In the past decade, policymakers and researchers have become increasingly interested in social programs that promote and support healthy marriages.В A growing body of research evidence suggests that marriage has benefits for families and children, including improved economic well-being and mental health, and that children raised in two-parent families perform better in school and have more positive developmental outcomes than children from single-parent families (Amato and Booth 1997; McLanahan and Sandefur 1994; Waite and Gallagher 2000; Wood et al. 2007).В Inspired in part by these potential benefits of marriage, a wide range of programs have been developed to encourage and support healthy marriages (Dion 2005).
Reflecting this growing interest in healthy marriage programs, the U.S. Department of Health and Human Services (DHHS) has recently sponsored efforts to expand the understanding of the effectiveness of these programs and to support their expansion through funding and technical assistance, as part of the Healthy Marriage Initiative (HMI).В For example, the HMI currently supports three large-scale, multi-state demonstration projects:В (1) Building Strong Families, an evaluation of programs to help expectant unwed couples fulfill their aspirations for a healthy marriage and a stable family life; (2) Supporting Healthy Marriage, a project to develop and test healthy marriage programs for low-income married parents; and (3) the Community Healthy Marriage Initiative, an evaluation of community-level interventions to support healthy marriages.В In addition, with funding from the Deficit Reduction Act of 2005, DHHS has awarded grants to a diverse range of state and local agencies to increase access to healthy marriage programs and raise awareness of the potential benefits of marriage for families and children.
As interest in healthy marriage programs continues to grow and new programs are developed, a key issue policymakers and program operators will face is deciding which populations to serve.В Healthy marriage programs aim to serve a broad mix of target populations, including expectant unmarried parents, low-income married parents, high school students, engaged couples, single adults, and other groups.В The design and content of the programs can vary substantially, depending on which of these populations are served.В Policymakers will also need to make choices about whether to focus their programs on specific social or demographic groups, such as residents of certain cities or counties, individuals living in rural or urban areas, or members of certain racial/ethnic groups.
The Marriage Measures Guide is designed to assist policymakers and marriage program operators with this decision making process.В Drawing on data from several sources, the guide provides policymakers and program operators with a broad range of state-level statistical information they can use to better assess the characteristics and needs of their state populations, identify high-priority target populations, and make informed decisions about the design and implementation of their healthy marriage programs.В The guide can also help policymakers decide which healthy marriage programs are best targeted to their statewide populations and which are more appropriate for local or targeted groups.В In addition to these uses for the development and implementation of healthy marriage programs, the guide also serves as a general resource for anyone wanting to better understand current marriage patterns in their state.
This chapter provides a general introduction to the guide and the best ways to use it.В To make the most of the guide, users should read this chapter carefully before turning to the statistical tables.В The chapter begins by describing the overall content and layout of the guide.В It then describes in greater detail the content of the statistical tables and explains how to correctly interpret each statistic.В The chapter ends with a brief discussion of additional resources.В A more detailed discussion of the various data sources and methods used to construct the statistical tables appears at the end of the guide in the Technical Appendix.
Overview and Layout
The Marriage Measures Guide is a collection of stand-alone five-page reports for each of the 50 U.S. states. Each report has the same layout and content. The report begins with a one-page summary sheet that briefly describes the size and racial/ethnic makeup of the states population and highlights key findings from the companion statistical tables. Following the summary sheet, there are four pages of detailed statistical tables, with information to help local policymakers and program operators better understand the characteristics of their state populations. The main topics covered in the tables include (1) marriage and childbearing, (2) marriage and divorce, and (3) marriage and low-income children. We describe these topics in greater detail below. After reading this introductory chapter, most readers should use the guide by turning directly to the summary sheet and statistical tables for their individual state. However, important information can also be gained by scanning the tables for several states to develop points of comparison.
The information included in the Marriage Measures Guide can be used to answer the following types of questions policymakers might have as they design and implement their healthy marriage programs:
- What percentage of children in my state are born to unmarried women?
- What is the overall racial/ethnic makeup of my states population?
- How does the percentage of births to unmarried women vary among racial/ethnic groups?
- What is the overall divorce rate in my state?
- How does the divorce rate in my state rank in comparison to those of other states?
- How many low-income children live in my state?
- What percentage of the low-income children in my state live with married parents?
- Do most of the low-income children in my state live in rural or urban areas?
For most of these statistics, the guide also provides national and regional benchmark estimates to help identify the distinctive characteristics of each state, information policymakers can use to better understand marriage patterns in their state and to design an approach that best serves the needs of their local populations.
Marriage and Childbearing
The first statistical table for each state presents information related to marriage and childbearing. Rates of nonmarital childbearing vary substantially by state and these rates may provide useful guidance to policymakers as they design healthy marriage programs. For example, states such as Louisiana, New Mexico, and Mississippi have very high rates of nonmarital childbearing. Therefore, these states may want to place a particular focus on unmarried expectant partners when designing their marriage programs. By contrast, states such as Utah, Idaho, and New Hampshire have low rates of nonmarital childbearing and thus may not want to target this population statewide. In addition to this variation among states, rates of nonmarital childbearing vary substantially among demographic groups. For example, in many states, the percentage of births to unmarried women is highest among women who are younger, less educated, and from racial or ethnic minority groups. Therefore, policymakers might want to target these groups when designing healthy marriage programs for unmarried expectant parents.
The statistics included in this table are based on analyses of 2004 data from the National Center for Health Statistics. The key statistics are as follows:
- Number of Births to Unmarried Women. Reported in the first column of the table, this statistic is an annual measure that includes all births registered to unmarried women in 2004. The second column in the table reports the total number of births to all women in the state, regardless of marital status.
- Percentage of Births to Unmarried Women. Reported in the third column of the table, this statistic expresses the number of births to unmarried women (column 1) as a percentage of the total number of births to all women (column 2). Higher numbers correspond to a higher percentage of births to unmarried women. This statistic was used to determine the overall ranking of states by the percentage of births to unmarried women, reported on the summary sheet for each state. The ranking of states by percentage of births to unmarried women is also displayed near the end of the guide in Appendix A.
- Subgroup Estimates for Selected Demographic Groups. The rows of the table report separate estimates for key population subgroups, defined by the mothers demographic characteristics. The subgroup estimates for geographic areas report separate statistics for the one or two largest counties in each state, as well as a combined estimate for all other counties. This county-level information is available for all states except Wyoming (see Technical Appendix for details).
- Regional and National Benchmark Estimates. The far right-hand columns of the table show how the percentage of births to unmarried women in the state compares to regional and national averages. The national averages include data for all 50 states. The regional estimates are based on the standard nine geographic divisions defined by the U.S. Census Bureau (Figure I.1).
Marriage and Divorce
The second table for each state reports statistics related to marriage and divorce. Divorce rates vary widely by state and this information could provide useful guidance to state policymakers and program operators designing healthy marriage programs. States such as Arkansas, Nevada, and Wyoming have high divorce rates. For this reason, these states may want to focus their marriage programs on reducing divorce among married couples statewide. By contrast, states such as Massachusetts and Pennsylvania have low divorce rates and thus may want to make their programs for married couples more local or targeted. Divorce rates also vary by demographic characteristics such as gender, race/ethnicity, education level, and rural or urban residence. Information on these differences may help policymakers further target their healthy marriage programs to key segments of their state populations.
Geographic Divisions Defined by the U.S. Census Bureau
The divorce statistics reported in the top half of the table are based on administrative records collected by the National Center for Health Statistics. Data are available for all states except California, Georgia, Hawaii, Indiana, Louisiana, and Minnesota. The three main statistics reported in this part of the table are defined as follows:
- Number of Divorces Granted. This statistic indicates the total number of divorces granted in the state in 2005, as reported to the National Center for Health Statistics by various state agencies. The figures include reported annulments and count more than one divorce by the same person as separate events. The regional and national estimates exclude the six states for which data are not available.
- Divorce Rate. The states divorce rate is calculated by dividing the total number of divorces granted in 2005 by the size of the states population. The resulting statistic indicates the number of divorces granted per 1,000 residents. The statistic should not be interpreted as the number of marriages that end in divorce or the number of divorces a person can expect in his or her lifetime, because it is based on data from a single year and does not account for the timing of divorce. Rather, the statistic provides a rough estimate of the average number of people who became divorced during 2005, adjusting for population size and not accounting for people who filed for more than one divorce.
- State Rankings. The state rankings by divorce rate run from highest to lowest among the 44 states for which data are available. For example, Kentuckys rank of ninth means that it has the ninth-highest divorce rate among the 44 states that report these statistics.
The additional divorce statistics reported in the bottom half of the table are based on survey data from the 2006 American Community Survey (ACS), a large nationally representative survey of U.S. households conducted annually by the U.S. Census Bureau. Key statistics reported in this part of the table include the following:
- Number of Divorced People. This statistic is reported in the first column of the table and indicates the total number of people who reported their current marital status as divorced in 2006. The sample is limited to adults ages 15 and older. Because these totals refer to current marital status, they do not count people who have remarried following a divorce. The numbers would be higher if they included people who have ever been divorced.
- Percentage of People who Are Divorced. This statistic expresses the number of people who reported their current marital status as divorced (column 1) as a percentage of the total number of people in the group who have ever been married. The sample is limited to adults ages 15 and older. For example, the national average reported in the bottom row of the far right-hand column indicates that about 15 percent of ever-married adults in the United States reported their current marital status as divorced in 2006. This statistic does not represent the percentage of all marriages that will end in divorce, because it is based on data for a single year and does not account for remarriage. Rather, the statistic provides a snapshot estimate of the relative size of the divorced population in 2006. Moreover, as explained in the Technical Appendix, the statistic can also be viewed as a rough proxy for the states overall divorce rate, at least for the purpose of making basic rankings or comparisons of states.
- Subgroup Estimates. The rows of the table report separate estimates for key demographic subgroups. For example, the national estimates for men show that about 13.7 percent of ever-married men in the United States reported their marital status as divorced in 2006. To calculate the subgroup estimates for rural and urban areas, we merged geographic information from the 2000 U.S. Census with survey data from the 2006 ACS. For reasons explained in the Technical Appendix, this approach may overstate the size of the urban population in some states. Moreover, the Census definitions of rural and urban areas are based on statistical criteria that may not correspond with how state or local residents define these areas. For example, some small towns or lightly populated areas defined as rural by local residents may be classified as urban in Census tabulations, especially areas that are near larger towns or cities. Therefore, readers should interpret these statistics with caution. The symbol NA means that the information is not available because there are fewer than 5,000 ever-married adults in that area or group.
Marriage and Low-Income Children
The third and fourth tables for each state report statistics related to marriage and low-income children. Much of the policy interest in marriage and relationship quality stems from concerns about how parental marital status affects the well-being of children, particularly low-income children. Therefore, information concerning the distribution of low-income children across various family types would be very helpful for policymakers setting priorities for healthy marriage programs. For example, if a state has a particularly high concentration of low-income children living with married parents in rural areas, then state policymakers may want to focus their resources on programs for low-income married parents in these areas. Alternatively, if most low-income children in the state live with unmarried parents in urban areas, then state policymakers may want to focus their healthy marriage programs on this population.
The statistics in these tables are based on survey data from the 2006 ACS, the same data source used to calculate state-level divorce statistics (described earlier). Some of the key statistics reported in these tables include the following:
- Distribution of Children Living in Low-Income Families. The top panel of the third table for each state shows the overall distribution of low-income children by family type. For example, the national averages in the far right-hand column of the table indicate that 45 percent of the countrys low-income children are living with married parents, 15 percent are living with never-married single parents, and 18 percent are living with single parents who are separated, widowed, or divorced. The data set does not distinguish between biological parents, adoptive parents, and stepparents. In our analyses, we defined low-income children as those living in families with incomes below 200 percent of the federal poverty level. Most of the children classified as having unknown family types live in complex multifamily or multigenerational households that cannot be accurately distinguished in the 2006 ACS data (see Technical Appendix for details). Nationally, the unknown category accounts for less than 10 percent of all low-income children. Additional statistics on the number of low-income children in each state are presented near the end of the guide in Appendix A.
- Distribution of Low-Income Children Between Rural and Urban Areas. The subgroup estimates for rural and urban areas (third table) were calculated following the same approach we used to calculate geographic differences in state-level divorce statistics. As explained earlier, this approach may overstate the size of the urban population in some states. Moreover, the definitions of rural and urban areas follow statistical standards established by the U.S. Census, which may not correspond with how state or local residents define these areas. For these reasons, readers should use caution when interpreting the statistics for rural and urban areas. The symbol NA means that the information is not available because either (1) the area includes a small group of fewer than 5,000 low-income children or (2) the specific family type listed accounts for less than 2 percent of the areas low-income children. Appendix A near the end of the guide reports additional statistics on the total number of low-income children living in rural and urban areas for nine regions of the country.
- Distribution of Low-Income Children by Racial/Ethnic Background. The fourth table for each state reports racial/ethnic differences in the distribution of low-income children across family types. The estimates for whites and African Americans are limited to non-Hispanics in these groups. The estimates for Hispanics include children from all races. The symbol NA means that the information is not available because either (1) there are fewer than 5,000 low-income children in that racial/ethnic group or (2) the specific family type listed accounts for less than 2 percent of the groups low-income children. Appendix A near the end of the guide reports additional statistics on the number of low-income children by racial/ethnic background for nine regions of the country.
A more detailed discussion of the data sources and methods used to calculate the statistics reported in this guide appears in the Technical Appendix. In addition, readers interested in obtaining additional statistics or conducting their own analyses of state-level data should consult the following sources:
- National Center for Health Statistics. The National Center for Health Statistics (NCHS) collects and reports a broad range of state-level marriage statistics. Basic counts of the number of marriages, divorces, and births in each state are published in a monthly series of National Vital Statistics Reports. State-level data on nonmarital childbearing can be accessed and analyzed online using the interactive VitalStats website. Most of the state-level data available from NCHS are based on administrative records collected from states, not from surveys of state residents.
- American Community Survey (ACS). The ACS is a new national survey conducted annually by the U.S. Census Bureau as an alternative to the long form of the decennial Census. The ACS is especially well suited for calculating state-level marriage statistics because it has an extremely large sample size. For example, the 2006 ACS collected social and demographic information from more than 1.2 million households. Basic tables and statistics based on ACS data can be accessed online through the interactive American FactFinder website. More detailed analyses can be conducted by downloading the ACS public use microdata sample (PUMS), which includes individual-level survey responses for an anonymous sample of ACS respondents.
- Current Population Survey (CPS). The CPS can also be used to calculate state-level marriage statistics. The CPS is conducted on a monthly basis, but the broadest range of social and demographic information is collected in a special supplemental survey administered from late February through early April, called the Annual Social and Economic (ASEC) Supplement. Data from the CPS can be used to calculate a range of state-level statistics related to marriage and family structure. State-level statistics based on CPS data are generally less precise than comparable statistics based on data from the ACS, because the CPS has a smaller sample size. However, the CPS also has advantages over the ACS, including more detailed measures of family income and household structure. Data from the CPS are available through a range of Census Bureau reports and from the CPS website.
 The tables for each state are labeled using the states two-letter postal code followed by the table number. For example, the first table for Alabama is labeled Table AL-1.
 National Vital Statistics Reports. Available at: [http://www.cdc.gov/nchs/products/pubs/pubd/nvsr/nvsr.htm]. Accessed November 21, 2007.
 VitalStats. Available at: [http://www.cdc.gov/nchs/VitalStats.htm]. Accessed November 21, 2007.
 American FactFinder. Available at: [http://factfinder.census.gov]. Accessed November 21, 2007.
 Current Population Survey (CPS). Available at: [http://www.census.gov/cps/]. Accessed November 21, 2007.