1

FCD CWIFCD CWI

HOW ARE THE KIDS DOING?HOW DO WE KNOW?

Recent Trends and International Comparisons in Child and Youth Well-Being*

Kenneth C. Land, Duke UniversityPresidential Address

NCSA Annual Meeting, Durham, NCFebruary 29, 2008

Other members of the Foundation for Child Development Child and Youth Well-Being Research Project Team are Vicki L. Lamb (NCCU and Duke U), Sarah O. Meadows (Princeton U), and Hui Zheng (Duke U)

2

FCD CWIFCD CWI

The Basic Social Indicators Questions

• How are we doing?

• More specifically, with respect to children, how are the kids (including adolescents and youths) doing?

3

FCD CWIFCD CWI

The Basic Social Indicators Questions

These questions can be addressed by comparisons: to past historical values, to other contemporaneous units (e.g., comparisons

among subpopulations, states, regions, countries), or

to goals or other externally established standards,

The Child and Youth Well-Being Index (CWI), described below, uses all three points of comparison.

4

FCD CWIFCD CWI

Three Basic References on the CWI

• Land, Kenneth C. Vicki L. Lamb, and Sarah Kahler Mustillo 2001 “Child and Youth Well-Being in the United States, 1975-1998: Some Findings from a New Index,” Social Indicators Research, 56, (December):241-320.

• Land, Kenneth C. Vicki L. Lamb, Sarah O. Meadows, and Ashley Taylor 2007 “Measuring Trends in Child Well-Being: An Evidence-Based Approach,” Social Indicators Research, 80:105-132.

• Hagerty, Michael R. and Kenneth C. Land 2007 “Constructing Summary Indices of Quality of Life: A Model for the Effect of Heterogeneous Importance Weights,” Sociological Methods and Research, 35(May):455-496.

5

FCD CWIFCD CWI

What is the CWI?

• A composite measure of trends over time in the quality of life, or well-being, of America’s children and young people.

• It consists of several interrelated summary or composite indices of annual time series of 28 social indicators of well-being.

6

FCD CWIFCD CWI

The Principal Objective of the CWI:

• To give a sense of the overall direction of change in the well-being of children and youth in the United States as compared to 1975.

7

FCD CWIFCD CWI

The CWI is designed to address the following types of questions:

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

• Did it improve or deteriorate, and by how much?

• In which domains or areas of social life?

8

FCD CWIFCD CWI

• For specific age groups?

• For particular race/ethnic groups?

• For each of the sexes?

• And did race/ethnic group and sex disparities increase or decrease?

9

FCD CWIFCD CWI

Methods of Index Construction

• Annual time series data (from vital statistics and sample surveys) have been assembled on some 28 national level indicators in seven quality-of-life domains.– Family Economic Well-Being

– Health

– Safety/Behavioral Concerns

– Educational Attainment

– Community Connectedness

– Social Relationships (with Family and Peers)

– Emotional/Spiritual Well-Being

10

FCD CWIFCD CWI

• These seven domains have been well-established in over two decades of empirical studies of subjective well-being, including studies of children and youths, by social psychologists and other social scientists.

• In this sense, the CWI is an evidence-based measure of trends in averages of the social conditions encountered by children and youths in the United States.

11

FCD CWIFCD CWI

Table 1. Twenty-Eight Key National Indicators of Child and Youth Well-Being in the United States.

Family Economic Well-Being Domain• Poverty Rate (All Families with Children)• Secure Parental Employment Rate• Median Annual Income (All Families with Children)• Rate of Children with Health InsuranceHealth Domain• Infant Mortality Rate• Low Birth Weight Rate• Mortality Rate (Ages 1-19)• Rate of Children with Very Good or Excellent Health (as reported by parents)• Rate of Children with Activity Limitations (as reported by parents)• Rate of Overweight Children and Adolescents (Ages 6-19)Safety/Behavioral Domain• Teenage Birth Rate (Ages 10-17)• Rate of Violent Crime Victimization (Ages 12-19)• Rate of Violent Crime Offenders (Ages 12-17)• Rate of Cigarette Smoking (Grade 12)• Rate of Alcohol Drinking (Grade 12)• Rate of Illicit Drug Use (Grade 12)

12

FCD CWIFCD CWI

Table 1, ContinuedEducational Attainment Domain• Reading Test Scores (Ages 9, 13, and 17)• Mathematics Test Scores (Ages 9, 13, and 17) Community Connectedness• Rate of Persons who have Received a High School Diploma (Ages 18-24)• Rate of Youths Not Working and Not in School (Ages 16-19)• Rate of Pre-Kindergarten Enrollment (Ages 3-4)• Rate of Persons who have Received a Bachelor’s Degree (Ages 25-29)• Rate of Voting in Presidential Elections (Ages 18-20)Social Relationships Domain• Rate of Children in Families Headed by a Single Parent• Rate of Children who have Moved within the Last Year (Ages 1-18)Emotional/Spiritual Well-Being Domain:• Suicide Rate (Ages 10-19)• Rate of Weekly Religious Attendance (Grade 12)• Percent who report Religion as Being Very Important (Grade 12)Note: Unless otherwise noted, indicators refer to children ages 0-17.

13

FCD CWIFCD CWI

• Each of the 28 Key Indicators is indexed by percentage change from the base year, 1975.– That is, subsequent annual observations are

computed as percentages of the base year.– Three indicators begin in the mid-1980s and use

corresponding base years.

• The base year is assigned a value of 100. – The directions of the indicator values are oriented

such that a value greater (lesser) than 100 in subsequent years means the social condition measured has improved (deteriorated).

14

FCD CWIFCD CWI

• The time series of the 28 indicators are grouped together into the seven domains described above and domain-specific summary well-being indices are constructed.– Within these summary indices, each indicator is

equally weighted.

• The seven component indices are then combined into the equally-weighted composite Child and Youth Well-being Index (CWI).

15

FCD CWIFCD CWI

On Equal Weighting

In an article on statistical methodology, Hagerty and Land (2007) consider the general question of how to construct composite, summary indices for a social unit that will be endorsed by a majority of its members.

They assume that many social indicators are available to describe the social unit, but individuals disagree about the relative weights to be assigned to each social indicator.

The composite index that maximizes agree among individuals can then be derived, along with conditions under which an index will be endorsed by a majority in the social unit.

16

FCD CWIFCD CWI

On Equal Weighting Using both a theoretical analysis of a statistical model to

measure the extent of agreement among individuals and computer simulations, Hagerty and Land (2007) show that intuition greatly underestimates the extent of agreement among individuals, and that it is often possible to construct a composite index with which most individuals agree (at least in direction).

In particular, they show that the equal-weighting strategy is privileged in that it is what statisticians call a minimax estimator—it minimizes disagreement among all possible individuals’ weights for the indicators. Hagery and Land (2007) demonstrate these propositions by calculating real composite quality-of-life indices from two sample surveys of individuals’ actual importance weights.

Some illustrations from Hagerty and Land (2007) follow.

17

FCD CWIFCD CWI

18

FCD CWIFCD CWI

19

FCD CWIFCD CWI

Significant Findings

• The following charts show changes over time in the CWI and its various components.– Overall Composite Index of Child and Youth

Well-Being– Domain-Specific Indices

20

FCD CWIFCD CWI

85

90

95

100

105

110

Pe

rce

nt

of

Ba

se

Ye

ar

Year

Figure 1: Child Well-Being Index, 1975-2005, with Projections for 2006

21

FCD CWIFCD CWI

Figure 2. Domain-Specific Summary Indices, 1975-2005, with Projections for 2006.

60

70

80

90

100

110

120

130

140

150

Year

Perc

en

t o

f B

ase Y

ear

Family Economic Well-Being

Health

Safety/Behavioral Concerns

Educational Attainment

Community Connectedness

Social Relationships

Emotional/Spiritual Well-Being

22

FCD CWIFCD CWI

Figure 1 shows that improvements in the CWI essentially have slowed and stalled in the early years of this first decade of the 21st century.

Just as the CWI allowed us be the first to signal that the steady increases in numerous Key Indicators in the period 1994-2002 were indicative not just of isolated trends, but rather of an overall improvement in well-being, the CWI now is telling us that this trend of overall improvement has come to an end.

23

FCD CWIFCD CWI

Index Validation

Do changes in the CWI actually indicate anything about trends in the subjective well-being of America’s children/youths?

To address this question, Land, Lamb, Meadows, and Taylor (2007) compared trends in the CWI with those of smoothed data on overall life satisfaction for High School Senior from the Monitoring the Future Study, as shown in the following chart:

24

FCD CWIFCD CWI

Figure 6. CWI and Smoothed MTF Life Satisfaction, 1975-2003.

80

85

90

95

100

105

110

Year

44.5

45

45.5

46

46.5

47

47.5

48

48.5

49

Child Well-Being Composite Index

Monitoring the Future Life SatisfactionResponses- Moving Average

Child

Wel

l-Bei

ng C

ompo

site

Inde

x

Mon

itorin

g th

e Fu

tire

Life

Sat

isfa

ctio

n Re

spon

ses-

Mov

ing

Aver

age

25

FCD CWIFCD CWI

Race/Ethnic Disparities and Trends

In brief, the CWI tells us something about trends in child and youth well-being in the U.S. across the past three decades and appears to have some validity. But are the trends concentrated only among the white majority kids and not shared in general?

26

FCD CWIFCD CWI

Race/Ethnic Disparities and Trends

This question recently was addressed in detail by Don Hernandez and Suzanne Macartney in:

• Hernandez, Donald J. and Suzanne E. Macartney, “Child Well-Being 1985-2004: Black-White and Hispanic-White Gaps Narrowing, But Persist.” Presentation, January 29, 2008. Washington, DC: New America Foundation.

Space and time limitations do not permit a complete review of their findings, but the following three slides show the rationale and some overall results.

27

FCD CWIFCD CWI

0.0

25.0

50.0

75.0

100.0

1980 1990 2000 2010 2020 2030 2040 2050 2060 2070 2080 2090 2100

American Indian (2000-2050 with NHOPI, Other)Asian/NHOPI (2000-2050 Asian alone, 2060-2100 includes American Indian)Black, Non-Hispanic Hispanic originWhite, Non-Hispanic

Slide 2. Percent of U.S. Children Ages 0-17 in Specified Race/Ethnic Groups, 1980-2100

Projections for 2000-2050 were released by the Census Bureau March 18, 2004. These projections take into account the much larger Hispanic population identified in Census 2000. Projections and estimates for other years are from an earlier series released by the Census Bureau January 13, 2000, and were based on the count of Hispanics in Census 1990.

28

FCD CWIFCD CWI

Overall CWI, 1985-2004

70

80

90

100

110

120

White

Hispanic

Black

Inde

x va

lue

Average of 7 Domains

"Measuring Social Disparities" (2008) by Donald J. Hernandez and Suzanne Macartney; University at Albany, SUNY.

29

FCD CWIFCD CWI

Black-White gap

● 18 years, based on 1993-2004 trends

● 54 years, based on 1985-2004 trends

Hispanic-White gap

● 14 years, based on 1994-2004 trends

● 43 years, based on 1985-2004 trends

Closing the Gaps:

How long might it take?

30

FCD CWIFCD CWI

International Comparisons of Child and Youth Well-Being

But the question remains:

How well are America’s children and youth doing in recent years as compared to the children and youth of other nations?

31

FCD CWIFCD CWI

International Comparisons of Child and Youth Well-Being

• To address this question, we compare U.S. trends in child and youth well-being with those of four other English-speaking counties, specifically: – Australia, – Canada, – New Zealand, and – the United Kingdom.

32

FCD CWIFCD CWI

These nations were chosen for a number of reasons:• all share a common language;• Australia, Canada, New Zealand, and the U.S. are

former colonies of the United Kingdom;• all five nations are liberal democracies that have

representative democratic forms of government;• all five also place considerable emphasis on the use of

economic markets for the production and distribution of goods and services; and

• because of all the above, all share some common elements of culture.

33

FCD CWIFCD CWI

• We assembled data on 19 Key International Indicators of child and youth well-being that were measured around the year 2000.

• The 19 Key International Indicators can be classified into 7 domains: Family Economic Well-Being, Social Relationships, Health, Safety and Behavior, Educational Attainment, Community Connectedness, and Emotional Well-Being.

34

FCD CWIFCD CWI

• Table 1 presents a “report card” comparison of child and youth well-being by domain for the United States and the four English-speaking countries. The table indicates the measures used within each domain.

• The B [W] indicates the rates for the U.S. are better [worse] than for the comparison country. An = means the rates are equal.

• A blank cell indicates no country-level Key International Indicator was available.

35

FCD CWIFCD CWI

Table 1. Comparison of Child and Youth Well-Being in US and Four English-Speaking Countries: Canada, UK, Australia and New Zealand.

Countries Domains

Canada

United

Kingdom

Australia

New

Zealand 1/2 1/2 1/2 1/2 W W W W

Family Economic Well-Being Poverty Rate: All Children (Age 0-17) Percentage of Working Age Households with

Children Without An Employed Parent B B B B

0/1 1/1 0/1 Social Relationships Percent of All Children Ages 0-17 Living in Single

Mother Families W B W

0/5 1/5 0/4 0/3 W = W W W W W W W W W = W W W

Health Low Birth Weight Infant Mortality Child and Youth Mortality (Age 1-19) Overweight (Age 13 and 15) Self rated "poor or fair health" (Age 11, 13, & 15 ) W B

3/4 3/4 0/1 0/1 W W W W B B B B

Safety/Behavioral Concerns

Teenage Birth Rate (Age 15-19) Smoking Daily (Age 11, 13, & 15) Drunk Twice or More (Age 11, 13, & 15) Having Used Cannabis (Age 15) B B

0/2 0/2 0/2 0/2 W W W W

Educational Attainment Reading (Age 15) Math (Age 15) W W W W

2/4 3/4 3/4 2/3 = B B B = B W B B B B

Educational Attainment/Community Connectedness High School Completion (Age 25-34) Not Working or In School (Age 15-19) Bachelor’s Degree (Age 25-34) Preschool Enrollment Rate (Age 3-4) B W B W

1/1 0/1 1/1 1/1 Emotional Well Being Suicide Rate (Age 15-24) B W B B

Overall Tally 7/19 9/19 5/15 4/12

36

FCD CWIFCD CWI

• Table 2 presents a different perspective on these international comparisons:– a summary of the relative ranking of the

five Anglophone countries based on each of the seven child and youth well-being domains and indicators.

37

FCD CWIFCD CWI

• The domain-specific rankings are based on the averages of the rankings of the indicators within each domain.

• They range from ‘1’, the highest ranking of child well-being, down to ‘5’ (or ‘4’ for social relationships), which indicates the lowest ranking among the five countries.

• Two composite rankings for each country are given at the bottom of the table – – the average rank across the 7 domains, and – the average rank across all 19 indicators.

38

FCD CWIFCD CWI

Table 2. Relative Ranking of Five English-Speaking Countries for Child and Youth Well-Being by Each Domain and Across All Domains and All Indicators

Domain

Canada Australia UnitedStates

New Zealand

UnitedKingdom

Family Economic Well-Being 1 2 2 2 5

Social Relationships 1 2 3 4

Health 1 3 5 4 2

Safety/Behavioral Concerns 3 1 2 5 4

Educational Attainment 1 3 5 1 4

Educational Attainment/Community Connectedness

2 3 1 3 5

Emotional Well-Being 3 4 2 5 1

Average Rank Across All 7 Domains 1.7 2.6 2.9 3.3 3.6

Average Rank Across All 19 Indicators 2.0 2.6 2.9 3.0 3.1

39

FCD CWIFCD CWI

Trends and Implication of Our International Comparisons:

Overall Conclusions:Although no country outscores the United States on

all domains of child and youth well-being, our comparison shows deficiencies in U.S. child and youth well-being, particularly in Family Economic Well-Being, Health, and Educational Attainment domains.

On the other hand, the U.S. does relatively well on indicators in the Safety/Behavior, Community Connectedness, and Emotional Well-Being domains.

40

FCD CWIFCD CWI

Best-Practice AnalysesAs noted earlier, the basic social indicators

question of how are we doing can be answered in three ways.

In addition to comparisons with how we are doing in comparison to the past and in comparison to other countries or social units, we can address this question in terms of comparisons to goals or other externally-constructed standards.

41

FCD CWIFCD CWI

Best-Practice Analyses

An externally-constructed standard that we have used for the CWI is a “best practice” standard, where best practice refers either to the

“best historical value” of each indicator ever observed in the U.S., or to the

“best internationally observed” value of each indicator among countries for which data are available.

42

FCD CWIFCD CWI

Best-Practice Analyses

Recent calculations of the values the CWI would obtain if all of the 28 indicators were at their best values are:

126.53 for historical best-practice U.S. values, and

147.36 for international best-practice values.

43

FCD CWIFCD CWI

Conclusions

• Trends: The composite CWI indicates improvement in overall child and youth well-being since the early-1990s to levels just above those of a generation ago. Race/ethnic disparities have decreased somewhat, but persist.

• International Comparisons: Among English-speaking nations, child well-being in the U.S. appears to be in the middle of the pack—ahead of New Zealand and the U.K. but behind Canada and Australia.

• Best-Practice Analyses: These show that we could be doing substantially better.

44

FCD CWIFCD CWI

The CWI on the Web:

http://www.soc.duke.edu/~cwi/

This project is funded by grants from the Foundation for Child Development