SECTION A:

BRIEF DESCRIPTION OF THE

CHILD AND YOUTH WELL-BEING INDEX (CWI)

The Child and Youth Well-Being Index (CWI) is an evidence-based measure of trends over time in the quality of life or well-being of America's children and young people.  It comprises several interrelated composite/summary indices of annual time series of numerous social indicators of the well-being of children and young people in the United States.  The composite indices give a sense of the overall direction of change in well-being, as compared to two base years of the indicators, 1975 and 1985. 

This Index addresses the questions such as the following: 

n      Overall, on average, how did child and youth well-being in the United States of America change in the last quarter of the 20th century and beyond? 

n      Did it improve or deteriorate? 

n      By how much? 

n      In which domains or areas of social life? 

n      For specific age groups? 

n      For particular race/ethnic groups? 

n      And did race/ethnic groups disparities increase or decrease? 

Methods of Construction:  To address these questions, the following index construction methods have been applied (for a complete description of research methods, see the Land, Lamb, and Mustillo, 2001, article accessible in Section I).   Annual time series data (from vital statistics and sample surveys) on some 28 national-level social indicators (for a complete list, see the table in Section C) in seven quality-of-life domains family economic well-being, social relationships (with family and peers), health, safety/behavioral concerns,educational attainment, community engagement (participation in schooling or work institutions), and emotional/spiritual well-being have been assembled.  Twenty-five of these time series date back to 1975 or earlier, while three are based on indicators that commenced in the mid-1980s. [1] 

The seven domains of quality of life or well-being have been well-established in over two decades of empirical research on subjective well-being by numerous social psychologists and other social scientists.  In this sense, the Child and Youth Well-Being Index is an evidence-based measure of trends in averages of the social conditions encountered by children and youth in the United States.

The 25 time series that date back to 1975 have been indexed by percentage change from the base year 1975-that is, subsequent annual observations are computed as percentages of the base year values. The three time series beginning after 1975 are also indexed by percentage change from the base year; however, their base years correspond to the year that specific series began. The overall index for 1975 takes into account the fact that some of the component time series did not begin in 1975, and thus utilizes all 28 time series indicators for the years subsequent to the three that commence in the 1980s (see footnote 1). Similarly, all 28 time series that are available by 1985 also have been indexed separately with 1985 as the base year. 

Equally-Weighted Summary Indices:  The time series of social indicators have been grouped into the seven domains of well-being identified above and a domain-specific summary well-being index has been constructed.  In this summary index, each of the component social indicator time series is equally weighted.  The seven component indices then are combined into two equally-weighted summary indices of child well-being-the first of which is based on the 25 social indicator time series that date back to 1975, as well as the three indicators that date back to the mid-1980s, and the second of which is based on the 28 time series that date back to 1985. 

Research on a model of summary index construction by Hagerty and Land (2006), accessible in Section I, shows that, in the absence of evidence-based data on importance weights for component indicators, the equally-weighted method of index construction is likely to receive the highest consensus among all possible weighting schemes.  Tables reporting the annual index values of all of the component social indicators of our Index are accessible in Section H.  From these, it is possible to construct alternative summary indices based on any possible weighting scheme.

Sensitivity of Trend Analyses to Base Years and Component Indicators and Domains:  Our research (reported in the Land, Lamb, and Mustillo, 2001, article accessible in Section I) has shown that conclusions about trends in child well-being in the United States in the last quarter of the 20th century depend on both the base year and the formula by which the summary indices are calculated.  Findings about child well-being also are dependent upon the specific indicators and domains used in the composition of the summary indices. 

Projected Values for Recent Years:  Since many of the economic, demographic, and social statistical time series used to construct the Child and Youth Well-Being Index are published with a one to two year lag behind the current year, time series models have been estimated for all of the component social indicator time series of the Index.  These models then have been used to project values for each component indicator to the current year and the corresponding projected values of the domain-specific and summary indices have been calculated.  The graphical representations of the overall summary and domain-specific indices in Section D and Section E show these projected values for the most recent two years.

What Composite/Summary Indices Can and Cannot Tell Us:  In its broadest sense, an index number, such as the Child and Youth Well-Being Index, is a measure of the magnitude of a variable at one point (say, a specific year that is termed the current year) relative to its value at another point (called the reference base or base year).  The index number problem occurs when the magnitude of the variable under consideration is nonobservable.  In economics, where index numbers are widely used, this is the case, for example, when the variable to be compared over time is the general price level, or its reciprocal, the purchasing power of money.  In the present case, the variable to be compared over time is the overall well-being of children and young people in the United States - defined in terms of averages of social conditions encountered by children and young people.  In economics, the index number problem arises in measuring the general price level due to the fact that there are multiple indicators of well-being to be compared.  Over any given historical period, the prices of some economic goods will have risen and some will have fallen.  Similarly, over any period of years, some indicators of child well-being likely will have risen and some will have fallen.

A key point is that in any given year no single consumer is likely to purchase all of the items that comprise the market basket of goods used in constructing the Consumer Price Index.  On the other hand, fluctuations over time in the Consumer Price Index signal changes in general price levels that generally are encountered by consumers, and most consumers are interested in how the general price level is changing.  Similarly, in any given year no single child encounters all of the social conditions that enter into the overall community engagement that is developed in this paper.  Fluctuations over time in the Child and Youth Well-Being Index can be taken, however, as signaling changes in the overall context of social conditions encountered by children and youths.  And many policy makers, officials, adults, and parents (and some children and youths as well) are interested in how the general level of social conditions faced by children in a recent year, such as 1998, compares to the corresponding level in a previous year, such as 1985.

For additional commentary on the uses and abuses of social indicators see The Uses (and Misuses) of Social Indicators:Implications for Public Policy by Kristin A. Moore, Brett V. Brown, and Harriet J. Scarupa..

For More Details and Causal Understanding of Trends in Well-Being:  In all cases, in order to more fully understand the indications of year-to-year changes in overall child well-being given by our Index and its several variants, one must study the corresponding changes in the domain-specific indices and their component social indicators.  And, for an even deeper understanding of the causal mechanisms lying beneath the year-to-year changes in the summary indices, domain-specific indices, and component social indicator time series, one often must access specific detailed studies in the social sciences, demography, and epidemiology.  A number of such studies are cited in the Land, Lamb, and Mustillo (2001) article accessible in Section I.

Links to Other Sites with Child Well-Being Statistics and Indicators:  Two other Web sites that contain a wealth of social indicators, statistics, and statistical time series on child and youth well-being are:

n      the Child Trends Data Bank (http://www.childtrendsdatabank.org/) which provides data and analyses for over 70 indicators on the well-being of children, including many of those used in the Child Well-Being Index Project; and

n      the federal ChildStats data site (http://www.childstats.gov/) which offers easy access to federal and state statistics and reports on children and their families; this includes the annual reports of the Federal Interagency Forum on Child and Family Statistics - America's Children:  Key National Indicators of Well-Being.  

Acknowledgments:  The Child and Youth Well-Being Index Project is coordinated by Kenneth C. Land, Ph.D., John Franklin Crowell Professor of Sociology and Demography, Department of Sociology, P.O. Box 90088, Duke University, Durham, NC 27708-0088 (e-mail: kland@soc.duke.edu).  Other researchers involved in the project include Vicki L. Lamb, Ph.D., Professor of Sociology, North Carolina Central University, and Xiaolu Zang, M.Phil., Duke University.  The Project is supported by grants from the Foundation for Child Development (http://www.fcd-us.org/).  We especially acknowledge the support and encouragement of Deborah Phillips, President, and Mark Bogosian, Communications and Grants Officer, Foundation for Child Development.  We also thank Kristin A. Moore, Ph.D. and Brett Brown, Ph.D. of Child Trends, Inc. (http://www.childtrends.org), Donald Hernandez, Ph.D. of the Graduate School at Hunter College, and the FCD-CWI Advisory Board for invaluable advice and assistance in this project.

General questions and data inquiries can be addressed to the following individuals. For general questions about the Child Well-Being Index, contact Kenneth C. Land at the address given in the preceding paragraph. For questions concerning data sources and the database used for the construction of the CWI and its component indices, contact Xiaolu Zang (xiaolu.zang@duke.edu).  

 



[1] The percent of children with health insurance begins in 1987, while the percent of children who's health is rated very good or excellent as well as the percent of children with activity limitations begin in 1984.