflict lead to less parental authority and insufficient family communication and thus have negative effects on youths’ self-esteem, psychosocial well-being, and academic aspirations.³⁴ Thus, in the category of family time, adaptation to American cultural norms may actually be undesirable for immigrant youths. In the analysis that follows, it is hypothesized that teens from immigrant backgrounds spend more time with family; hence, the analysis goes on to examine whether time spent with family is a characteristic of immigrant families or is mostly the result of their family composition and lower socioeconomic status. Also examined is whether youths with immigrant parents engage in different leisure pursuits than adolescents with native-born parents do, although no strong hypotheses about potential differences are presented.

Finally, time allocation by gender is examined within the two groups of adolescents; it is expected that teens of immigrant parents will show greater differentiation in activities by gender than teens of native-born parents, because immigrant parents may have more traditional gender expectations.

Data and analysis

This study takes advantage of the first-ever U.S. time use data with sufficiently large sample sizes to examine differences in teenagers’ time allocation by whether or not they have parents who immigrated to the United States. The data source is the ATUS, a nationally representative cross-sectional time use survey launched in 2003 by the Bureau of Labor Statistics and conducted by the Census Bureau. The ATUS interviews randomly selected individuals age 15 years and older from a subset of households that have completed their eighth and final interview for the Current Population Survey (CPS). Interviews for the ATUS typically take place between 2 and 5 months after the household's final CPS interview.³⁵

The ATUS collects information on all the activities that took place during a 24-hour period in an individual's life—including the time each activity started, the time it ended, the nature of the activity, and where and with whom it took place³⁶—thereby providing a comprehensive and contextualized picture of time allocation. During a computer-assisted telephone interview, researchers ask respondents what they did between 4 a.m. of the previous day and 4 a.m. of the interview day. The diary method has been shown to yield more reliable estimates of time use than do stylized questions about time spent performing a certain activity during a reference period.³⁷ ATUS interviewing occurs continuously over the course of the year, with each respondent interviewed once. Data files are released annually. In the analysis to be presented, data are pooled from the 2003–2010 ATUS to increase sample sizes. The response rate ranged from 53 percent to 58 percent during those years. Fifty percent of the sample is interviewed about the respondents’ time use on a weekday (Monday through Friday) and 50 percent about time use on a weekend. Weights are applied that adjust for nonresponse and for oversampling of some groups and that equally weight all 7 days of the week. Evidence suggests that using the weights helps to correct for sources of nonresponse bias.³⁸

The sample³⁹ includes 5,198 respondents who were 15–17 years of age at the time of the ATUS interview and who were living with at least one parent: 4,203 youths with a native-born background and 995 youths with an immigrant background (i.e., the respondent and/or at least one co-resident parent were born outside the United States). One hundred sixty-eight cases in which the 15- to 17-year-old adolescent did not live with at least one parent were deleted. All teenagers ages 15-17 who lived with at least one parent—even those teenagers who were not enrolled in school—were included, so as not to artificially reduce gaps in educational behavior between youths with immigrant parentage and those with native-born parentage. Previous studies have found that a nontrivial share of adolescents who were born in Mexico never enroll in U.S. schools because they migrated to the United States for work.⁴⁰

Variables: immigrant and native background

This study specifically compares teenagers with immigrant backgrounds with teenagers who have no immigrant background. Yet, immigrant status and native status are in no way self-evident and could be construed in a number of ways.⁴¹ Herein, teenagers who are themselves foreign born (commonly known as first-generation immigrants) or who have at least one parent who is foreign born (second-generation immigrants) are categorized as youths with immigrant backgrounds; teens who are not living with a foreign-born parent are classified as native born. The place of birth of the parents in the household is used to make this determination. In families in which the respondent lives with only one parent, only the place of birth of that parent is known. Hence, some of the respondents classified as having U.S.-born parents may also have an immigrant parent who is not living with them. However, there is no way to identify this situation with the available data. Children are classified as children of native-born parents in these households because we have information on only one parent's immigrant status. The alternative would have been to restrict the analysis to children in two-parent families, thereby severely restricting the sample size.

Variables: race or ethnicity and place of birth

In addition to the broad distinction between teenagers with native-born parentage and those with immigrant parentage, the sample of teenagers is further disaggregated by race or ethnicity (for those with a native-born household background) and by parental place of birth (for those with an immigrant household background). A considerable amount of research has uncovered differences in the socioeconomic status among native-born non-Hispanic Whites and minority youths. In this article, the native-born group is disaggregated into four categories: non-Hispanic White, Black, Hispanic, and “other” (which includes third-generation and later Asians and Native Americans). Small sample sizes do not permit disaggregation of the “other” group.

Immigration to the United States has been substantial in recent decades, with particularly large flows from Latin America and Asia. On average, immigrants from Asia are much more highly educated than immigrants from Latin America.⁴² Given differences in the average socioeconomic status of youths, depending on where their parents were born, we might expect differences in the productive activities of paid work and study. On average, the families from Latin America might have a greater need for youths to be employed. Although all immigrant groups tend to experience upward mobility across generations, given the higher educational attainment, on average, of immigrant Asian families, youths with immigrant Asian parents are expected to be especially likely to commit more time to educational activities, such as studying.

The group of youths with an immigrant parent was dis-aggregated into three categories: those with a parent (or parents) born in Latin America (including Mexico, with Mexicans being the largest immigrant group in the United States), those with a parent born in an Asian country, and “other.”⁴³ The “other” category is a heterogeneous mixture of those with foreign-born parents, mostly from Europe, Canada, Australia, New Zealand, Africa, or one of the Pacific Islands. Small sample sizes do not permit further elaboration of immigrant backgrounds. In sum, the full race–immigrant status variable comprises seven categories: native-born non-Hispanic White, native-born Hispanic, native-born Black, native born of some other race, Latin American immigrant, Asian immigrant, and other immigrant family background.

Variables: age, gender, and region of residence of the adolescent

Because we focus on adolescents living with at least one parent, we restrict the age range of adolescents to those ages 15–17 because it is far more common for youths to reside apart from parents after they reach age 18, the age of majority in the United States. Parental characteristics are not available in the ATUS for those who do not co-reside with parents. Multivariate models include a control for single years of age. Estimates of time use are disaggregated by gender. Regression models include a control for gender, labeled female and coded as 1, with male as the omitted category.

The regional distribution of immigrant and nonimmigrant families is quite distinct. Region of residence is captured with a four-category variable that identifies those in the Northeast (the reference group), Midwest, South, and West.

Variables: teen's family composition

The analysis distinguishes youths who are living in a single-parent family and youths living in a two-parent family. Two-parent families are further disaggregated by whether the mother is not employed, employed part time, or employed full time. In the regression models, two-parent families with a mother who is employed full time serves as the reference category for the family-type variable.

The number of siblings under 18 years who live in the adolescent's household is captured with an indicator variable of 0, 1, 2, and 3 or more. In the regressions, 0 (the adolescent is the only child present in the household) is the reference category.

Variables: socioeconomic status of the adolescent's family

Three indicators of the socioeconomic status of the adolescent's household are also examined: an indicator of parental educational attainment, an indicator of family income, and an indicator of whether the family owns a business. Parental educational attainment is coded into three categories: less than high school diploma (the omitted category in the regressions), high school diploma or some college, and college degree or higher. In two-parent families, the parental educational attainment variable is based on the parent with the higher level of educational attainment. An indicator variable of family income is disaggregated into five categories: less than $25,000 (the reference category in the regressions), $25,000–$49,999, $50,000–$74,999, $75,000–$99,999, and $100,000 or more. Income data are missing for 12 percent of the respondents; cases of missing respondents are assigned to the modal category ($25,000–$49,000). A dummy flag for missing income is included in the regressions and is not significant in any of the models.

Measuring time: the total day

This study analyzes the time teenagers spend on activities in five major categories: contracted time or paid work, committed time or unpaid work in the home (housework and caregiving), personal care (sleeping, eating, grooming, and other personal care activities), free-time activities, and educational activities. Unlike most studies of adults, this study considers education separately because adolescents spend a substantial amount of time in school and related activities and because education is central to upward mobility.⁴⁴

Measuring time: teens’ productive activities

The analysis considers three activities to be the “productive” activities of teens: paid work, household work, and educational activities. Paid work is measured as minutes spent on the diary day in paid work and in commuting to or from paid work. Total household work comprises minutes per day spent on housework and on caregiving. Educational activities captures three activities: minutes per day spent in school, studying, and in extracurricular activities. The time spent in extracurricular activities is classified into productive activities rather than leisure because of the importance of these activities for teenagers vying for college admission.

Measuring time: teenagers' time spent in leisure activities and personal care

The category of personal care includes the number of minutes per day spent sleeping, eating meals, and grooming. Total minutes spent in free-time activities are assessed and are disaggregated into a number of categories of interest, including watching television, computer use and games (but not computer games), reading, sports activities, religious activities, and volunteering.

Measuring time: time spent with family

Qualitative studies consistently highlight the central role that family plays in the lives of immigrants and their children.⁴⁵ To test whether youths with immigrant parents are embedded in a denser social network of family and kinship ties than are youths with native-born parents, the “with whom” data in the time diary are used to construct indicators of the number of minutes per day that the adolescent spends with a parent; the number of minutes the adolescent spends with relatives other than parents, excluding siblings; and the number of minutes the adolescent spends with relatives other than parents, including siblings. That is, in addition to asking respondents what they were doing, the ATUS asks them who was with them while they were doing an activity. The number of minutes spent in that manner are summed across the entire diary day to construct the number of minutes per day spent with a parent, with relatives other than siblings, and with any relatives (including siblings).

Table 1 shows the distribution of the sample by family characteristics.⁴⁶ Approximately 1 in 5 adolescents lives with a parent who was born outside the United States. Compared with those who live with native-born parents, adolescents with an immigrant parent are characterized by a slightly higher percentage of males, have parents with much lower levels of educational attainment and lower family incomes, are more often in two-parent families, and have a greater number of siblings. The percentage of adolescents in immigrant households who reside in the West is far higher than the percentage of adolescents who live with native-born parents and reside in the West.

Table 1.

Percent distribution of personal and family characteristics, 15- to 17-year-olds, 2003–2010

Characteristic	Total	Native-born household	Immigrant household
Household background
Native-born parent(s)	77.6	...	...
Non-Hispanic White	56.9	73.3	...
Hispanic	6.3	8.2	...
Black	11.6	14.9	...
Other	2.8	3.6	...
Immigrant parent(s)	22.4	...	...
Latin America	15.0	...	66.9
Asia	3.9	...	17.3
Other	3.5	...	15.8
Gender of teen
Male	51.1	49.9	55.2
Female	48.9	50.1	44.8
Age of teen
15	27.2	27.2	27.1
16	36.7	36.9	35.9
17	36.2	35.9	37.0
Region of residence
Northeast	17.6	17.5	18.0
Midwest	23.6	27.9	8.5
South	33.9	35.0	30.1
West	24.9	19.5	43.4
Parental education
Less than high school diploma	12.8	6.6	34.3
High school diploma or some college	49.8	53.8	36.1
College degree or higher	37.4	39.6	29.5
Type of family
Two-parent family, mother employed full time	41.3	41.2	41.5
Two-parent family, mother employed part time	15.1	15.4	14.3
Two-parent family, mother not employed	17.2	14.7	25.9
Single mother	22.2	23.9	16.3
Single father	4.2	4.8	2.1
Family business
Family doesn't own a business	81.9	81.1	84.8
Family owns a business	18.1	18.9	15.2
Family income
Less than $25,000	14.9	12.7	22.4
$25,000–$49,999	32.7	30.1	41.9
$50,000–$74,999	19.6	20.8	15.2
$75,000–$100,000	15.0	16.3	10.3
$100,000 or more	17.8	20.0	10.1
Missing income data
Not missing income	88.1	88.4	87.0
Missing income	11.9	11.6	13.0
Number of siblings in household
Teen is only child in household	37.7	39.9	29.8
1 sibling	34.1	35.0	31.2
2 siblings	17.7	15.8	24.1
3 or more siblings	10.5	9.3	14.9
Sample size	5,198	4,203	995

SOURCE: U.S. Bureau of Labor Statistics, American Time Use Survey.

Measuring time: analysis plan

The next section describes the total number of minutes per day, on average, that teenagers spend in each of their activities. These averages are broken down by immigrant background and by gender and immigrant background. Then, weighted ordinary least squares regressions are used to standardize for differences in family characteristics in order to assess whether, net of differences in family background, children of immigrants spend differential amounts of time in productive activities.

Results

Table 2 shows the number of minutes per day teenagers spend in each of their various activities. There are notable differences in the productive activities of youths who live with and without immigrant parents. Even though household income levels are lower in immigrant households, a factor that might increase the need for additional workers, adolescents with native-born parents spend more time in paid work than do adolescents with immigrant parents. The former average 52 minutes per day, compared with an average of 27 minutes for the latter (p <.001, two-tailed t-test). Conversely, immigrants and children of immigrant parents spend more time in education-related activities: an average of 26 minutes more per day, compared with the time spent by youths living with native-born parents (p <.01, two-tailed t-test). As a group, immigrant teens spend about 8 minutes more on unpaid household activities than do teens with native-born parents (p <.05, two-tailed t-test).

Table 2.

Average number of minutes per day teenagers with native-born and immigrant parents spend in various activities, 15- to 17-year-olds, by gender, 2003–2010

Activity	All teens		Male		Female
	Native-born parent(s)	Immigrant parent(s)	Native-born parent(s)	Immigrant parent(s)	Native-born parent(s)	Immigrant parent(s)
Total paid work	51.9	26.7	54.0	28.6	49.7	24.3
Total household work	49.9	57.6	41.0	44.3	58.8	74.0
Housework	38.4	40.8	32.6	31.6	44.2	52.1
Total caregiving	11.5	16.8	8.4	12.7	14.6	22.0
Caregiving, own household	5.3	8.7	3.2	6.2	7.4	11.8
Caregiving, outside of household	6.2	8.1	5.2	6.5	7.2	10.1
Total education	210.7	236.8	200.9	235.6	220.4	238.3
School	166.5	181.3	166.7	187.3	166.3	173.8
Study	37.9	50.0	29.2	44.2	46.5	57.1
Extracurricular activities	6.3	5.6	5.0	4.1	7.6	7.3
Total personal care	665.2	675.6	657.9	676.5	672.4	674.4
Eating	51.4	56.5	52.0	56.0	50.8	57.1
Sleeping	564.6	570.7	566.2	579.9	563.0	559.3
Grooming	47.3	48.0	37.3	40.1	57.3	57.8
Other personal care activities	1.8	.3	2.4	.4	1.3	.2
Total free time	380.8	359.8	407.0	377.1	354.7	338.5
Computer use	19.8	20.1	20.1	20.8	19.5	19.3
Visiting	52.8	51.0	47.3	51.5	58.3	50.3
Television	130.5	137.8	135.1	142.0	125.9	132.6
Games	32.8	23.3	52.5	35.0	13.2	8.9
Reading	8.6	6.2	6.2	4.4	11.0	8.3
Total sports	53.9	41.6	71.2	56.2	36.7	23.5
Sports or exercise	47.8	37.7	65.2	52.3	30.4	19.6
Attending sports	6.1	3.9	6.0	3.8	6.2	3.9
Religious and spiritual activities	8.0	5.9	7.4	5.9	8.6	5.8
Volunteer activities	10.5	8.7	11.6	7.5	9.4	10.3
Shopping	15.6	22.4	10.7	15.9	20.4	30.3
Telephone	15.7	13.6	12.7	8.4	18.7	20.1
Other leisure	29.7	26.5	30.1	28.0	29.3	24.6
Obtaining services	2.9	2.8	2.0	1.4	3.7	4.5
Time spent traveling to activities	70.5	70.5	69.7	67.3	71.3	74.5
Unaccounted minutes1	11.0	13.0	10.0	10.6	12.6	16.0
Total minutes	1,440.0	1,440.0	1,440.0	1,440.0	1,440.0	1,440.0
Time spent with at least one parent2	143.6	145.3	134.0	121.2	153.2	175.0
Time spent with any relative (not including siblings)2	172.6	150.5	170.6	142.0	174.5	161.1
Time spent with any relative (including siblings)2	203.5	249.5	191.6	218.0	215.4	288.2
N	4,203	995	2,107	544	2,096	451

¹These minutes were not coded in the survey because of insufficient detail, incorrect words, missing travel information, simultaneous activities incorrectly recorded, refusal of the respondent to provide information, or a gap in memory.

²Respondents are not asked the “where” and “with whom” questions for sleeping, grooming, and personal activities, or for any times for which they could not remember what they were doing. From 2003 to 2009, respondents were not asked the “with whom” questions for work activities.

SOURCE: U.S. Bureau of Labor Statistics, American Time Use Survey.

Adolescents in immigrant households spend more time with relatives (including siblings) than do those in native-born households. Whereas time spent with parents does not differ between adolescents with immigrant parents and those with native-born parents, adolescents with immigrant parents average 250 minutes, or 4.2 hours, per day in the company of some family member, compared with 204 minutes, or 3.3 hours, per day for youths who live with native-born parents (p <.001, two-tailed t-test). However, when time spent with siblings is not included, it is actually native-born teenagers who spend more time with their relatives, averaging 173 minutes per day compared with 151 minutes per day for youths who live with immigrant parents (p <.05, two-tailed t-test).

Table 2 also disaggregates the time use of teens by gender, and table 3 provides further detail by showing the percentage of male and female respondents who report each activity on their diary day and the mean number of minutes for those who engage in the activity. The difference in time spent in paid work between those in immigrant households and those in native-born households is present for girls and boys alike, with both sons and daughters in native-born households averaging nearly an hour a day and with averages closer to a half hour a day in immigrant households (see table 2; p <.001, two-tailed t-test). About 20 percent of adolescents living with native-born parents report paid work on the diary day, compared with approximately 11-12 percent of youths in immigrant households (see table 3; p <.001, two-tailed t-test). Thus, the major source of the difference among teens with native-born backgrounds and teens with immigrant backgrounds is the fact that a smaller proportion of teenagers with immigrant backgrounds do any paid work on their diary day; on days when children with immigrant backgrounds report paid work, they average as much or more time at work as do their counterparts with native-born backgrounds.

Table 3.

Percentage of youths who do selected activities, and average number of minutes per day engaged in each activity, conditional on doing the activity, 15- to 17- year-olds, by gender, 2003–2010

Activity	Male				Female
	Native-born parents		Immigrant parents		Native-born parents		Immigrant parents
	Percentage who do activity	Mean1	Percentage who do activity	Mean1	Percentage who do activity	Mean1	Percentage who do activity	Mean1
Total paid work	20.3	264.0	11.8	269.5	19.6	257.0	11.1	250.5
Total household work	56.9	74.0	55.0	80.7	70.1	82.8	71.3	103.1
Housework	47.9	69.2	44.9	71.4	59.4	73.8	61.9	85.9
Total caregiving	22.0	38.5	21.5	55.1	30.8	46.6	31.9	64.7
Caregiving (own household)	7.9	36.8	9.4	52.9	12.5	53.7	17.5	63.9
Caregiving (outside of household)	15.9	34.5	13.6	49.1	21.0	35.0	16.4	60.0
Total education	44.0	384.0	49.3	402.6	49.9	383.4	54.1	392.0
School	32.1	375.5	33.5	391.1	31.5	371.3	33.7	375.1
Study	26.7	98.2	36.9	108.1	36.1	119.9	43.5	57.1
Extracurricular activities	3.9	127.8	2.4	216.6	5.7	110.8	4.2	180.5

¹Mean for those who participate in the activity.

SOURCE: U.S. Bureau of Labor Statistics, American Time Use Survey.

The other striking difference—both an immigrant–native-born difference and a gender difference—is in time allocated to education-related activities: overall, immigrant teens report spending 237 minutes per day (an average of 27.7 hours per week) on educational activities, while native-born teenagers report 211 minutes per day (an average of 24.6 hours per week; p <.05, two-tailed t-test). Daughters in immigrant households report 238 minutes on educational activities, followed by sons in these households, who average 236 minutes (not significantly different, p = .30). Daughters in native-born households average 220 minutes, about 20 minutes more than the average for sons in these households (see table 2; p < .05, two-tailed t-test). In other words, daughters in immigrant households end up committing about three-quarters of an hour per day more, or 4.4 hours more per week, to educational activities than do sons in native-born households.

One might argue that time spent in school is less discretionary and less under the influence of either the youth or his or her parents than is time studying (although immigrants tend to report more minutes of school time than do native-born youths (see table 2)). Table 3 shows that 44 percent of daughters in immigrant households report studying on their diary day, compared with 36 percent of daughters in native-born households (p < .01, two-tailed t-test) and 37 percent of sons in immigrant households (p < .05, two-tailed t-test). Only 27 percent of sons in native-born households report doing any studying on their diary day.

Daughters do more household work than sons, and daughters of immigrants spend more time in household work than do daughters of native-born parents. There is only a slight immigrant–nonimmigrant difference for sons, with both groups of sons averaging 41–44 minutes per day in unpaid household work, compared with 59 minutes per day for daughters of native-born parents and 74 minutes per day for daughters in immigrant households. (See table 2; the difference in means for females is significant at the p < .001 level, two-tailed t-test.) Relatively high percentages of youths report doing some household work on their diary day: 45 percent to 48 percent of sons and 59 percent to 62 percent of daughters. (See table 3.) This likelihood of doing unpaid household work is not what differs between daughters in immigrant and nonimmigrant households; rather, a difference exists regarding the amount of time spent doing housework and caregiving, with an average of 103 minutes per day for daughters who do some of this type of work in immigrant households, compared with 83 minutes per day for daughters in native-born households (p < .05, two-tailed t-test).

There are hints of other differences in time allocation in these tables as well, with time spent on games being much higher for boys than girls and higher by about 18 minutes per day for sons in native-born households compared with sons in immigrant households (see table 2; p < .05, two-tailed t-test). Boys spend more time in sports-related activities than girls do, and there is a suggestion that both sons and daughters in native-born households spend more time in these activities than do their counterparts in immigrant households. Overall, there is a gradient in free time, with all groups having a sizeable amount of free time on their diary day, but with immigrant girls having the least amount: an average of 339 minutes, followed by 355 minutes for native-born girls and 377 minutes for immigrant boys. Native-born boys have the most free time, 407 minutes, a half hour more each day than immigrant girls. (A Wald test indicates that the differences in means between girls and boys and between immigrants and nonimmigrants are significant, F = 49.3, p < .001).

Finally, time spent with parents and family members shows interesting differences once gender is disaggregated. Daughters in immigrant households appear to spend 22 minutes more per day with parents (p < .01, two-tailed t-test), and 73 minutes more per day with any relative (p < .001, two-tailed t-test), than do daughters in native-born households. Still, both groups spend a sizable amount of time in the company of parents and relatives in general. Sons spend less time than daughters with parents, with sons in immigrant households reporting the lowest number of minutes so spent: 121, compared with 134 for sons in native-born households (this difference between sons is not significant, p = .24, two-tailed t-test), 153 for daughters in native-born households, and 175 for daughters in immigrant households. (A Wald test indicates that the differences in means between daughters and sons and between immigrants and nonimmigrants are significant, F = 11.3, p < .001). As regards overall time spent with any relatives, daughters in immigrant households stand apart, spending much more time with family members than the other groups do when time spent with siblings is included. However, with time spent with siblings omitted from the analysis, immigrant girls spend less time with relatives than do native-born girls or boys.

To assess the extent to which differences between the time allocations of children of immigrant parents and children of native-born parents reflect differences in household structure and socioeconomic status, multivariate analysis is applied. Here, differences in observed characteristics of immigrant and nonimmigrant households are controlled in order to see if these factors explain or reduce the differences between groups in amounts of time allocated to productive activities.

Multivariate analysis

Table 4 shows ordinary least squares regression results for immigrant status in models predicting teenagers’ time in the productive activities of (paid) work, total education, and the subcategory of study time. Table 5 gives regression results for time spent in total household work (including both housework and caregiving) and, separately, for housework and caregiving activities. Table 6 presents regression results for time spent with relatives not including siblings and time spent with relatives including siblings. In all three tables, Panel 1 shows results for models that include a bivariate indicator of immigrant status and Panel 2 shows results for models that include the seven-category variable for race or ethnicity and immigrant background. For each activity listed at the head of each pair of columns, results are shown from two models: a bivariate model (“no controls”) that includes only the categorical variable for race or ethnicity and immigrant background, and a multivariate model (“with controls”) that adjusts for family size, family structure, and family resources and that shows the relationship of living in an immigrant household (by race or ethnicity and place of origin), net of these factors. Control variables include type of family, parental level of education, family income, whether the family owns a business, number of siblings, gender, and region and age of the teenager.

Table 4.

Results from ordinary least squares regressions predicting number of minutes per day teenagers spend in productive activities, 15- to 17-year-olds, 2003–2010

Panel	Work		Total education		Study time
	Without controls1	With controls1	Without controls1	With controls1	Without controls1	With controls1
Panel 1: Immigrant status
Native born2	...	...	...	...	...	...
Immigrant	3–25.2 (4.3)	3–20.1 (4.4)	426.1 (10.2)	430.3 (11.2)	312.1 (3.3)	318.2 (3.5)
Panel 2: Race and immigrant status
Native-born household:
Non-Hispanic white2	...	...	...	...	...	...
Hispanic	–12.6 (8.1)	–8.2 (8.2)	–19.7 (17.7)	–1.9 (17.8)	4–14.5 (4.3)	–5.4 (4.5)
Black	5.8 (9.7)	3.9 (10.0)	5–34.9 (13.8)	–7.9 (14.5)	3–20.7 (3.3)	–6.6 (3.4)
Other	5–17.3 (8.6)	–18.1 (8.8)	–9.4 (21.0)	5.7 (20.4)	–10.8 (5.9)	–5.0 (5.8)
Immigrant household:
Latin American	3–28.9 (5.0)	3–23.5 (5.5)	18.1 (12.5)	535.4 (14.8)	–2.2 (3.8)	510.7 (4.3)
Asian	4–25.0 (8.3)	4–22.8 (8.6)	551.1 (21.3)	550.9 (21.4)	336.8 (8.7)	334.2 (8.4)
Other	–14.5 (9.4)	–14.0 (9.4)	–12.5 (21.9)	–12.2 (21.6)	516.0 (7.4)	514.9 (7.4)
N	5,198	5,198	5,198	5,198	5,198	5,198

¹Results are for coefficient B; standard errors are in parentheses.

²Reference category.

³p < .001.

⁴p <.01.

⁵p < .05.

NOTE: Controls include gender of teenager, age of teenager, region of residence of respondent, parental education, type of family, family income, family business ownership, and number of siblings.

SOURCE: U.S. Bureau of Labor Statistics, American Time Use Survey.

Table 5.

Results from ordinary least squares regressions predicting number of minutes per day teenagers spend in household work, 15- to 17-year-olds, 2003–2010

Panel	Total household work		Housework		Caregiving
	Without controls1	With controls1	Without controls1	With controls1	Without controls1	With controls1
Panel 1: Immigrant status
Native born2	...	...	...	...	...	...
Immigrant	37.7 (3.7)	3.8 (4.2)	2.3 (2.9)	.3 (3.4)	35.3 (2.2)	3.5 (3.6)
Panel 2: Race and immigrant status
Native-born household:
Non-Hispanic White2	...	...	...	...	...	...
Hispanic	–8.8 (5.8)	4–14.8 (5.5)	3–11.1 (4.5)	4–13.4 (4.5)	2.4 (4.0)	–1.4 (3.6)
Black	4–12.2 (4.2)	4–14.8 (4.5)	4–9.5 (3.6)	3–9.0 (3.9)	–2.7 (1.7)	3–5.8 (1.8)
Other	5.1 (8.4)	.3 (8.1)	1.2 (7.0)	–1.2 (7.0)	3.9 (4.2)	1.4 (3.9)
Immigrant household:
Latin American	5.7 (4.5)	–4.0 (5.2)	.8 (3.6)	–3.8 (4.4)	4.8 (2.7)	–.2 (2.9)
Asian	–1.7 (7.0)	–2.3 (7.4)	–6.0 (5.6)	–7.0 (5.8)	4.3 (4.4)	4.6 (4.5)
Other	11.5 (9.7)	13.5 (9.6)	3.4 (7.3)	4.7 (7.4)	8.1 (6.5)	8.8 (6.2)
N	5,198	5,198	5,198	5,198	5,198	5,198

¹Results are for coefficient B; standard errors are in parentheses.

²Reference category.

³p < .05.

⁴p < .01.

NOTE: Controls include gender of teenager, age of teenager, region of residence of respondent, parental education, type of family, family income, family business ownership, and region and number of siblings. Caregiving (p < 05) is restricted to teens with siblings.

SOURCE: U.S. Bureau of Labor Statistics, American Time Use Survey.

Table 6.

Results from ordinary least squares regressions predicting number of minutes per day teenagers spend with relatives, 15- to 17-year-olds, 2003–2010¹

Panel	Time spent with relatives, not including siblings		Time spent with relatives, including siblings
	Without controls2	With controls2	Without controls2	With controls2
Panel 1: Immigrant status
Native born3	...	...	...	...
Immigrant	6–22.0 (8.0)	–7.7 (9.0)	445.9 (9.1)	524.8 (9.9)
Panel 2: Race and immigrant status
Native-born household:
Non-Hispanic White3	...	...	...	...
Hispanic	–20.1 (15.0)	–5.1 (15.4)	21.8 (15.7)	6.9 (15.7)
Black	–15.3 (12.2)	–11.9 (13.0)	–18.3 (11.8)	–9.4 (12.0)
Other	–7.2 (19.7)	–.4 (19.0)	2.3 (17.6)	.6 (17.6)
Immigrant household:
Latin American	6–28.9 (9.9)	–9.7 (12.0)	446.9 (10.5)	18.6 (11.9)
Asian	5–30.6 (14.1)	–22.0 (14.0)	34.6 (19.4)	25.8 (19.4)
Other	–10.1 (19.4)	1.4 (18.8)	48.6 (25.3)	40.9 (24.7)
N	5,198	5,198	5,198	5,198

¹Respondents are not asked the “where” and “with whom” questions for sleeping, grooming, and personal activities, or for any times for which they could not remember what they were doing. From 2003 to 2009, respondents were not asked the “with whom” questions for work activities.

²Results are for coefficient B; standard errors are in parentheses.

³Reference category.

⁴p < .001.

⁵p < .05.

⁶p < .01.

NOTE: Controls include gender of teenager, age of teenager, region of residence of respondent, parental education, type of family, family income, family business ownership, and region and number of siblings. Caregiving (p < .05) is restricted to teens with siblings.

SOURCE:: U.S. Bureau of Labor Statistics, American Time Use Survey.

Youths in immigrant households spend significantly less time in paid employment and more time in total educational activities, including studying; these differences remain significant for both Latin American and Asian immigrants once family compositional differences are controlled. (See table 4.) Youths in households with immigrant parents from Latin America and Asia average significantly less time in paid work than do youths in households with only native-born parents, and this difference remains sizable—about 24 minutes less per day for the Latin American youths and 23 minutes less for the Asian youths—after adjusting for differences in family socioeconomic status and composition.

Time spent in educational activities is significantly higher in immigrant households than in nonimmigrant households, even after adjusting for family compositional differences. Adjusting for such differences increases the size of the coefficient from (a nonsignificant) 18 minutes per day to a statistically significant 35 minutes per day for youths with Latin American immigrant backgrounds (compared with non-Hispanic native-born White youths). This increase most likely reflects the fact that Latino immigrant parents have characteristics, such as lower income and less education, that are typically associated with lower educational achievement of children. Adjusting for these “suppressor” variables actually somewhat widens the gap in time spent studying between youths in Latino immigrant households and those in native-born, non-Hispanic White households. For teens with an Asian immigrant background, controlling for compositional differences does not alter the coefficients for minutes per day spent in educational activities: in both the bivariate and multivariate models, the coefficient is 51 minutes per day. Time spent studying follows a similar pattern for both Latin American and Asian immigrant teens. Without controls, the study time of Latino immigrant youths is slightly less than that for native-born White youths. After controls, Latino immigrant youths average 11 minutes more per day, or over an hour more, studying per week than do native-born White youths. Asian immigrant youths average a half hour more per day than native-born White youths do, before and after controlling for family background.

Results for total household work, housework, and care-giving in table 5 show that time allocations to total household work are slightly higher (by about 7 minutes per day) for youths with immigrant parents, but this difference is no longer statistically significant once family compositional factors are controlled. Nonetheless, some of the results discussed earlier suggest that differences in these productive behaviors might be especially pronounced for girls. Hence, total household work, housework only, and caregiving were examined for the restricted sample of female adolescents (data not shown). Bivariate models suggested that immigrant girls—in particular, Latina immigrant girls—might do about 15 minutes more total household work per day than native-born girls. However, this difference in time spent in total household work becomes nonsignificant once family compositional factors are controlled. Moreover, no other differences emerge from the analysis.

Table 6 shows results for the time teenagers spend with relatives. When time spent with siblings is included, immigrant teens appear to spend more time with relatives, even after controlling for family composition and socioeconomic status (although there is no statistically significant difference in models using the disaggregated race-or-ethnicity and immigrant status variables and including controls for family composition and socioeconomic status). However, if time spent with siblings is not included, immigrant teens actually spend less time with relatives than do native-born teens, although this difference, too, does not remain statistically significant once family compositional and socioeconomic factors are controlled. All these results suggest that immigrant teens spend more time caring for siblings than native-born teens do, but this difference is not picked up by the “caregiving” measure because caregiving may be done informally, at the same time that the immigrant teens are doing other activities. Given the attention paid by the assimilation literature to the role of family in the lives of immigrant youths, it is somewhat surprising that native-born teens spend more time with relatives other than siblings, but this seeming anomaly may be explained by the fact that the relatives of immigrant teens have to spend more time away from the household at work or that they live farther away (e.g., outside the United States).

Other covariates of adolescent time use

Tables A-1 through A-4 in the appendix present the full multivariate results. Girls do significantly more studying, housework, and caregiving than boys. Time in paid work increases with age of the teen: seventeen-year-olds average 47 more minutes per day in paid work than 15-year-olds, and 16-year-olds average 22 more minutes per day in paid work than 15-year-olds. Youths who live in two-parent households with a full-time employed mother (the reference group) do more paid work than those who live with two-parent households in which mothers are not employed. (See tables A-1 and A-2 for the full multivariate results.) Youths in households in which the family owns a business also do more paid work, presumably often working in the family business: teens whose parents own a business report 21 more minutes of paid work than do teens whose parents do not own a business. Finally, youths with one or more siblings spend more time in paid employment than do only children.

Older youths spend less time on total educational activities. Higher parental education is associated with more time spent studying: teens with a college-educated parent spend about half an hour more studying than teens whose parents have less than a high school degree. Likewise, youths in higher income households spend significantly more time studying than those in low-income households. Youths with one or more siblings spend somewhat less time studying than do only children.

Tables A-3 and A-4 show that most characteristics are not significantly related to time spent with relatives, especially when siblings are excluded. In models that include siblings, not surprisingly, time spent with relatives increases with family size. Time spent with relatives is lower in single-mother families, and is higher in two-parent families in which the mother is not employed or in which the mother is employed part time, compared with two-parent families in which the mother is employed full time.

This article takes advantage of the relatively large samples in the American Time Use Survey (ATUS) to explore whether the time allocation of youths differs by the native-born versus immigrant status of their parents. Evidence is presented that is consistent with the overall conclusions of classical and neoclassical assimilation theory: that immigrant youths are successfully assimilating into mainstream American society. If similarity in daily behavior among native-born and immigrant youths is considered an indicator of assimilation, then the range of activities in which immigrant youths participate is reasonably similar to that of youths with native-born parents. On the whole, native-born and immigrant youths spend their days in similar ways: doing household chores, eating, sleeping, grooming, using the computer, socializing, watching television, engaging in religious activities, volunteering, talking on the phone, and traveling.

There are, however, intriguing differences between youths in immigrant and native-born households. The literature would lead one to expect to find immigrants spending more time in paid work, but that is not the case: both Latino and Asian immigrant youths spend more time studying and less time in paid employment than do native-born youths. This difference suggests that immigrant teens may be pursuing a strategy of successful assimilation into U.S. society through investment in their educational capital. The findings presented here also give some support to the “immigrant advantage” literature, by pointing to a potential mechanism—investment in schooling—through which some immigrants are able to achieve mobility.

The analysis also finds evidence of segmented assimilation, a theory which argues that some immigrant groups will have divergent rather than convergent outcomes, with some groups perhaps even experiencing downward mobility. Even with socioeconomic factors controlled, Asian immigrant youths study more than Latino immigrant youths and native-born youths. Thus, studying may be one mechanism by which immigrant youths experience segmented assimilation, in addition to or in concert with their parents’ human capital, their parents’ mode of incorporation into the nation, and community resources. Still, on the whole, a comparison of the actual behaviors of teens from different ethnic backgrounds yields little evidence to support John Ogbu's and others’ argument that those ethnic groups which feel excluded or exploited are more likely to reject adapting to the dominant culture, especially with regard to schooling.⁴⁷

A limitation of the study presented here is that it lacks outcome measures and therefore cannot say for certain that more study time is associated with greater achievement. It is possible, for example, that students who achieve less actually study more: their poor academic performance spurs them to increase their time spent studying. Still, gender differences suggest that study time is a positive, rather than negative, correlate of academic achievement: girls get better grades than boys in high school and college,⁴⁸ and girls spend more time studying than boys.⁴⁹

Despite limitations, the look at time use afforded by the ATUS time diary data opens up the black box of how immigrants may achieve upward mobility. Time studying may be key, particularly for immigrant girls, for whom study time is high despite relatively high unpaid workloads (in the aggregate). Native-born boys spend the least time studying. A question for future research to consider is whether an assimilation pattern by gender is occurring, one that may harm immigrant boys’ chances for mobility more so than immigrant girls’ chances. That is, if immigrant boys assimilate to the patterns of native White male youths, they will spend less time studying and perhaps limit their chances of upward mobility.