How Well Do National Surveys Measure Hispanic Families and Households?

This brief has been updated on June 26, 2018 to include additional information in Table 7.

Overview

National surveys provide important information about the United States population. Researchers, policymakers, program developers, and government officials use data from these surveys to describe the characteristics of the population, study patterns of behavior and how they differ for subgroups of individuals, make decisions about how to allocate resources, and inform programs and policies. As a result, it is critical to maintain a data infrastructure that reflects the current U.S. population.

The United States is increasingly becoming more racially and ethnically diverse, and Hispanics are, in many respects, leading many of the shifts in the population.^1,2 The Hispanic population has not only grown rapidly over the past few decades, but has diversified in terms of nativity, country of origin, citizenship status, and geographic location within the United States. As the U.S. Hispanic population continues to diversify and comprise an increasing proportion of the general population, data are needed to understand what Latino/a  families and households look like, how their family life is organized, how Hispanic couples interact with one another, how they parent their children, and how their experiences differ (if at all) from other racial/ethnic groups and within Latino subgroups. This information is necessary to identify the strengths and needs of the Hispanic population today, and to inform the design of culturally relevant policies and programs. However, the ability to do so is contingent on the availability of current data on this population.

This brief examines the capacity of our nation’s data infrastructure to measure, describe, and understand the structure, diversity, complexity, and dynamics of Hispanic family life. We reviewed more than 20 mostly national surveys with large Latino sample sizes to assess the extent to which they include measures critical to understanding the characteristics and experiences of Hispanic families and households. We specifically examined the extent to which these surveys collect information about family and household composition, family formation and stability, relationship dynamics, and parenting and co-parenting. We selected these domains based on their relevance to Hispanic family life and their importance in predicting children’s outcomes.^b

As the Hispanic population diversifies, it has become increasingly important to measure and describe the varied experiences within Hispanic families. For this reason, we also assessed the extent to which data sets include information that can characterize the heterogeneity of Hispanic families. We searched for the availability of 10 key data elements identified by a Hispanic Research Work Group Group convened by the Administration for Children and Families (ACF) as those that are central to understanding the diverse social experiences of low-income Hispanic populations in the United States. Our review is intended to serve as a resource for researchers interested in studying Hispanic family life and those interested in identifying data sets that can inform policies and programs specifically designed for Latinos. The scan can also inform future data collection efforts.

Key Findings

The capacity of our nation’s current data infrastructure to describe the characteristics and experiences of Latino families and households is limited. On the one hand, large-scale and national surveys are collecting the needed information to count and describe the types of families and households in which Hispanic children and adults live. Additionally, several surveys provide at least some information to understand how Hispanic families are formed and how stable they are. However, our data infrastructure provides less information about what happens inside the Latino family, and is limited in its ability to describe the diversity within Latino communities.

Although many surveys capture some (albeit limited) information about parenting, few, for example, collect information on family functioning and processes that include couples’ relationship quality, co-parenting, and fathers’ involvement with children—restricting our ability to understand family processes among Latinos. More specifically, we found that:

• Data are available to adequately describe the structure of Hispanic families and households.
  • All surveys reviewed here contain information about the number of individuals, adults, and children in the household; most contain at least partial information about how individuals are related to one another— information that is critical to determining household and family composition.
  • Additionally, most surveys include questions about how families are formed, and many ask about stability over time. For all data sets, we can determine whether individuals are currently married and, often, their marital and cohabitation history, as well as family and relationship changes over time.

• Our nation’s data infrastructure has not kept up with the demographic shifts in the country.
  • The majority of surveys reviewed are longitudinal, providing a valuable opportunity to understand changes in family life and how they shape adult and child well-being over time.
  • However, the sampling frame of most longitudinal data sets does not adequately represent the current demographic composition of the country in general, and the Hispanic population specifically. Due to the nature of longitudinal studies, the sampling frame of most longitudinal data sets included in our review is at least a decade old, and about half is at least 15 years old. Therefore, these data miss much of the recent growth in, and diversification of, the Latino population.
• There is a dearth of information on relationship dynamics among Hispanic couples.
  • The majority of surveys contained no information about couples’ relationship quality.
  • Notably, relationship conflict was the most commonly examined dimension of couples’ relationship quality, yet only six surveys included questions about this aspect of couple relationships. Only two surveys measured multiple aspects of relationship dynamics extensively; however, these surveys include samples that may not be representative of all Hispanic families.
• Additionally, our knowledge of parenting behaviors—and fathering, in particular—among Latino families is limited.
  • Most surveys collect at least some information on parenting, but information is often minimal and restricted to the responding parent, which is usually the mother. Consequently, limited information is available about the father.
  • Moreover, there is limited information about co-parenting in two-parent families. Questions about how co-resident parents come together in their parenting role are rarely included in surveys.
• No survey allows us to get a complete picture of Hispanic diversity and family life.
  • Most surveys contain at least some information on five or more key data elements needed to unpack Hispanic diversity, but none has complete information on all 10 elements.^c
  • Notably, most data sets allow for comparisons between foreign-born and U.S.-born Hispanics, and most contain information on basic indicators of acculturation (specifically, time spent in the United States and language spoken at home).
  • However, limitations in the availability of information on Hispanic diversity, coupled with insufficient information about family life, hinder our ability to adequately describe Hispanic families and the diversity in their experiences.

Recommendations

Overall, based on our review, limitations in the sampling frame of existing large data sets—together with the lack of sufficient information about what occurs inside the home—signal a need for a new national survey of families and households. To adequately assess existing gaps in knowledge, this survey should:

• Assemble a new population-based cohort that captures the current demographic composition of the United States
• Obtain more granular demographic information that can help unpack the diversity within Hispanic families—namely, the 10 key data elements
• Inquire about couple dynamics, parenting, and co-parenting from both parents’ perspectives, regardless of residential status
• Collect data across multiple points in time to allow for examinations of change over time

Method

In this scan, we reviewed surveys of large-scale data sets that are commonly used to understand the well-being of children and families. We began our data set selection with the list of data sets included in the prior Center briefs in this series (see About this Brief box). Our list of data sets includes those funded by federal agencies such as the Administration for Children and Families, the Centers for Disease Control and Prevention, the National Center for Health Statistics, the National Center for Education Statistics, the Census Bureau, and Bureau of Labor Statistics, among others; as well as data sets funded through academic, government, and foundation partnerships. To ensure that we captured the full range of data sets, we also reviewed the Inter-University Consortium for Political and Social Research website and various scholarly journal databases, and consulted with scholars in relevant disciplines to identify additional data sets. We relied on study documentation available online to identify the study design and sample characteristics.

The inclusion criteria for data sets reviewed in this scan were as follows:

• Data available from the last 10 years (as of 2016, when analyses were conducted)
• National U.S. sample, with a few exceptions for rigorously designed surveys with large samples of Hispanics (e.g., Building Strong Families, Supporting Healthy Marriage, and Los Angeles Family and Neighborhood Survey)
• Presence of at least one of the following domains of family life: family and household composition; family formation and stability; relationship dynamics; parenting and co-parenting
• Large sample of Hispanics (at least 10 percent of the full sample, or 500 cases)
• Publicly available data and codebooks

In total, 22 data sets met our inclusion criteria (see Table 1). We reviewed the survey instruments used for each data set and assessed the extent to which they captured family life across four domains—(1) family and household composition, (2) family formation and stability, (3) relationship dynamics, and (4) parenting and co-parenting—with several dimensions under each domain.

Table 1. Design Characteristics of Large-scale Data Sets Relevant to Studying Latino Families and Household Experiences

We also assessed the extent to which the surveys included information on the 10 key data elements to unpack Hispanic diversity identified by the Hispanic Research Work Group. These key data elements are Hispanic ancestry/heritage subgroup, country of birth, parent country of birth, U.S. citizenship, time in the United States, language(s) spoken at home, English speaking proficiency, literacy in any language, highest educational level outside of the United States, and legal residency.^3,4

Many of the data sets reviewed include multiple waves or administrations. For cross-sectional surveys that contain repeated administrations, we describe the most recent assessment for which data were available at the time we initiated our review, unless otherwise noted. For longitudinal surveys, in most cases, we assessed the first wave of data collection with a few exceptions—notably, when a topical module of interest was included in later waves. For instance, the National Longitudinal Survey of Youth 1979 sample did not include an explicit focus on fathers until 1998. Similarly, the National Longitudinal Survey of Adolescent to Adult Health did not collect information on the focal child’s own children and their parenting until the third wave of assessment. In those cases, we reviewed the first instance when the relevant information was collected, in addition to the baseline survey. More broadly, all of our findings are based on reviews conducted in the latter part of 2016.

We summarized the availability of information for each dimension by indicating whether the study included (✔) or did not include (-) questions that capture that data element, whether extensive information was available (✔+), or whether partial information was obtained (✤). An example of extensive information for the dimension “relationship of responding adult to other adults in the household” (in Table 3) would be the availability of data on how all individuals in the household are related to one another, as opposed to data just on how the responding adult is related to other adults in the household. In most cases, a partial mark indicates that the measure provided incomplete information about the dimension of interest. For example, a partial measure for the number of individuals living in a household would be one that asked about the number of individuals under age 18 who lived in the household, but did not collect information on the number of individuals above this age. Information that could be partially inferred from another question or source (but was not directly asked) was also considered to be a partial measure. For example, if childbearing history was not inquired about directly, but if the number of children living in a household and their relationship to the respondent are known from a household roster, this would be considered a partial measure of childbearing. Additionally, when data are available for individuals other than the respondent, we specify whether data are available for the responding parent (“P”), child (“C”), up to two parents (“2P”), both partners (“BP”), or at the household level (“H”).

Findings

This section describes the data elements assessed, reports on their availability across data sets reviewed, and highlights key findings.

Data sources available to measure and describe Hispanic families and households 

Table 1 describes the methodological characteristics of the surveys reviewed in this scan, including the study design (e.g., longitudinal, cross-sectional), time frame, sampling frame, overall sample size and number of Hispanics in the sample, informant (e.g., parent, randomly selected household member), individuals for whom data are available (e.g., all household members, focal child), and whether geographic or other linking variable(s) are available. For surveys that had repeated assessments, we specify the specific survey wave or administration used for our review under the column labeled “featured survey and timeframe.” Exceptions to our general assessment rules are indicated in footnotes in Table 1 and in the relevant tables throughout.

The capacity of our current data infrastructure to adequately capture recent growth and diversification of the Latino population in the United States is mixed. On the one hand, seven out of the 22 (32 percent) surveys reviewed are repeated cross-sectional surveys and the majority are ongoing, thereby providing current snapshots of the Hispanic population and how it may be changing over time. On the other hand, most of the data sets reviewed (15 out of 22, or 68 percent) are longitudinal and most (10 out of 15) assembled their samples more than 10 years ago. These samples do not capture recent growth and diversification of Latinos in the United States, so our ability to understand changes in family life over time among newer subgroups of Hispanics may be hampered.

Collectively, data are available to examine Hispanic families and households from more than one perspective. However, few surveys collect data from both parents, limiting our ability to understand parenting and family dynamics. Fifteen data sets included more than one informant, often at least one parent and a child (nine data sets); only four data sets collected information from two parents.

The capacity is available to link geographic information and, perhaps, to understand the geographic diversity of Hispanics and how it shapes family life. With the exception of the National Survey of Family Growth, all data sets provide at least some geographic information, although the amount of information available varies greatly across data sets—ranging from a mere indication of the state where the participant lived to detailed geocoded data that can be linked to census and other data.

Measuring diversity in Hispanic family life

Table 2 shows the extent to which each data set includes information on the 10 key data elements needed to unpack Hispanic diversity.^3,4  Findings presented in this portion of the table are based on those reported in our previous brief, Improving Data Infrastructure to Recognize Hispanic Diversity in the United States. The right panel of this table summarizes the degree to which each survey captures the different dimensions within each of the four domains of family life examined. We also indicate whether information was collected for all dimensions within a domain (“ALL”), regardless of whether the information obtained was extensive, complete, or partial; whether information was obtained for some but not all dimensions (“SOME”); and whenever no information was obtained on any dimensions (“-”). Together, this information allows us to determine the extent to which we can understand the diversity of family life experiences among Latinos in the United States.

Table 2. Data Elements Measuring Key Characteristics of Latino Families and Households, by Data Set

The majority (19 out of 22, or 86 percent) of surveys had at least some information on five or more of the 10 key data elements for measuring Hispanic diversity, but only one (NAWS) had at least partial information on all 10 elements. Although most data sets (86 percent) collected information on both parent and child country of birth, only 13 of the 22 surveys inquired about U.S. citizenship. Both of these data elements are critical to understanding social capital and how it shape families’ experiences. The largest gap in information relevant to unpacking Hispanic diversity is related to legal status. Only four data sets obtained information on legal status (L.A. FANS, NAWS, NLSY79, SIPP), limiting our ability to understand how families fare when members lack legal status. Less than one-third of the data sets (six out of 22) collected information about parents’ literacy in any language and their educational attainment outside of the United States—critical components to understanding challenges to social integration and intergenerational mobility.

Overall, it was uncommon for data sets to collect information on both parents or all household members. For example, the Survey of Income and Program Participation was the only study that obtained information on U.S. citizenship status for everyone in the household. Roughly half of the data sets that included information about parent country of birth inquired about both parents’ country of birth, an important piece of information for establishing generational status.

All data sets collected data to describe Latino family and household composition. All include at least some data on current family status (e.g., married, cohabiting) and many collect information to examine changes in family status over time. Most data sets included at least some information to describe parenting in Latino families. However, there were pronounced gaps in our ability to measure, describe, and understand relationship dynamics among Latinos and, to a lesser degree, co-parenting (particularly among co-residential parents).

Data elements measuring key characteristics of Latino family and household composition

Table 3 summarizes the availability of data on family and household composition. Specifically, we assessed whether surveys included information on the number of individuals, children, and adults in the household, and the relationship of the responding adult to children and other adults in the household. These data allow researchers to determine household and family type and/or structure.^d

Table 3. Data Elements Measuring Key Characteristics of Latino Family and Household Composition, by Data Set

All data sets contained at least partial information on all dimensions of family and household composition. All surveys included complete information about the number of children in the household, and all included at least some information about the number of individuals and adults in the household. For the most part, information is available regarding the relationship of the responding adult to all children in the household (18 had complete information, and four partial) and to other adults in the household (all had at least partial information). However, in some cases (six surveys), only partial information is available about the relationship of the responding adult to other adults. Only a few data sets obtained detailed information about how all adults and children were related to all adults in the household. This information is needed to fully understand the arrangements of complex households (e.g., multi-family households).

Data elements measuring key characteristics of Latino family stability

Table 4 indicates the extent to which surveys include information about marital, cohabiting, and childbearing history. This table also contains information about whether surveys included questions about family (in)stability. Specifically, we searched for the availability of information regarding the number of family or relationship transitions (i.e., change in romantic residential relationship status), the type of transition (i.e., change in residential, marital, or relationship status), and the timing of these transitions.

Table 4. Data Elements Measuring Key Characteristics of Latino Family Stability, by Data Set

Nearly half of the surveys (10 of 22) captured, to some degree, all dimensions of family formation and stability examined. However, eight of the 22 surveys collected only partial or no information on this domain. All data sets asked about at least partial marital history, and most asked about cohabitation history (82 percent) and childbearing history (82 percent). Half (11 of 22) of the surveys asked about the number of family or relationship transitions, roughly two-thirds inquired about the type of transitions experienced, and over half asked about the timing of these transitions. Notably, surveys rarely contained information about family formation and family stability for both parents.

Data elements measuring key characteristics of Latino relationship dynamics

Table 5 focuses on the presence of data on six dimensions of relationship dynamics: relationship quality, happiness, communication, conflict, physical violence, and intimacy.

Table 5. Data Elements Measuring Key Characteristics of Latino Relationship Dynamics, by Data Set

In general, information on couple relationship dynamics is extremely limited. More than half of the data sets (12) did not contain any information on this domain of family life. Only two data sets, Building Strong Families and Supporting Healthy Marriage, inquired about all six dimensions. Both of these data sets are based on samples of couples who participated in evaluations of government-funded programs aimed at strengthening the relationships of low-income couples with their young children. Currently, no national data set is available that can speak to the characteristics of couples who were not targeted by these programs, or couples who may benefit from such programs but were not included in the studies.

Data on relationship dynamics is rarely collected, but when it is, it is often collected from both partners. Of the 10 surveys that collected information on relationship dynamics, seven did so from both partners, thereby allowing researchers to understand relationship experiences from the perspective of both partners, and how these experiences may differ.

Conflict is the most commonly available data element related to relationship dynamics, yet only six of the 22 surveys include it. Roughly one in four data sets included information on happiness and physical violence. Data on relationship quality and physical intimacy were collected in four surveys; only two surveys collected data on how couples communicate.

Data elements measuring key characteristics of Latino parenting and co-parenting

Tables 6 and 7 document the extent to which surveys included information on the six dimensions of parenting and co-parenting. The aspects of parenting reviewed here are parent-child activities (e.g., having meals together, reading); parent-child relationship/parenting behaviors (e.g., communication, monitoring); parenting knowledge, attitudes, and beliefs (e.g., educational expectations, parenting values); and fathers’ early involvement (e.g., participation in prenatal visits). Co-parenting refers to the ways in which both parents come together in their parenting roles. For this dimension, we searched for the inclusion of items regarding parents’ attitudes about co-parenting, relationship quality with the co-parent, roles and responsibilities, and conflict. Finally, we reviewed surveys for the availability of information about parental support and involvement in children’s lives among nonresidential parents. Because information on the dimensions examined was often available from several sources, we indicate the extent to which the survey included questions that reflect each dimension collectively for the mother (m), the father (f), and the resident (r) or nonresident (nr) parent.

Table 6. Data Elements Measuring Key Characteristics of Latino Parenting, by Data Set: Parent-Child Activities, Relationships/Behavior and Parenting Knowledge, Attitudes, and Beliefs

Overall, our review indicates that most data sets lack complete information about parenting and co-parenting. Indeed, five of the 22 surveys reviewed contained zero or just one of the 28 dimensions examined. Early paternal involvement has received little attention, with 16 (73 percent) of the data sets lacking information in this area. On the other hand, nearly three-quarters of the surveys (73 percent) examined parent-child activities, and nearly two-thirds (64 percent) examined parenting behaviors to some degree. For information about co-parenting and nonresidential parents’involvement, more information is available on nonresidential fathers than on how the two co-residential parents come together in their parenting roles. Below, we describe our findings across the six areas of parenting and co-parenting examined.

Roughly three-quarters of the surveys contained information on activities that parents and children do together, but in many cases, information was collected at the household level without specifying the person interacting with the child. For parent-child activities, most surveys collect information that falls into “other” general activities with parents (14 surveys), which includes activities like going to the museum, sporting events, and running errands. Ten surveys asked whether, or how often, the parent read to the child. A smaller number of surveys (nine) included questions about learning opportunities and activities at home (e.g., playing with blocks, letters, numbers). Approximately one-third of the surveys (eight) asked whether or how often the family ate meals together. Fewer surveys (seven) included questions about involvement in physical activities with the child (e.g., playing sports); general caregiving (five), including feeding, bathing, and dressing the child; and homework (four).

The communication between parent and child (including frequency, quality, and content) is the aspect of the parent-child relationship and parenting behavior that is most commonly inquired about; still, only about half (12 of 22) of the data sets included items in this area. Fewer than half (10) of the surveys contain data on the quality of the parent-child relationship, which includes warmth, affection, and global relationship quality. Less than half (nine) of the data sets included items on parental monitoring and rules. Questions in this dimension often asked about parents’ knowledge of their children’s whereabouts, but some asked about monitoring and rules regarding specific activities (e.g., watching television). Measures of parental behaviors and attitudes regarding discipline are included in more than one-third (eight) of the data sets. Nearly one-third (seven) of the surveys ask about parents’ involvement in school.

For parenting knowledge, attitudes, and beliefs, we reviewed the extent to which surveys included information about parental educational expectations, parental self-efficacy, parenting values and beliefs, and parental knowledge about parenting and child development. At most, data sets included three of the four dimensions of parenting knowledge, attitudes, and beliefs. Parenting values and beliefs appeared in 41 percent of the surveys. Only one survey (ECLS-B) included questions about parents’ knowledge about parenting and child development.

Fathers’ early involvement includes questions about fathers’ participation in prenatal visits, their presence at birth, and whether paternity was established. We found that only two data sets, both birth cohorts (ECLS-B and FFCWS), collected data on all three dimensions of fathers’ early involvement. Just six data sets asked questions about whether parents had established paternity, and only two collected data on fathers’ participation in prenatal visits and their presence at birth.

Table 7. Data Elements Measuring Key Characteristics of Latino Parenting, by Data Set: Father’s Early Involvement, Co-parenting, Nonresidential Parent Support and Involvement

To determine how much information our current data infrastructure provides about co-parenting, or the ways in which two parents work together and relate to each other in their parenting roles, we reviewed the extent to which surveys assessed parents’ attitudes toward co-parenting, the quality of the relationship between the two parents, their roles and responsibilities in childrearing, and communication and conflict about co-parenting. Our review indicated that 13 of the 22 data sets contained at least some information on co-parenting; only one asked questions on all co-parenting dimensions examined. The most common dimension of co-parenting assessed—in eight of the 22 data sets—was the quality of the relationship with the co-parent. Nearly one-quarter of the surveys included questions about attitudes toward co- parenting. Only four data sets asked about communication or conflict in the co-parenting relationship, and three asked about roles and responsibilities regarding childrearing. Questions were generally asked from the perspective of one parent.

We also reviewed the degree to which surveys included information about nonresident parents. In this area, we searched for the inclusion of questions about legal arrangements (e.g., custody agreements), financial aid provided by a nonresident parent (both formal, through child support, and informal), in-kind support provided by a nonresident parent, and information about the amount of time the nonresident parent spends with the child. Seventeen data sets included at least some information about the nonresidential parent’s support and involvement in their child’s life. Most of the information obtained addressed the time spent with children (68 percent included questions in this dimension). For the types of support provided by the nonresident parent, most data sets focused on the provision of financial support (59 percent), and only five asked about other types of nonfinancial support (e.g., social support). Eight data sets included questions about legal arrangements (e.g., custody).

Summary and Implications

Our review of 22 primarily national surveys suggests that the current data infrastructure allows us to describe how Latino families are structured, but that data are limited for understanding family life beyond basic sociodemographic descriptors. Specifically, there is insufficient information about relationships, and parenting and co-parenting. Overall, the data available may not represent the current demographic composition of the Hispanic population in the United States.

Where information exists

There are adequate data to describe how Hispanic families are structured, which is critical to our understanding of Hispanic family life. Family structure and stability are key indicators of family well-being that are related to both parental and child outcomes.5-7 With the current data infrastructure, we are able to answer questions about the number of individuals that live in the same household, whether children are growing up in single or two-parent households, and (to some degree) whether extended family members and nonrelated individuals live in the household. Many surveys also collect information about changes in family composition. Importantly, many existing data sets contain sample sizes of Latinos large enough for subgroup analyses (e.g., by nativity status and, to a lesser extent, country of origin).

The information available can be used to identify potential resources and stressors that are present in Latino families, and how these differ for foreign-born and U.S.-born Latinos. Program developers can rely on this information to define their target population and the appropriate timing of interventions. For example, analyses using national data sets like those reviewed here have revealed that most births to low-income Hispanics occur in two-parent unions,^8,9 and that the first five years of a child’s life are characterized by high levels of family stability.10 Thus, the first few years following a child’s birth are prime periods to engage fathers in parenting programs, and engage couples in family strengthening interventions. The importance of a child’s early years also calls attention to the need for programs to support two-parent unions, in addition to single-parent families.

Where gaps exist

Our review indicates that current surveys do not adequately capture the more dynamic aspects of family life that are indicative of how a family functions, including couple dynamics, parenting, and co-parenting. For example, existing surveys allow us to determine whether fathers are present in Hispanic children’s homes, but there is insufficient information about how fathers interact with their children, how the two parents come together in their parenting roles, or the quality of the relationship between the two parents. Importantly, most surveys do not capture fathering; when they do, they focus on nonresidential fathers. This gap is especially noteworthy for Hispanic families, given that most Hispanic fathers live with all of their children.¹¹

The absence of information on family processes hinders our ability to design programs and policies that respond to the specific needs and challenges of Latino families. Previous work has shown a deficit of marriage programs specifically designed with Latino populations in mind.¹² Programs could be optimized with the aid of research that focuses on the unique challenges that Latinos relationships face, and by devising specific strategies to support Hispanic families. Only two surveys contain extensive information about couples’ relationship dynamics. In both cases, the samples are select groups of individuals who agreed to be part of an evaluation of programs targeting couples, but who may not represent Latino couples typically not reached by these programs.

In addition, there continues to be insufficient information about critical variables that speak to the diverse experiences of Latino families in the United States. Our and others’ work have noted divergent experiences among immigrant and nonimmigrant families and individuals. As we continue to delve into these differences, we must better understand the drivers of those differences: legal or citizenship status, linguistic isolation, English language fluency, education, or other factors. Each explanation points to different challenges and implications for programs and policies. To address these questions, adequate data are needed that capture the diverse experiences of Latinos. Legal status is missing in most surveys, and parental literacy and educational attainment outside the United States are also not often collected. These three elements play a central role in determining Hispanic families’ access to resources and opportunities, and their potential for upward mobility. However, obtaining sensitive information like legal status presents challenges that must be considered when designing future surveys.

Importantly, our nation’s data infrastructure has not kept up with national demographic shifts. The Latino population has not only grown significantly over the recent decades, but has diversified and dispersed to areas with little previous representation of Latinos. Specifically, a sizable increase in the presence of Hispanics in the Southeast and in rural and suburban communities since the 1990s has altered the demographic landscape of the United States.^1,13 The countries of origin for Latinos are also more diverse. Although repeated cross-sectional surveys likely adequately capture these demographic shifts—especially by providing potential links to geographic data—many of the longitudinal data sets reviewed compiled their samples prior to these changes and may not reflect these emerging Hispanic communities.

Implications

This review signals a need for improvements to the nation’s data infrastructure that enhances our ability to study and serve Hispanic families and children. Future data collection efforts should assemble a new population-based cohort that captures the current demographic composition of the United States and follows participants over time. The survey should obtain more granular demographic information that can unpack the diversity within Latino families and inquire about indicators of family functioning beyond family structure (i.e., couple dynamics, parenting, and co-parenting, and from both parents’ perspectives). Even though some studies not reviewed here have focused specifically on family dynamics among Latinos, they tend to be small and based on convenience samples that do not represent the diversity of the Hispanic population in the United States. While the task at hand may be perceived as daunting, the foundation for a future national survey of families and households exists in our current array of surveys.

Caveats

Our review is constrained by our decision to select the first (in longitudinal data sets) and most current (in repeated, cross-sectional data sets) survey available for review, with a few exceptions. We recognize that, in doing so, we may not capture a data set’s full potential to inform on Hispanic family life. However, we based our review on baseline surveys (in the case of longitudinal data sets) because these usually contain the most comprehensive demographic information and the most complete data on their sample, compared with later waves that often have attrition.

For repeated cross-sectional data sets, we selected the latest survey available for review, as it generally contains the most current information. More generally, we provide a rich array of information that can serve as a resource to researchers interested in studying Hispanic family life and guide future plans for data collection efforts.

Notes

^a In this brief, we use the terms Hispanic and Latino interchangeably. Most large-scale surveys included in this review give respondents the option of identifying themselves (or their children) as being “of Spanish, Hispanic, or Latino origin.”

^b Other important dimensions not examined here include family and household emotional and economic well-being, child abuse or neglect, family stress, and parenting by nonbiological parents. We also did not search for the availability of information on culturally relevant factors that influence family life, such as culture-related intergenerational conflict, religiosity, and acculturation gap between parents and their children. These exclusions were primarily due to the absence of such information in national data sets and a desire to contain the scope of the project

^c The 10 data elements include Hispanic heritage, child and parent country of birth, U.S. citizenship, time spent in the United States, language used in the home, English language proficiency of the parent, parent literacy in any language, parental educational attainment outside of the United States, and legal status. For more information on the 10 data elements needed to measure diversity within the Hispanic population, see https://www.acf.hhs.gov/opre/report/survey-data-elements-unpack-diversity-hispanic-populations.

^d Household and family type often, but not always, overlap. The household refers to all persons who occupy a housing unit, whereas family refers to those linked by blood, marriage, or adoption.