National Academies Press: OpenBook

Databases for Estimating Health Insurance Coverage for Children: A Workshop Summary (2010)

Chapter: 10 Income and Poverty Measurement in Surveys of Health Insurance Coverage--John L. Czajka

« Previous: 9 Health Insurance Coverage in the American Community Survey: A Comparison to Two Other Federal Surveys--Joanna Turner and Michel Boudreaux
Suggested Citation:"10 Income and Poverty Measurement in Surveys of Health Insurance Coverage--John L. Czajka." National Research Council. 2010. Databases for Estimating Health Insurance Coverage for Children: A Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/13024.
×

10
Income and Poverty Measurement in Surveys of Health Insurance Coverage

John L. Czajka

Mathematica Policy Research, Inc.


Each year the Census Bureau produces a set of state-level estimates of children under age 19 in families with incomes at or below 200 percent of the federal poverty level and the subset of these who are without health insurance coverage. The estimates, which are derived from the Annual Social and Economic Supplement (ASEC) to the monthly Current Population Survey (CPS), were mandated by Congress for the purpose of allocating federal funds for matching state expenditures under the Children’s Health Insurance Program (CHIP), but they have become a critical source of data for monitoring the states’ progress in reducing the number of low-income uninsured children. Because the CPS ASEC samples for most states are not large enough to support estimates of low-income uninsured children at satisfactory levels of precision, the Census Bureau combines 3 years of data to produce the estimates for each year. Even with 3 years of data, however, the estimates for many states are still not as precise as might be desired, and the need to combine 3 years of data makes the estimates less sensitive to year-to-year change in the number of low-income uninsured children. While this reduced sensitivity may have merits for funding allocation, it limits the usefulness of the estimates for monitoring state levels of health insurance coverage among low-income children.

With the recent reauthorization of CHIP, Congress made provision for potential enhancements to the annual estimates of low-income uninsured children while simultaneously ending their role in CHIP funding allocation. One of the options that the law mandates the Census Bureau to consider is to supplement or replace the CPS estimates with alternative

Suggested Citation:"10 Income and Poverty Measurement in Surveys of Health Insurance Coverage--John L. Czajka." National Research Council. 2010. Databases for Estimating Health Insurance Coverage for Children: A Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/13024.
×

estimates drawn from the American Community Survey (ACS)—a relatively new Census Bureau survey with an annual sample that is 30 times that of the CPS ASEC supplement. With its ability to support not only state, but also substate estimates, the ACS would appear particularly well suited to serve as the principal source of data for monitoring the number of low-income uninsured children. For this reason, measures of health insurance coverage were added to the ACS questionnaire in 2008.

Although much attention is being focused on the quality of these new estimates of health insurance coverage as a major factor in the choice between the CPS or the ACS, the quality of the income data collected in the ACS may be as important in determining its ultimate viability as a source of annual estimates of low-income children and low-income uninsured children. Arguably, the measurement of income is more challenging than the measurement of health insurance coverage, which requires only binary responses as opposed to dollar amounts. For income measurement, the CPS holds clear advantages over the ACS. For one, the CPS is the official source of annual estimates of income and poverty in the United States, whereas the ACS collects comparatively little information on personal income and does so through a questionnaire that two-thirds of the respondents complete without the assistance of an interviewer and submit by mail. Given these potential limitations, do the ACS income data measure up to the CPS sufficiently well to warrant their use in producing the mandated annual estimates of low-income children and low-income uninsured children?

Policy analysts also have reason to be interested in the ACS as a potential data source for a wide variety of analyses of state and local variation in health insurance coverage, and income is likely to play a key role in many such analyses. For such applications, the demands placed on the ACS income data are likely to exceed those that must be met in providing satisfactory estimates of low-income children.

This paper compares estimates of income obtained from the ACS and the CPS. Selected comparisons include a third survey as well—the National Health Interview Survey (NHIS), which is sponsored and designed by the National Center for Health Statistics (NCHS) with field-work conducted by the Census Bureau. In certain respects the design of the NHIS resembles that of the ACS, with interviews conducted on a rolling basis throughout the calendar year. The measurement of health insurance coverage is a major element of the NHIS, but the measurement of income is much more limited than even that in the ACS. Two other surveys, the Survey of Income and Program Participation (SIPP) and the Household Component of the Medical Expenditure Panel Survey (MEPS) also collect data on both income and health insurance coverage, but their longitudinal designs are less well suited to monitoring the number of low-

Suggested Citation:"10 Income and Poverty Measurement in Surveys of Health Insurance Coverage--John L. Czajka." National Research Council. 2010. Databases for Estimating Health Insurance Coverage for Children: A Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/13024.
×

income uninsured children than those of the CPS, the ACS, or the NHIS, and, like the NHIS, their samples are not designed for state-level estimation. For this reason, this paper does not include comparative estimates from these other surveys.

Following an overview of income measurement and other features that differentiate among the three surveys, I present findings from two empirical analyses comparing the surveys. The first set of findings is drawn from a recent, comprehensive, and systematic assessment of the income data in eight major surveys and their utility for policy-related analyses. This research was conducted by Mathematica Policy Research, Inc., and its subcontractor, Denmead Services & Consulting, under a contract with the Office of the Assistant Secretary for Planning and Evaluation in the U.S. Department of Health and Human Services.1 The second set of findings, which is restricted to the CPS and the ACS, was prepared to examine specific issues related to the use of income data in the analysis of health insurance coverage.2 The paper closes with a brief assessment of the implications of these findings for applications of income and health insurance data from the three surveys.

INCOME MEASUREMENT IN HOUSEHOLD SURVEYS

Despite its wide use in analysis of social and economic phenomena and its critical role in policy analysis, income is difficult to measure well in household surveys. Surveys with a major focus on income devote hundreds of questions to its measurement, whereas surveys that emphasize other topics may devote just a few questions to capturing income. Ideally, those who design the income modules of surveys would be able to draw on established practice to determine how to obtain with a given number of questions the best all-around measure of income or the best measure focusing on a specific application. Such is not the case, however. The most rigorous study of income measurement conducted in the United States to date, by the Income Survey Development Program, was focused on the collection of comprehensive data on income and laid the cornerstone for the design of the SIPP, but it did not shed much light on how a survey could most effectively get by with less. Indeed, the Mathematica study suggests that the SIPP approach, with its focus on monthly income, may be less effective than simpler approaches to measuring, say, annual earnings or even total annual income. Below I provide a brief overview of income measurement in the CPS, the ACS, and the NHIS and then discuss

1

Czajka and Denmead (2008).

2

Some of these findings were presented at the annual meeting of the American Association for Public Opinion Research, May 13-16, 2010, Chicago.

Suggested Citation:"10 Income and Poverty Measurement in Surveys of Health Insurance Coverage--John L. Czajka." National Research Council. 2010. Databases for Estimating Health Insurance Coverage for Children: A Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/13024.
×

other differences among the surveys that have implications for the respective survey estimates of income and poverty.

DIFFERENCES IN THE MEASUREMENT OF INCOME ACROSS SURVEYS

As the official source of annual income and poverty estimates for the United States, the CPS in a sense defines how these constructs should be measured. Thus one can refer to the CPS income concept (pretax money income as defined in the survey), the CPS family (two or more persons living in the same household and related by blood, marriage, or adoption), and CPS residence rules (which include in a household any usual members who are temporarily living elsewhere). The CPS collects data on the presence of more than 50 sources of income and captures up to 24 annual dollar amounts for each sample member age 15 and older. The reported incomes of individual family members at the time of the interview are summed to obtain a measure of total family income for the preceding calendar year.

By contrast, the ACS collects income for up to eight sources for each sample person age 15 and older, combining many sources for which the CPS collects separate reports. The eight sources listed in the questionnaire are

  1. wages, salary, commissions, bonuses, or tips from all jobs (prior to deductions);

  2. self-employment income from own nonfarm businesses or farm businesses, including proprietorships or partnerships;

  3. Interest, dividends, net rental income, royalty income, or income from estates and trusts;

  4. Social Security or Railroad Retirement;

  5. Supplemental Security Income;

  6. any public assistance or welfare payments from the state or local welfare office;

  7. retirement, survivor, or disability pensions; and

  8. any other sources of income received regularly such as veterans’ payments, unemployment compensation, child support, or alimony (not to include lump-sum payments).

This last source serves as a catchall for individual sources that are not explicitly mentioned in the ACS but are identified explicitly in the CPS questionnaire. In addition, each of the preceding sources is captured by multiple questions in the CPS.

CPS interviews are conducted in person or by telephone using

Suggested Citation:"10 Income and Poverty Measurement in Surveys of Health Insurance Coverage--John L. Czajka." National Research Council. 2010. Databases for Estimating Health Insurance Coverage for Children: A Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/13024.
×

computer-assisted personal interviewing (CAPI) or computer-assisted telephone interviewing (CATI). By contrast, about two-thirds of the ACS responses are collected by mail, using a mailout-mailback questionnaire. Most of the remaining responses are collected by CAPI, with the remainder being collected by CATI. The CATI/CAPI responses are collected from a sample of nonrespondents to the mail questionnaire. Unlike the CPS, responding to the ACS is mandatory by law, and the Census Bureau obtains an overall response rate of about 97 percent (weighted to reflect the sampling of nonrespondents to the mail questionnaire). The response rate for the monthly labor force component of the CPS reaches the low 90s, but an additional 8 percent or more of the respondents to the labor force questionnaire do not complete the ASEC supplement interview.

The NHIS collects total family income from a single question asked of the family respondent. Earnings from employment are collected from all persons aged 18 and older, but this is separate from and not reconciled with reported family income. The NHIS interviews are conducted in person using CAPI, and they make extensive use of flash cards—but not for income.

Both the CPS and the NHIS ask their respondents to report their income for the previous calendar year. In the ACS, respondents are asked to report their income for the past 12 months. For persons completing the ACS at the beginning of the year, the past 12 months are January through December of the prior year; for persons completing the survey at the end of the year, the past 12 months are December of the prior year through November of the current year. Thus, the income data collected from ACS households during a given calendar year span a 23-month period centered on December of the prior year.3 To convert the ACS income to a common reference period—specifically, the calendar year in which the data were collected—the Census Bureau applies an inflation adjustment, defined as the ratio of the average monthly price index for the survey year to the average index for the reference period. These monthly adjustment factors are used internally and are applied to published estimates, but the ACS public-use file contains only an average adjustment factor for the 12 survey months, because the Census Bureau has elected not to reveal the survey month. To calculate income relative to poverty, the Census Bureau adjusts the poverty thresholds rather than the reported income. That is, the income reported for a given reference period is divided by the average monthly

3

Income reported in ACS published data and online tables that are based on internal files is adjusted across the rolling reference period to the same real dollars, based on the consumer price index. Income in ACS public-use files is not adjusted for inflation, although an average inflation factor is provided.

Suggested Citation:"10 Income and Poverty Measurement in Surveys of Health Insurance Coverage--John L. Czajka." National Research Council. 2010. Databases for Estimating Health Insurance Coverage for Children: A Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/13024.
×

threshold for that reference period. The estimates of income relative to poverty derived in this manner are reported on the public-use file.

OTHER DIFFERENCES AMONG THE SURVEYS

In addition to these differences in the measurement of income, there are several additional differences that users need to take into account.

First, the surveys represent populations at different times. The CPS ASEC is conducted primarily in March, with additional interviews in February and April. The survey is weighted to represent the population as of March 1.4 Both the ACS and the NHIS represent an average of populations over a calendar year. The ACS is weighted to July 1 population totals, and the four segments of the NHIS are weighted, separately, to February 1, May 1, August 1, and November 1 of the survey year, then combined to create a single annual weight on the public-use file, with an effective reference date of mid-June.

Second, the CPS includes in the household any college students living away from home, whereas the NHIS and the ACS do not. The NHIS samples and interviews students independently and conducts interviews in dormitories to collect data from students in campus housing, which the CPS does not do. The ACS follows decennial census practice in counting college students where they live at the time of the interview, although it did not begin to collect data from residents of dormitories and other group quarters until 2005.5

Third, both the CPS and the ACS define the family unit to include only those persons residing in the same housing unit who are related by blood, marriage, or adoption. The NHIS family definition is more expansive, including unmarried partners (and any relatives living with them) and foster children in the same family. Differences in the residence rules and the family definition affect the number and composition of CPS, ACS, and NHIS families. This, in turn, has implications for the measurement of poverty, which is defined at the family level.

Fourth, while employing the same family definition as the CPS, the ACS does not collect data on relationships among persons who are unrelated to the householder. Thus, if a household includes the householder

4

CPS ASEC interviews are conducted in February, March, and April in the week that includes the 19th. Historically, the annual income supplement was conducted solely in March, so a March-based weight was appropriate. The population estimates that are used as controls in the CPS and other surveys have a reference data corresponding to the first of each month as key components of these population estimates—births and deaths—are collected by calendar month.

5

Until 2005, students living in dormitories were excluded from the ACS universe, and the population totals used in weighting the survey were adjusted accordingly.

Suggested Citation:"10 Income and Poverty Measurement in Surveys of Health Insurance Coverage--John L. Czajka." National Research Council. 2010. Databases for Estimating Health Insurance Coverage for Children: A Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/13024.
×

and a husband, wife, and child who are all unrelated to the householder, the ACS would report this as a householder and three unrelated individuals, and none of the relationships among the three family members would be recorded.

SURVEY ESTIMATES OF INCOME AND POVERTY, 2002

The estimates reported in this section, for calendar year 2002, were adjusted in order to make the estimates as comparable as possible. Under the heading Adjustments to Enhance Comparability, the adjustments that were applied to the surveys included in this paper are described. Total Income compares survey estimates of total income, both in the aggregate and by quintile of family income. The next section examines earned and unearned income aggregates and quintiles. Under Poverty, survey estimates of the number of persons and the fraction of the population classified as poor and near-poor among the population as a whole and among children, nonelderly adults, and the elderly are compared. It also provides estimates of the impact of the NHIS family definition on measured poverty. Finally, Aspects of Data Quality examines alternative measures of income data quality, including allocation of missing data due to item nonresponse and the prevalence of rounding.

Adjustments to Enhance Comparability

The adjustments that were applied to the eight surveys covered by Czajka and Denmead (2008) were designed to make the estimates comparable in terms of income reference period, universe, income concept, and family concept. Most of the adjustments addressed features of the surveys other than the three covered here. I mention adjustments that affected the CPS, the ACS, and the NHIS. ACS income data were multiplied by the average adjustment factor described above. Because of differential inclusion and weighting of members of the active-duty armed forces, all families including such members were removed. Unrelated children under age 15 were also removed, as they are excluded from official estimates of poverty. The most complex adjustment involved the creation of CPS families within the subset of NHIS families that included members based on the broader NHIS family concept—specifically, unmarried partners and their children as well as foster children. In each case, the family members were reassembled into two or more CPS families, and the income of the original family was apportioned among the new families. I did not attempt to adjust for differences in survey timing, which means that the NHIS estimates will reflect a slightly larger population than the CPS estimates, and both surveys’ estimates will reflect a larger population than the ACS estimates.

Suggested Citation:"10 Income and Poverty Measurement in Surveys of Health Insurance Coverage--John L. Czajka." National Research Council. 2010. Databases for Estimating Health Insurance Coverage for Children: A Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/13024.
×
Total Income

As a summary statistic, total or aggregate income is appealing for its simplicity and its use of all of the income data collected in each survey. However, aggregate income is heavily dependent on the amount of income captured from the upper end of the income distribution, which holds the least interest for policy analysis. In presenting estimates of aggregate income, I include a breakdown by quintile of family income, which makes it possible to compare the surveys with respect to their collection of income from different segments of the income distribution. Table 10-1 presents estimates of aggregate income for the whole population and by quintile of family income for the three surveys.6 It also shows these amounts as a percentage of the corresponding amounts for the CPS. There is no gold standard for estimates of income, nor do I mean to suggest that the CPS estimates are the best. But because the CPS is the official source of household income and poverty statistics for the United States, expressing other survey estimates of income as a percentage of the CPS provides a useful standardization.

Aggregate income ranges from $6.12 trillion in the NHIS to $6.47 trillion in the CPS—a spread of just 5 percent. That is, with a single question the NHIS captures 95 percent as much total income as the CPS; with a simple instrument primarily filled out by respondents rather than a trained interviewer, the ACS captures 98 percent as much total income as the CPS.

In examining income by quintile, one finds that the ACS aggregates lie within a percentage point of the CPS aggregates (both above and below) through the first three quintiles before dropping to 98 and 97 percent, respectively, of the CPS in the fourth and fifth quintiles. The NHIS compares least favorably in the bottom quintile, capturing only 85 percent as much total income as the CPS (and the ACS). This fraction rises to a peak of 98 percent in the fourth quintile before falling back to 94 percent in the top quintile.

Aggregates in the top quintile may be affected by outliers and by differences in survey practice with respect to the topcoding of public-use data. For example, the CPS assigns the means of topcoded values as their respective topcodes, which preserves overall means and totals, but not all surveys do this for all income items. For this reason, the survey aggregates are summed through the bottom four quintiles. Through four quintiles, the ACS captures 99.1 percent as much total income as the CPS, and the NHIS improves only marginally to 95.3 percent.

6

In each survey, each of the five quintiles contains the same number of people (weighted) except when the numbers are affected by heaping at quintile boundaries.

Suggested Citation:"10 Income and Poverty Measurement in Surveys of Health Insurance Coverage--John L. Czajka." National Research Council. 2010. Databases for Estimating Health Insurance Coverage for Children: A Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/13024.
×

TABLE 10-1 Aggregate Income by Quintile of Family Income: Three Surveys

Income Estimate

CPS

ACS

NHIS

 

Billions of Dollars

Aggregate Income, All Persons

6,468.40

6,346.30

6,116.20

Family Income Quintile

 

 

 

Lowest

370.50

368.70

313.70

Second

774.10

778.40

717.70

Third

1,090.20

1,087.40

1,058.40

Fourth

1,446.80

1,415.80

1,420.70

Highest

2,786.70

2,696.00

2,605.80

Sum Through Four Quintiles

3,681.70

3,650.30

3,510.40

 

Percentage of CPS

Aggregate Income, All Persons

100.00

98.10

94.80

Family Income Quintile

 

 

 

Lowest

100.00

99.50

84.70

Second

100.00

100.60

92.70

Third

100.00

99.70

97.10

Fourth

100.00

97.90

98.20

Highest

100.00

96.70

93.50

Sum Through Four Quintiles

100.00

99.10

95.30

SOURCE: Mathematica Policy Research, Inc., from tabulations of calendar year 2002 income from the 2003 CPS ASEC supplement; the 2003 NHIS; and prior 12 months income, inflation-adjusted to calendar year 2002, from the 2002 ACS.

Because the CPS and the NHIS are weighted to population totals almost a year later than the ACS, the ACS is at a comparative disadvantage for the measurement of aggregate income. One can compensate by putting aggregate income on a per capita basis. Table 10-2 compares the three surveys with respect to per capita income by quintile. By this measure, the ACS captures 99.8 percent as much total income as the CPS, and the NHIS, which is weighted to slightly larger population totals than the CPS, declines to 94.2 percent of the CPS total. ACS per capita income equals or exceeds CPS per capita income through the first three quintiles and is within 0.4 percentage points in the fourth quintile and within 1.6 percentage points in the top quintile. Putting income on a per capita basis has an uneven impact on the NHIS estimates, which implies that the weighted NHIS sample is not distributed as uniformly by quintile as the other two surveys. A likely explanation is discussed below.

Suggested Citation:"10 Income and Poverty Measurement in Surveys of Health Insurance Coverage--John L. Czajka." National Research Council. 2010. Databases for Estimating Health Insurance Coverage for Children: A Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/13024.
×

TABLE 10-2 Average Income Per Capita by Quintile of Family Income: Three Surveys

Income Estimate

CPS

ACS

NHIS

 

Billions of Dollars

Aggregate Income, All Persons

22,893.00

22,854.00

21,558.00

Family Income Quintile

 

 

 

Lowest

6,513.00

6,526.00

5,528.00

Second

13,789.00

14,259.00

12,649.00

Third

19,293.00

19,576.00

18,493.00

Fourth

25,604.00

25,496.00

25,151.00

Highest

49,316.00

48,543.00

46,114.00

 

Percentage of CPS

Aggregate Income, All Persons

100.00

99.80

94.20

Family Income Quintile

 

 

 

Lowest

100.00

100.20

84.90

Second

100.00

103.40

91.70

Third

100.00

101.50

95.90

Fourth

100.00

99.60

98.20

Highest

100.00

98.40

93.50

SOURCES: Mathematica Policy Research, Inc., from tabulations of calendar year 2002 income from the 2003 CPS ASEC supplement; the 2003 NHIS; and prior 12 months income, inflation-adjusted to calendar year 2002, from the 2002 ACS.

Earned and Unearned Income

Economists divide income into earned and unearned. Earnings include wages and salaries plus self-employment income. Earnings account for 82.8 percent of total income in the CPS, 82.1 percent in the ACS, and 86.0 percent in the NHIS (see Table 10-3).7

The similarity between the ACS and the CPS hides a difference in the composition of earnings in the two surveys. The ACS captures 4 percent less wage and salary income than the CPS but 19 percent more self-employment income, which raises the ACS earned income to 97.3 percent of the CPS total. The ACS also captures slightly more (2.2 percent) unearned income than the CPS, which contributes to an overall total income that is 98.1 percent of the CPS total. The NHIS does not collect unearned income, but the difference between total income and earned income collected in the NHIS implies unearned income that is 77 percent

7

Recall that total family income and earnings by individual family members are measured independently in the NHIS and not reconciled.

Suggested Citation:"10 Income and Poverty Measurement in Surveys of Health Insurance Coverage--John L. Czajka." National Research Council. 2010. Databases for Estimating Health Insurance Coverage for Children: A Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/13024.
×

TABLE 10-3 Contribution of Earned and Unearned Income to Total Income: Three Surveys

Income Estimate

CPS

ACS

NHIS

 

Billions of Dollars

Total Income

6,468.40

6,346.30

6,118.20

Earned Income

5,354.30

5,207.90

5,261.40

Wages and salaries

5,026.30

4,817.20

n/a

Self-employment

328.00

390.70

n/a

Unearned Income

1,114.10

1,138.30

854.80

 

Percentage of Total Income

Total Income

100.00

100.00

100.00

Earned Income

82.80

82.10

86.0

Wages and salaries

77.70

75.90

n/a

Self-employment

5.10

6.20

n/a

Unearned Income

17.20

17.90

14.0

 

Percentage of CPS Income by Source

Total Income

100.00

98.10

94.60

Earned Income

100.00

97.30

98.30

Wages and salaries

100.00

95.80

n/a

Self-employment

100.00

119.10

n/a

Unearned Income

100.00

102.20

76.70

NOTE: n/a = not available.

SOURCES: Mathematica Policy Research, Inc., from tabulations of calendar year 2002 income from the 2003 CPS ASEC supplement; the 2003 NHIS; and prior 12 months income, inflation-adjusted to calendar year 2002, from the 2002 ACS.

of the CPS total. This implied shortfall is simply an indication that the NHIS does not do as well in obtaining total income with its single question as it does in collecting earned income from all adults.8

Comparing survey estimates of earned income by quintile of family income, one finds, interestingly, that both the ACS and the NHIS have more earnings in the lowest quintile of family income than does the CPS (see Table 10-4). The additional earnings range from 12 to 17 percent of the CPS total. The ACS has progressively less total earnings relative to the CPS as the quintile increases. The NHIS, in contrast, has progressively more aggregate earnings relative to the CPS over quintiles two through four.

8

In the 2002 NHIS, more than a fifth of individuals were in families with more reported total earnings than total family income. Each of these families would have implied negative unearned income.

Suggested Citation:"10 Income and Poverty Measurement in Surveys of Health Insurance Coverage--John L. Czajka." National Research Council. 2010. Databases for Estimating Health Insurance Coverage for Children: A Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/13024.
×

TABLE 10-4 Aggregate Earned Income by Quintile of Family Income: Three Surveys

Income Estimate

CPS

ACS

NHIS

 

Billions of Dollars

Aggregate Earned Income

5,354.30

5,207.90

5,261.40

Family Income Quintile

 

 

 

Lowest

176.10

206.50

196.40

Second

542.90

565.30

514.40

Third

889.20

878.80

888.20

Fourth

1,255.90

1,225.50

1,301.90

Highest

2,490.20

2,332.00

2,360.50

Sum Through Four Quintiles

2,864.10

2,876.00

2,900.90

 

Percentage of CPS

Aggregate Income, All Persons

100.00

97.30

98.30

Family Income Quintile

 

 

 

Lowest

100.00

117.30

111.60

Second

100.00

104.10

94.70

Third

100.00

98.80

99.90

Fourth

100.00

97.60

103.70

Highest

100.00

93.60

94.80

Sum Through Four Quintiles

100.00

100.40

101.30

SOURCES: Mathematica Policy Research, Inc., from tabulations of calendar year 2002 income from the 2003 CPS ASEC supplement; the 2003 NHIS; and prior 12 months income, inflation-adjusted to calendar year 2002, from the 2002 ACS.

Unearned income does not show such clear patterns. Overall, the ACS has slightly more unearned income than the CPS, but, unlike earned income, the ACS has progressively more than the CPS as the quintile rises (see Table 10-5). In the top quintile, the ACS has 23 percent more unearned income than the CPS. With its unearned income measured as a residual rather than a reported amount, the NHIS is erratic. The difference between aggregate total and aggregate earned income is as low as 60 percent of the CPS aggregate in one quintile and as high as 88 percent (in the adjacent quintile).

Moving from dollars to receipt, which the NHIS collects for a number of income sources, I observe striking differences among the surveys with respect to programs that serve low-income families. Both the ACS and the NHIS find a much greater incidence of food stamp and welfare receipt among higher income families than does the CPS (see Table 10-6). While the ACS identifies 9 to 10 percent more participants than the CPS in the bottom two quintiles, this proportion rises sharply through the top three quintiles. In the top quintile the ACS finds nearly four times as

Suggested Citation:"10 Income and Poverty Measurement in Surveys of Health Insurance Coverage--John L. Czajka." National Research Council. 2010. Databases for Estimating Health Insurance Coverage for Children: A Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/13024.
×

TABLE 10-5 Aggregate Unearned Income by Quintile of Family Income: Three Surveys

Income Estimate

CPS

ACS

NHIS

 

Billions of Dollars

Aggregate Unearned Income

1,114.10

1,138.30

854.80

Family Income Quintile

 

 

 

Lowest

194.40

162.20

117.30

Second

231.20

213.10

203.30

Third

201.00

208.60

170.20

Fourth

190.90

190.30

118.70

Highest

296.50

364.00

245.30

Sum Through Four Quintiles

817.60

774.30

609.50

 

Percentage of CPS

Aggregate Income, All Persons

100.00

102.20

76.70

Family Income Quintile

 

 

 

Lowest

100.00

83.40

60.30

Second

100.00

92.20

88.00

Third

100.00

103.80

84.60

Fourth

100.00

99.70

62.20

Highest

100.00

122.80

82.70

Sum Through Four Quintiles

100.00

94.70

74.50

NOTE: Unearned income is the difference between total income, reported in Table 10-1, and earned income, reported in Table 10-4.

SOURCES: Mathematica Policy Research, Inc., from tabulations of calendar year 2002 income from the 2003 CPS ASEC supplement; the 2003 NHIS; and prior 12 months income, inflation-adjusted to calendar year 2002, from the 2002 ACS.

many participants as the CPS (a difference of 700,000 persons). The NHIS is not as extreme as the ACS, but it still has nearly twice the incidence of food stamp and welfare receipt in the top quintile as does the CPS. In the bottom quintile the NHIS compares closely to the ACS. Given the target populations for these two programs, one suspects that the substantially higher receipt of food stamps and welfare in the top quintiles in the ACS and the NHIS relative to the CPS is more indicative of problems with the former two surveys than the CPS. While all three surveys show the expected steep decline in receipt with rising family income, it is likely that both the ACS and the NHIS have far too many high-income families with food stamps or welfare. Without further study, however, it is not clear that the problem necessarily lies with the survey data on receipt as opposed to family income.

Suggested Citation:"10 Income and Poverty Measurement in Surveys of Health Insurance Coverage--John L. Czajka." National Research Council. 2010. Databases for Estimating Health Insurance Coverage for Children: A Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/13024.
×

TABLE 10-6 Persons in Families with Welfare and/or Food Stamps by Quintile of Family Income: Three Surveys

Income Estimate

CPS

ACS

NHIS

 

Thousands of Persons

All Participants

20,496.00

24,325.00

21,990.00

Family Income Quintile

 

 

 

Lowest

13,562.00

14,879.00

14,783.00

Second

4,461.00

4,854.00

4,355.00

Third

1,748.00

2,396.00

1,685.00

Fourth

493.00

1,273.00

719.00

Highest

233.00

923.00

447.00

Sum Through Four Quintiles

20,263.00

23,402.00

21,543.00

 

Percentage of CPS

Aggregate Income, All Persons

100.00

118.70

107.30

Family Income Quintile

 

 

 

Lowest

100.00

109.70

109.00

Second

100.00

108.80

97.60

Third

100.00

137.00

96.40

Fourth

100.00

258.40

145.90

Highest

100.00

396.80

192.20

Sum Through Four Quintiles

100.00

115.50

106.30

SOURCE: Mathematica Policy Research, Inc., from tabulations of the 2003 CPS ASEC and the 2003 NHIS.

Poverty

Another useful summary statistic, but one that is informative about only the lower end of the income distribution, is the poverty rate—that is, the percentage of persons whose family incomes lie below the official poverty threshold. Estimates of the number of poor and near-poor (defined as those between 100 and 200 percent of the poverty threshold) are important measures for policy analysis.9 Marked differences across surveys in estimates of the poor and near-poor would be a source of concern among policy analysts and other data users. They would imply, for example, that the poor in one survey do not represent the same people as the poor in another survey.

Estimates of the Poor and Near-Poor

The CPS and the ACS compare closely, with poverty rates of 12.2 percent and 12.5 percent, respectively (see Table 10-7). The NHIS, in contrast,

9

Near-poor does not have a standard definition. I use the term to give a name to those with low income (below 200 percent of poverty) but not poor. Elsewhere, near-poor is sometimes used to identify persons between 100 and 125 percent of poverty.

Suggested Citation:"10 Income and Poverty Measurement in Surveys of Health Insurance Coverage--John L. Czajka." National Research Council. 2010. Databases for Estimating Health Insurance Coverage for Children: A Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/13024.
×

TABLE 10-7 Estimates of the Poor and Near-Poor: Three Surveys

Estimate

CPS

ACS

NHIS

 

Millions of Persons

All Persons

282.55

277.69

283.71

Poverty Status

 

 

 

Poor

34.38

34.61

41.58

Near-poor

51.81

49.28

53.91

Total Low Income

86.19

83.89

95.49

 

Percentage of the Population

All Persons

100.00

100.00

100.00

Poverty Status

 

 

 

Poor

12.20

12.50

14.70

Near-poor

18.30

17.70

19.00

Total Low Income

30.50

30.20

33.70

NOTES: The poor have a family income below the poverty threshold. The near-poor have a family income at or above the poverty threshold but below twice the poverty threshold.

SOURCES: Mathematica Policy Research, Inc., from tabulations of poverty status in calendar year 2002 from the 2003 CPS ASEC supplement and the 2003 NHIS, and poverty status in the prior 12 months, inflation-adjusted to calendar year 2002, from the 2002 ACS.

is an outlier with an estimate of 41.6 million poor and a poverty rate of 14.7 percent. The NHIS poverty rate with a CPS family definition is more than 2 percentage points higher than those of the other two surveys.

Combining the estimates of the poor and near-poor, which define the low-income population, the ACS assigns a slightly smaller fraction of the population to that status than the CPS: 30.2 versus 30.5 percent. The NHIS has a larger fraction—33.7 percent of the population or 95.5 million persons—to have low income. The number of persons estimated to have low income in the NHIS exceeds the CPS estimate by 9.3 million.

Estimates of Poor and Near-Poor Children, Nonelderly Adults, and Elderly

The CPS and ACS estimates of children in low-income families are very similar at about 27.4 million, or 38.2 to 38.8 percent of the child population (see Table 10-8). The NHIS has somewhat more at 29.7 million or 41.4 percent. The NHIS also has the most poor children, with a child poverty rate that exceeds the other surveys by 2 to 3 percentage points, but it has no more near-poor children than the CPS. In fact, the estimates of near-poor children across the three surveys vary from only 14.9 to 15.4 million or 21.1 to 21.5 percent of the child population.

Suggested Citation:"10 Income and Poverty Measurement in Surveys of Health Insurance Coverage--John L. Czajka." National Research Council. 2010. Databases for Estimating Health Insurance Coverage for Children: A Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/13024.
×

TABLE 10-8 Estimates of Poor and Near-Poor Children: Three Surveys

Estimate

CPS

ACS

NHIS

 

Millions of Persons

All Children Under Age 18

71.67

70.79

71.73

Poverty Status

 

 

 

Poor

12.03

12.51

14.29

Near-poor

15.38

14.94

15.41

Total Low Income

27.41

27.45

29.70

 

Percentage of the Population

All Children Under Age 18

100.00

100.00

100.00

Poverty Status

 

 

 

Poor

16.80

17.70

19.90

Near-poor

21.50

21.10

21.50

Total Low Income

38.20

38.80

41.40

NOTES: The poor have a family income below the poverty threshold. The near-poor have a family income at or above the poverty threshold but below twice the poverty threshold.

SOURCES: Mathematica Policy Research, Inc., from tabulations of poverty status in calendar year 2002 from the 2003 CPS ASEC supplement and the 2003 NHIS, and poverty status in the prior 12 months, inflation-adjusted to calendar year 2002, from the 2002 ACS.

Estimates of nonelderly adults who are poor and near-poor are strikingly similar between the CPS and ACS at 45.6 versus 45.3 million lowincome persons or 25.8 versus 26.1 percent of the nonelderly adult population (see Table 10-9). The NHIS, however, has substantially more at 53 million or nearly 30 percent of the nonelderly adult population with low income.

The ACS finds the fewest low-income elderly at 11.2 million or 33.3 percent, compared with 13.2 million or 38.5 percent for the CPS (see Table 10-10). Estimates of the number of poor elderly do not differ as much among the three surveys, however, with a range of only 3.2 to 3.8 million or 9.5 to 11.0 percent.

Impact of the Family Definition

Some surveys utilize family definitions that deviate from the CPS family concept, which is incorporated into the official measure of poverty in the United States. The NHIS, as noted, includes unmarried partners and their children in the same family, and it includes foster children as part of the family as well. Broadening the family concept relative to the CPS family produces major changes in family income and poverty rates.

In developing the NHIS estimates of income for comparison with the other surveys, I separated unmarried partners and foster children from

Suggested Citation:"10 Income and Poverty Measurement in Surveys of Health Insurance Coverage--John L. Czajka." National Research Council. 2010. Databases for Estimating Health Insurance Coverage for Children: A Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/13024.
×

TABLE 10-9 Estimates of Poor and Near-Poor Nonelderly Adults: Three Surveys

Estimate

CPS

ACS

NHIS

 

Millions of Persons

All Adults Aged 18 to 64

176.66

173.34

177.76

Poverty Status

 

 

 

Poor

18.77

18.91

23.53

Near-poor

26.85

26.36

29.40

Total Low Income

45.62

45.27

52.94

 

Percentage of the Population

All Adults Aged 18 to 64

100.00

100.00

100.00

Poverty Status

 

 

 

Poor

10.60

10.90

13.20

Near-poor

15.20

15.20

16.50

Total Low Income

25.80

26.10

29.80

NOTES: The poor have a family income below the poverty threshold. The near-poor have a family income at or above the poverty threshold but below twice the poverty threshold.

SOURCES: Mathematica Policy Research, Inc., from tabulations of poverty status in calendar year 2002 from the 2003 CPS ASEC supplement and the 2003 NHIS, and poverty status in the prior 12 months, inflation-adjusted to calendar year 2002, from the 2002 ACS.

TABLE 10-10 Estimates of Poor and Near-Poor Elderly: Three Surveys

Population Subgroup

CPS

ACS

NHIS

 

Millions of Persons

All Persons Aged 65 and Older

34.22

33.56

34.22

Poverty Status

 

 

 

Poor

3.58

3.20

3.76

Near-poor

9.58

7.98

9.10

Total Low Income

13.16

11.18

12.86

 

Percentage of the Population

All Persons Aged 65 and Older

100.00

100.00

100.00

Poverty Status

 

 

 

Poor

10.50

9.50

11.00

Near-poor

28.00

23.80

26.50

Total Low Income

38.50

33.30

37.60

NOTES: The poor have a family income below the poverty threshold. The near-poor have a family income at or above the poverty threshold but below twice the poverty threshold.

SOURCES: Mathematica Policy Research, Inc., from tabulations of poverty status in calendar year 2002 from the 2003 CPS ASEC supplement and the 2003 NHIS, and poverty status in the prior 12 months, inflation-adjusted to calendar year 2002, from the 2002 ACS.

Suggested Citation:"10 Income and Poverty Measurement in Surveys of Health Insurance Coverage--John L. Czajka." National Research Council. 2010. Databases for Estimating Health Insurance Coverage for Children: A Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/13024.
×

TABLE 10-11 Comparison of the CPS and NHIS Family Concepts with Respect to the Estimated Distribution of Persons by Income Relative to Poverty

Family Income as Percentage of Poverty

CPS Family

NHIS Family

Change

Total Percentage

100.00

100.00

 

Under 100

14.70

13.70

−0.90

100 to under 200

19.00

19.00

0.00

200 to under 400

30.70

30.90

0.20

400 or more

35.70

36.40

0.80

Total Percentage

283.70

283.90

0.20

Under 100

41.60

39.00

−2.60

100 to under 200

53.90

53.80

−0.10

200 to under 400

87.10

87.70

0.60

400 or more

101.20

103.40

2.30

SOURCE: Mathematica Policy Research, Inc., from tabulations of poverty status in calendar year 2002 from the 2003 NHIS.

the NHIS family and apportioned family income among the two or more family units created from each NHIS family and that conform to the CPS family definition. By comparing the income and poverty estimates prepared using the CPS family definition with estimates obtained from the original data, I assessed the impact of using the NHIS versus CPS family definition on groups and individuals for the purposes of estimating family income.

Table 10-11 shows estimates of the impact of the broader NHIS family definition. The NHIS family definition reduces the number of poor by 2.6 million and reduces the poverty rate by 0.9 percentage points. There is no impact on the percentage of persons between 100 and 200 percent of poverty, which means that the number of people who were moved above the poverty line by the NHIS family concept is offset by the number of people who were moved beyond 200 percent of poverty. Most of the upward shift is observed in the top category—that is, among people above 400 percent of poverty, where the broader family concept adds 2.3 million to the number in the NHIS.

Aspects of Data Quality

Two ways in which respondents can diminish the effectiveness of even very well designed income questions are by providing no answers at all or (which may be worse) by giving inaccurate answers. It is well known that income questions generate some of the highest item non-

Suggested Citation:"10 Income and Poverty Measurement in Surveys of Health Insurance Coverage--John L. Czajka." National Research Council. 2010. Databases for Estimating Health Insurance Coverage for Children: A Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/13024.
×

response rates in surveys generally. Frequently, this results in large amounts of missing income data. Unless the data producers choose to leave such missing data for their users to address, they must apply one or more methods of allocation (or imputation) to fill in the missing data. When the data producers elect to allocate their missing income data, high rates of nonresponse are likely to mean that large fractions of the income data they provide to their users will have been created by the data producers themselves rather than supplied by their respondents. This makes the quality of the income data dependent not only on the completeness and accuracy of the reported amounts, but also on the quality of the methods used to generate allocated amounts. High-quality allocations are unbiased, have the appropriate variances, and reflect the covariance structure that would be observed if the missing values were fully and correctly reported.

One can quantify the amount of income data that are allocated in a survey and, in so doing, measure the magnitude of nonresponse and its potential impact on data quality. One cannot assess in any direct way the accuracy of survey responses to income questions. However, one way in which respondents may reduce the accuracy of their responses is to use a high level of approximation—for example, by reporting a salary of $50,000 when the true salary lies somewhere between $45,000 and $55,000. When a significant number of respondents round their responses in this way, it distorts the distribution by creating spikes at the rounded values. In fact, rounding is a commonly used technique for protecting the confidentiality of income data in public-use files.10 The frequency of rounded responses can be quantified, and I do so for selected income sources for the three surveys.

Nonresponse and Allocation

The percentage of respondents with any income allocated is sensitive to the number of income questions asked in the survey. For this reason, it is more useful to look at the proportion of dollars that was allocated. Table 10-12 reports the fraction of total dollars allocated in the three surveys by source of income. For total income, this fraction in the ACS is about half what it is in the CPS: 17.6 percent versus 34.2 percent. The allocation rate in the NHIS is similar to the CPS at 32.4 percent. For the seven sources of income listed in the table, allocation rates in the CPS vary from a low of 28.0 percent (Supplemental Security Income) to a high of 62.6 percent (asset income). Allocation rates in the ACS show little variation by source,

10

Limiting the number of significant digits in reported incomes reduces their uniqueness, making them less identifiable. The ACS uses a very well-defined rounding rule.

Suggested Citation:"10 Income and Poverty Measurement in Surveys of Health Insurance Coverage--John L. Czajka." National Research Council. 2010. Databases for Estimating Health Insurance Coverage for Children: A Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/13024.
×

TABLE 10-12 Percentage of Income Allocated by Source: Three Surveys

Source of Income

CPS

ACS

NHIS

Total Income (NHIS family income)

34.20

17.60

32.40

Wages and salaries (NHIS earnings)

32.00

17.20

31.80

Self-employment

44.70

23.10

n/a

Asset income

62.60

19.40

n/a

Social Security or Railroad Retirement

35.50

18.50

n/a

Supplemental Security Income

28.00

16.70

n/a

Welfare

29.20

17.90

n/a

Pensions

35.40

16.20

n/a

NOTE: n/a = not available.

SOURCES: Mathematica Policy Research, Inc., from tabulations of calendar year 2002 income from the 2003 CPS ASEC supplement; the 2003 NHIS; and prior 12 months income, inflation-adjusted to calendar year 2002, from the 2002 ACS.

however, with the lowest (16.2 percent for pensions) and the highest (23.1 percent for self-employment income) being separated by only 7 percentage points. Thus, the ACS allocation rate for asset income is less than a third of the corresponding rate in the CPS.

One can speculate that the lower nonresponse to the ACS income questions is a carryover from the mandatory nature of the survey. This may also explain why the ACS allocation rates differ so little by source of income. Regardless of the explanation, the markedly lower allocation rate in the ACS is an important indicator of data quality.

Rounding

One reason to examine rounding in the context of policy analytic use of income data is that the heaping of incomes at well-spaced values can distort the results of policy simulations involving the use of income thresholds to establish program eligibility. An eligibility threshold that lies near an income amount with excessive heaping will produce dramatically different results depending on whether the threshold falls just below or just above that amount. If the former, a simulation will mildly understate the impact of a small change in policy; if the latter, it will grossly overstate the impact of the policy change. Another reason to be concerned about rounding is that a high level of rounding suggests inaccuracy or a lack of precision in the reported amounts generally.

I examined the extent of rounding in reported incomes below $52,500 for earnings, wages and salaries, Social Security benefits, other retirement income, total personal income, and total family income in the three

Suggested Citation:"10 Income and Poverty Measurement in Surveys of Health Insurance Coverage--John L. Czajka." National Research Council. 2010. Databases for Estimating Health Insurance Coverage for Children: A Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/13024.
×

surveys.11 Earnings and total family income are the only income amounts collected in the NHIS and therefore the only amounts on which all three surveys could be compared. I compared the CPS and the ACS with respect to four additional sources. Social Security benefits reported at the person level will have been collected as a single value in both surveys. Most respondents reporting wages and salaries, retirement income, and even earnings are likely to have supplied a single value in response to one question, even though multiple questions were asked.

The results show the differential impact of few versus many income questions. In the NHIS, which relies on a single, person-level question to collect earnings and a single, family-level question to collect total family income, 40 percent of the reported earnings and 36 percent of the reported family incomes below $52,500 are multiples of $5,000, and 23 percent of the earnings and 21 percent of the total family incomes are multiples of $10,000 (see Table 10-13).

On wages and salaries as well as earnings, the CPS is only marginally better than the ACS, with 27 to 28 percent of the amounts being divisible by $5,000 compared with 30 percent for the ACS. The ACS shows markedly more rounding than the CPS on total personal income (20 versus 14 percent) and total family income (16 versus 11 percent). Rounding in both surveys is much lower for Social Security and other retirement income than for the other sources, but rounding in the CPS is still several percentage points lower than in the ACS.

MORE CURRENT COMPARISONS

Comparing just a single year of data provides no basis for assessing how the rolling reference period for annual income data in the ACS may affect the measurement of income and poverty in the ACS relative to the CPS. To provide more data on this issue, I constructed simple comparisons using data from the 2007 and 2008 ACS and the 2008 and 2009 CPS ASEC supplements. In addition, with the 2008 ACS data it is possible to compare the two surveys with respect to their estimates of health insurance coverage by relative income. Below, the first section compares estimates of the poor and near-poor for the total civilian noninstitutionalized population and for children and both nonelderly and elderly adults. The next section compares estimates of the percentage uninsured by income relative to poverty for these same populations. The third section examines the impact of differences between the CPS and the ACS residency rules

11

I selected $52,500 in order to examine the frequency of rounding up to levels of $50,000. The ACS public-use file has incomes of $50,000 or greater rounded to the nearest $1,000.

Suggested Citation:"10 Income and Poverty Measurement in Surveys of Health Insurance Coverage--John L. Czajka." National Research Council. 2010. Databases for Estimating Health Insurance Coverage for Children: A Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/13024.
×

TABLE 10-13 Reporting of Rounded Values by Source of Income by Survey Among Positive Dollar Amounts Below $52,500

Income Source and Level of Rounding

CPS

ACS

NHIS

Earnings

 

 

 

Percentage divisible by $5,000

27.80

29.60

39.80

Percentage divisible by $10,000

15.80

17.40

22.90

Percentage of income in range

82.10

82.40

80.90

Wages and Salaries

 

 

 

Percentage divisible by $5,000

27.20

29.70

n/a

Percentage divisible by $10,000

15.40

17.40

n/a

Percentage of income in range

82.20

82.70

n/a

Social Security

 

 

 

Percentage divisible by $5,000

0.60

4.30

n/a

Percentage divisible by $10,000

0.40

1.90

n/a

Percentage of income in range

100.00

100.00

n/a

Retirement Income

 

 

 

Percentage divisible by $5,000

4.50

8.00

n/a

Percentage divisible by $10,000

2.70

4.30

n/a

Percentage of income in range

95.60

95.40

n/a

Total Personal Income

 

 

 

Percentage divisible by $5,000

13.70

19.70

n/a

Percentage divisible by $10,000

7.80

11.50

n/a

Percentage of income in range

84.60

84.20

n/a

Total Family Income

 

 

 

Percentage divisible by $5,000

11.00

16.20

35.60

Percentage divisible by $10,000

6.20

9.50

20.90

Percentage of income in range

66.90

66.00

60.30

NOTES: Allocated amounts are excluded from each source. Family income for the NHIS is based on the NHIS family, which is the level at which family income was reported.

n/a = not available.

SOURCES: Mathematica Policy Research, Inc., from tabulations of calendar year 2002 income from the 2003 CPS ASEC supplement and the 2003 NHIS and prior 12 months income, inflation-adjusted to calendar year 2002, from the 2002 ACS.

by comparing young adults aged 18 to 24 by college enrollment, relative income, and the percentage lacking health insurance coverage.

Poor and Near-Poor

For 2007 the ACS estimated a slightly higher poverty rate than the CPS, at 13.0 versus 12.5 percent, and a slightly smaller fraction between

Suggested Citation:"10 Income and Poverty Measurement in Surveys of Health Insurance Coverage--John L. Czajka." National Research Council. 2010. Databases for Estimating Health Insurance Coverage for Children: A Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/13024.
×

100 and 200 percent of poverty (near-poor), at 17.7 versus 18.0 percent (see Table 10-14). The total fraction of the population estimated as low income (below 200 percent of poverty) was essentially identical in the two surveys: 30.6 percent in the ACS versus 30.5 percent in the CPS. For 2008, when the United States entered a deep recession in the latter part of the year, the CPS showed an increase of 0.7 percentage points in both the percentage poor and the percentage near-poor, yielding an increase of 1.4 percentage points in the percentage of the population classified as low income. The ACS, with its longer reference period, showed little change between 2007 and 2008, with the poverty rate rising by just 0.2 percentage points and the fraction near-poor remaining essentially unchanged. The fraction of the population classified as low income rose by just 0.3 percentage points in the ACS, leaving the ACS a full percentage point below the CPS. These changes are consistent with the expectation that the ACS should be slower to respond to economic changes.

In effect, the same pattern can be seen among children, with no difference between the two surveys in the percentage with low income in 2007 but the CPS moving more than a full percentage point above the ACS in 2008 (Table 10-15). Among nonelderly adults, the ACS had a percentage point higher fraction classified as low income in 2007 but showed only a 0.4

TABLE 10-14 Estimates of the Poor and Near-Poor: CPS and ACS, 2007 and 2008

Estimate

2007

2008

CPS

ACS

CPS

ACS

 

Millions of Persons

All Persons

297.81

293.42

300.10

295.32

Poverty Status

 

 

 

 

Poor

37.22

38.01

39.74

39.03

Near-poor

53.64

51.84

56.04

52.33

Total Low Income

90.86

89.85

95.77

91.36

 

Percentage of the Population

All Persons

100.00

100.00

100.00

100.00

Poverty Status

 

 

 

 

Poor

12.50

13.00

13.20

13.20

Near-poor

18.00

17.70

18.70

17.70

Total Low Income

30.50

30.60

31.90

30.90

NOTES: The poor have a family income below the poverty threshold. The near-poor have a family income at or above the poverty threshold but below twice the poverty threshold.

SOURCES: Mathematica Policy Research, Inc., from tabulations of poverty status in calendar years 2007 and 2008 from the 2008 and 2009 CPS ASEC supplements and the 2007 and 2008 ACS.

Suggested Citation:"10 Income and Poverty Measurement in Surveys of Health Insurance Coverage--John L. Czajka." National Research Council. 2010. Databases for Estimating Health Insurance Coverage for Children: A Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/13024.
×

TABLE 10-15 Estimates of the Poor and Near-Poor Children Under 18: CPS and ACS, 2007 and 2008

Estimate

2007

2008

CPS

ACS

CPS

ACS

 

Millions of Persons

All Children Under Age 18

73.99

72.66

74.07

72.80

Poverty Status

 

 

 

 

Poor

13.32

13.03

14.05

13.23

Near-poor

15.66

15.47

16.00

15.53

Total Low Income

28.98

28.50

30.05

28.76

 

Percentage of the Population

All Children Under Age 18

100.00

100.00

100.00

100.00

Poverty Status

 

 

 

 

Poor

18.00

17.90

19.0

18.20

Near-poor

21.20

21.30

21.60

21.30

Total Low Income

39.20

39.20

40.60

39.50

NOTES: The poor have a family income below the poverty threshold. The near-poor have a family income at or above the poverty threshold but below twice the poverty threshold.

SOURCES: Mathematica Policy Research, Inc., from tabulations of poverty status in calendar years 2007 and 2008 from the 2008 and 2009 CPS ASEC supplements and the 2007 and 2008 ACS.

percentage point increase in this fraction between 2007 and 2008, while the CPS fraction increased by 1.6 percentage points (see Table 10-16). Among the elderly, for whom the ACS finds substantially fewer near-poor, the ACS had a 4 percentage point lower fraction with low income in 2007 (see Table 10-17). This difference increased by 0.4 percentage points between 2007 and 2008. The more modest increase among the elderly versus the nonelderly is consistent with the elderly receiving more of their income from sources that are less affected by recession than children and working-age adults.

Uninsured by Relative Income

For the civilian noninstitutionalized population as a whole, the uninsured rate measured in the CPS for 2008 was 15.4 percent. Using a very different approach to measuring health insurance coverage, the ACS obtained an uninsured rate of 15.2 percent (see Table 10-18). The other papers in this session address the differences in measurement and possible reasons why the uninsured rates are nevertheless so close. When persons are classified by income relative to poverty, similar uninsured rates are also found in each of the four classes. The ACS has a somewhat lower rate among the poor (29.1 versus 30.5 percent) but a slightly higher rate

Suggested Citation:"10 Income and Poverty Measurement in Surveys of Health Insurance Coverage--John L. Czajka." National Research Council. 2010. Databases for Estimating Health Insurance Coverage for Children: A Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/13024.
×

TABLE 10-16 Estimates of the Poor and Near-Poor Nonelderly Adults: CPS and ACS, 2007 and 2008

Estimate

2007

2008

CPS

ACS

CPS

ACS

 

Millions of Persons

All Adults Aged 18 to 64

187.03

184.53

188.25

185.32

Poverty Status

 

 

 

 

Poor

20.35

21.52

22.03

22.12

Near-poor

28.28

28.23

30.00

28.63

Total Low Income

48.63

49.75

52.03

50.75

 

Percentage of the Population

All Adults Aged 18 to 64

100.00

100.00

100.00

100.00

Poverty Status

 

 

 

 

Poor

10.90

11.70

11.70

11.90

Near-poor

15.10

15.30

15.90

15.50

Total Low Income

26.00

27.0

27.60

27.40

NOTE: The poor have a family income below the poverty threshold. The near-poor have a family income at or above the poverty threshold but below twice the poverty threshold.

SOURCES: Mathematica Policy Research, Inc., from tabulations of poverty status in calendar years 2007 and 2008 from the 2008 and 2009 CPS ASEC supplements and the 2007 and 2008 ACS.

among the near-poor (25.9 versus 24.3 percent). For the combined, lowincome population, the uninsured rates for the two surveys are therefore quite similar: 26.9 percent for the CPS and 27.3 percent for the ACS. The rates are essentially identical among persons between 200 and 400 percent of poverty, whereas the ACS rate is modestly lower among persons above 400 percent of poverty: 5.5 percent versus 6.0 percent in the CPS.

Among children, the ACS uninsured rate is fractionally higher than the CPS at 9.9 versus 9.8 percent (see Table 10-19). Uninsured rates are equally similar among poor children, whereas the ACS uninsured rate is more than a percentage point higher among the near-poor: 15.9 versus 14.7 percent. As with the population as a whole, the ACS is 0.1 percentage point higher among children between 200 and 400 percent of poverty and 0.5 percentage points lower among children above 400 percent of poverty.

The pattern changes for nonelderly adults, for whom the poor in the ACS are somewhat less likely to be uninsured than the poor in the CPS: 41.3 versus 44.0 percent (see Table 10-20). The ACS is a percentage point higher among the near-poor, and the differences among those above 200 percent of poverty are identical to those for children, even though the uninsured rates for nonelderly adults are more than twice as high as they are for children.

Suggested Citation:"10 Income and Poverty Measurement in Surveys of Health Insurance Coverage--John L. Czajka." National Research Council. 2010. Databases for Estimating Health Insurance Coverage for Children: A Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/13024.
×

TABLE 10-17 Estimates of the Poor and Near-Poor Elderly: CPS and ACS, 2007 and 2008

Estimate

2007

2008

CPS

ACS

CPS

ACS

 

Millions of Persons

All Persons Aged 65 and Older Poverty Status

36.79

36.23

37.79

37.20

Poor

3.56

3.46

3.66

3.68

Near-poor

9.70

8.14

10.03

8.16

Total Low Income

13.26

11.59

13.69

11.84

 

Percentage of the Population

All Persons Aged 65 and Older Poverty Status

100.00

100.00

100.00

100.00

Poor

9.70

9.50

9.70

9.90

Near-poor

26.40

22.50

26.60

21.90

Total Low Income

36.00

32.00

36.20

31.80

NOTE: The poor have a family income below the poverty threshold. The near-poor have a family income at or above the poverty threshold but below twice the poverty threshold.

SOURCES: Mathematica Policy Research, Inc., from tabulations of poverty status in calendar years 2007 and 2008 from the 2008 and 2009 CPS ASEC supplements and the 2007 and 2008 ACS.

TABLE 10-18 Number and Percentage Uninsured by Relative Income: All Persons, CPS and ACS, 2008

Estimate

CPS

ACS

 

Uninsured Persons (millions)

All Persons

46.25

44.76

Income Relative to Poverty

 

 

Under 100%

12.12

11.37

100 to under 200%

13.64

13.54

200 to under 400%

13.80

13.63

400 or more

6.68

6.22

Low Income (under 200%)

25.76

24.91

 

Percentage Uninsured

All Persons

15.40

15.20

Income Relative to Poverty

 

 

Under 100%

30.50

29.10

100 to under 200%

24.30

25.90

200 to under 400%

14.80

14.90

400 or more

6.00

5.50

Low Income (under 200%)

26.90

27.30

SOURCE: Mathematica Policy Research, Inc., from tabulations of poverty status in calendar year 2002 from the 2003 NHIS.

Suggested Citation:"10 Income and Poverty Measurement in Surveys of Health Insurance Coverage--John L. Czajka." National Research Council. 2010. Databases for Estimating Health Insurance Coverage for Children: A Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/13024.
×

TABLE 10-19 Number and Percentage Uninsured by Relative Income: Children Under 18, CPS and ACS, 2008

Estimate

CPS

ACS

 

Uninsured Persons (millions)

All Children Under Age 18

7.25

7.20

Income Relative to Poverty

 

 

Under 100%

2.21

2.10

100 to under 200%

2.35

2.48

200 to under 400%

1.99

2.01

400 or more

0.70

0.61

Low Income (under 200%)

4.56

4.58

 

Percentage Uninsured

All Children Under Age 18

9.80

9.90

Income Relative to Poverty

 

 

Under 100%

15.70

15.90

100 to under 200%

14.70

15.90

200 to under 400%

8.60

8.70

400 or more

3.40

2.90

Low Income (under 200%)

15.20

15.90

SOURCE: Mathematica Policy Research, Inc., tabulations of the 2009 CPS ASEC supplement and the 2008 ACS.

Uninsured rates among the elderly are not given much attention in the literature, as they are exceedingly low due to the broad coverage provided by Medicare, and the estimates are probably dominated by measurement error. Compared with the CPS, the ACS finds a markedly lower uninsured rate among the poor elderly—3.6 versus 6.2 percent—but at higher levels of relative income, the estimates are remarkably similar (see Table 10-21).

Residency Among Young Adults

By asking that college students living away from home be counted at their place of residence at the time of the survey, the ACS departs from the CPS in a way that could lead to higher estimates of the college student population. College students living away from home are given two opportunities to be counted, in effect: once at school and once at home. While the ACS asks that students be counted at school, parents responding to the survey may choose to count their children at home regardless. The fact that the ACS includes college dormitories in its sample frame while the CPS does not means that every college student has a chance to be counted twice in the ACS, whereas students in dormitories are precluded from this possibility in the CPS.

Suggested Citation:"10 Income and Poverty Measurement in Surveys of Health Insurance Coverage--John L. Czajka." National Research Council. 2010. Databases for Estimating Health Insurance Coverage for Children: A Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/13024.
×

TABLE 10-20 Number and Percentage Uninsured by Relative Income: Nonelderly Adults, CPS and ACS, 2008

Estimate

CPS

ACS

 

Uninsured Persons (millions)

All Nonelderly Adults

38.35

37.03

Income Relative to Poverty

 

 

Under 100%

9.68

9.14

100 to under 200%

11.12

10.91

200 to under 400%

11.66

11.47

400 or more

5.88

5.50

Low Income (under 200%)

20.80

20.05

 

Percentage Uninsured

All Nonelderly Adults

20.40

20.00

Income Relative to Poverty

 

 

Under 100%

44.00

41.30

100 to under 200%

37.10

38.10

200 to under 400%

20.30

20.40

400 or more

7.50

7.00

Low Income (under 200%)

39.50

40.00

SOURCE: Mathematica Policy Research, Inc., from tabulations of the 2009 CPS ASEC supplement and the 2008 ACS.

Comparative estimates from the two surveys for 2008 show 1.8 million more young adults aged 18-24 enrolled in college in the ACS than the CPS; 1.0 million fewer young adults not enrolled in college in the ACS than the CPS; and almost 0.9 million more young adults in total in the ACS than the CPS (see Table 10-22). Despite students in dormitories being excluded from the ACS poverty universe, it still finds a higher fraction of young adults to be poor (20.0 versus 18.4 percent). Virtually all of the difference is due to college students, whose poverty rate is 5 percentage points higher in the ACS than the CPS despite the exclusion of 18.7 percent of students from the ACS poverty universe. Among those who are not enrolled in college, virtually none is excluded from the ACS poverty universe, and the distribution of persons across the four classes of relative income is quite similar between the ACS and the CPS. It is clear from the comparative distributions that the students excluded from the poverty universe in the ACS are drawn from higher income levels, as that is where the ACS falls well short of the CPS.

A comparison of uninsured rates by college enrollment and relative income provides clear evidence that the college students identified as poor in the ACS look more like students from higher income families in the CPS. The uninsured rate of 19.9 percent for poor college students in the ACS compares with 32.5 percent in the CPS, whereas both surveys

Suggested Citation:"10 Income and Poverty Measurement in Surveys of Health Insurance Coverage--John L. Czajka." National Research Council. 2010. Databases for Estimating Health Insurance Coverage for Children: A Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/13024.
×

TABLE 10-21 Number and Percentage Uninsured by Relative Income: Elderly Persons, CPS and ACS, 2008

Estimate

CPS

ACS

 

Uninsured Persons (millions)

All Persons Aged 65 and Older

0.65

0.54

Income Relative to Poverty

 

 

Under 100%

0.22

0.13

100 to under 200%

0.17

0.15

200 to under 400%

0.15

0.15

400 or more

0.10

0.11

Low Income (under 200%)

0.39

0.28

 

Percentage Uninsured

All Persons Aged 65 and Older

1.70

1.40

Income Relative to Poverty

 

 

Under 100%

6.20

3.60

100 to under 200%

1.70

1.80

200 to under 400%

1.20

1.20

400 or more

0.90

0.80

Low Income (under 200%)

2.90

2.40

SOURCE: Mathematica Policy Research, Inc., from tabulations of the 2009 CPS ASEC supplement and the 2008 ACS.

find uninsured rates around 20 percent for college students with incomes between 200 and 400 percent of poverty (see Table 10-23). Among those not enrolled in college, the ACS has somewhat higher uninsured rates in every category of relative income. I have no explanation for this. Regardless of college enrollment, only 6 to 7 percent of young adults excluded from the ACS poverty universe are reported to be without health insurance coverage.

IMPLICATIONS

The principal findings can be summarized quite simply. At high levels of aggregation, the ACS produces estimates of income and health insurance coverage that look strikingly similar to those obtained from the CPS, despite notable differences in measurement. This suggests but by no means proves that, with its much larger sample size, the ACS could be used to develop direct estimates of low-income uninsured children at the state level and even lower levels of geography that would mimic what could be obtained from the CPS if it had a much bigger sample.

Although the ACS seems to fare well with a modest set of income questions, the NHIS illustrates potential adverse consequences from devoting

Suggested Citation:"10 Income and Poverty Measurement in Surveys of Health Insurance Coverage--John L. Czajka." National Research Council. 2010. Databases for Estimating Health Insurance Coverage for Children: A Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/13024.
×

TABLE 10-22 Persons Aged 18 to 24 by College Enrollment and Relative Income: CPS and ACS, 2008

Estimate

CPS

ACS

 

Millions of Persons

All Persons Aged 18 to 24

28.47

29.31

 

Percentage of Persons

Income Relative to Poverty

100.00

100.00

Outside the poverty universe

0.00

7.90

Under 100%

18.40

20.00

100 to under 200%

20.40

18.90

200 to under 400%

31.10

27.80

400% or more

30.00

25.40

 

Millions of Persons

Enrolled in College

10.40

12.22

 

Percentage of Persons

Income Relative to Poverty

100.00

100.00

Outside the poverty universe

0.00

18.70

Under 100%

15.60

20.70

100 to under 200%

16.70

12.90

200 to under 400%

28.10

21.10

400% or more

39.60

26.60

 

Millions of Persons

Not Enrolled in College

18.07

17.08

 

Percentage of Persons

Income Relative to Poverty

100.00

100.00

Outside the poverty universe

0.00

0.20

Under 100%

20.10

19.50

100 to under 200%

22.50

23.20

200 to under 400%

32.90

32.60

400% or more

24.50

24.60

SOURCE: Mathematica Policy Research, Inc., tabulations of the 2009 CPS ASEC supplement

too few questions to income measurement—particularly among families at the low end of the income distribution. Also, differences between the ACS and the CPS begin to emerge as income and/or the population are disaggregated. Such differences are often but certainly not always consistent with differences in the two surveys’ approaches to measuring income, including reference period as well as the way that income is defined. For

Suggested Citation:"10 Income and Poverty Measurement in Surveys of Health Insurance Coverage--John L. Czajka." National Research Council. 2010. Databases for Estimating Health Insurance Coverage for Children: A Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/13024.
×

TABLE 10-23 Percentage Uninsured among Persons Aged 18 to 24 by College Enrollment and Relative Income: CPS and ACS, 2008

Estimate

CPS

ACS

All Persons Aged 18 to 24

28.80

28.60

Income Relative to Poverty

 

 

Outside the poverty universe

n/a

6.30

Under 100%

41.90

36.60

100 to under 200%

39.80

43.00

200 to under 400%

28.70

30.90

400% or more

13.40

15.90

Enrolled in College

19.10

15.90

Income Relative to Poverty

 

 

Outside the poverty universe

n/a

6.30

Under 100%

32.50

19.90

100 to under 200%

30.60

29.60

200 to under 400%

20.40

20.70

400% or more

8.10

9.00

Not Enrolled in College

34.40

37.70

Income Relative to Poverty

 

 

Outside the poverty universe

n/a

6.70

Under 100%

46.10

49.40

100 to under 200%

43.80

48.30

200 to under 400%

32.80

35.70

400% or more

18.30

21.20

NOTE: n/a = not available.

SOURCE: Mathematica Policy Research, Inc., from tabulations of the 2009 CPS ASEC supplement and the 2008 ACS.

example, in recent years the ACS lags the CPS in its response to changes in the economy—as it should, given the longer reference period in the ACS. This could be viewed as a drawback for many prospective uses of the ACS, but for state- and lower level estimates it is important to remember that the Census Bureau combines 3 years of CPS data to produce the annual estimates of low-income uninsured children that the ACS would potentially replace. Compared with a 3-year moving average, the ACS data show greater rather than lesser sensitivity to economic change.

The Census Bureau and the research community are only beginning to understand the implications of many aspects of ACS data collection. From the perspective of income measurement, the chief issues revolve around the ACS reference period and the ways that the Census Bureau has attempted to annualize the estimates, the residence rules for college students living away from home, the small number of questions

Suggested Citation:"10 Income and Poverty Measurement in Surveys of Health Insurance Coverage--John L. Czajka." National Research Council. 2010. Databases for Estimating Health Insurance Coverage for Children: A Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/13024.
×

used to capture the diverse sources of unearned income, and the absence of data on the relationships among persons who are unrelated to the householder.

ACKNOWLEDGMENTS

Much of the work presented herein was prepared by Mathematica Policy Research, Inc., under contract to the Office of the Assistant Secretary for Planning and Evaluation, U.S. Department of Health and Human Services and presented in a final report, “Income Data for Policy Analysis: A Comparative Assessment of Eight Surveys” dated December 23, 2008, and coauthored by Gabrielle Denmead. The findings and conclusions reported here are those of the author and do not necessarily represent the views of ASPE or HHS.

REFERENCE

Czajka, J.L., and Denmead, G. (2008). Income Data for Policy Analysis: A Comparative Assessment of Eight Surveys. Report from Mathematica Policy Research, Inc. Available: http://www.mathematica-mpr.com/publications/pdfs/incomedata.pdf [October 2010].

Suggested Citation:"10 Income and Poverty Measurement in Surveys of Health Insurance Coverage--John L. Czajka." National Research Council. 2010. Databases for Estimating Health Insurance Coverage for Children: A Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/13024.
×
Page 109
Suggested Citation:"10 Income and Poverty Measurement in Surveys of Health Insurance Coverage--John L. Czajka." National Research Council. 2010. Databases for Estimating Health Insurance Coverage for Children: A Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/13024.
×
Page 110
Suggested Citation:"10 Income and Poverty Measurement in Surveys of Health Insurance Coverage--John L. Czajka." National Research Council. 2010. Databases for Estimating Health Insurance Coverage for Children: A Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/13024.
×
Page 111
Suggested Citation:"10 Income and Poverty Measurement in Surveys of Health Insurance Coverage--John L. Czajka." National Research Council. 2010. Databases for Estimating Health Insurance Coverage for Children: A Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/13024.
×
Page 112
Suggested Citation:"10 Income and Poverty Measurement in Surveys of Health Insurance Coverage--John L. Czajka." National Research Council. 2010. Databases for Estimating Health Insurance Coverage for Children: A Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/13024.
×
Page 113
Suggested Citation:"10 Income and Poverty Measurement in Surveys of Health Insurance Coverage--John L. Czajka." National Research Council. 2010. Databases for Estimating Health Insurance Coverage for Children: A Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/13024.
×
Page 114
Suggested Citation:"10 Income and Poverty Measurement in Surveys of Health Insurance Coverage--John L. Czajka." National Research Council. 2010. Databases for Estimating Health Insurance Coverage for Children: A Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/13024.
×
Page 115
Suggested Citation:"10 Income and Poverty Measurement in Surveys of Health Insurance Coverage--John L. Czajka." National Research Council. 2010. Databases for Estimating Health Insurance Coverage for Children: A Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/13024.
×
Page 116
Suggested Citation:"10 Income and Poverty Measurement in Surveys of Health Insurance Coverage--John L. Czajka." National Research Council. 2010. Databases for Estimating Health Insurance Coverage for Children: A Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/13024.
×
Page 117
Suggested Citation:"10 Income and Poverty Measurement in Surveys of Health Insurance Coverage--John L. Czajka." National Research Council. 2010. Databases for Estimating Health Insurance Coverage for Children: A Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/13024.
×
Page 118
Suggested Citation:"10 Income and Poverty Measurement in Surveys of Health Insurance Coverage--John L. Czajka." National Research Council. 2010. Databases for Estimating Health Insurance Coverage for Children: A Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/13024.
×
Page 119
Suggested Citation:"10 Income and Poverty Measurement in Surveys of Health Insurance Coverage--John L. Czajka." National Research Council. 2010. Databases for Estimating Health Insurance Coverage for Children: A Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/13024.
×
Page 120
Suggested Citation:"10 Income and Poverty Measurement in Surveys of Health Insurance Coverage--John L. Czajka." National Research Council. 2010. Databases for Estimating Health Insurance Coverage for Children: A Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/13024.
×
Page 121
Suggested Citation:"10 Income and Poverty Measurement in Surveys of Health Insurance Coverage--John L. Czajka." National Research Council. 2010. Databases for Estimating Health Insurance Coverage for Children: A Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/13024.
×
Page 122
Suggested Citation:"10 Income and Poverty Measurement in Surveys of Health Insurance Coverage--John L. Czajka." National Research Council. 2010. Databases for Estimating Health Insurance Coverage for Children: A Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/13024.
×
Page 123
Suggested Citation:"10 Income and Poverty Measurement in Surveys of Health Insurance Coverage--John L. Czajka." National Research Council. 2010. Databases for Estimating Health Insurance Coverage for Children: A Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/13024.
×
Page 124
Suggested Citation:"10 Income and Poverty Measurement in Surveys of Health Insurance Coverage--John L. Czajka." National Research Council. 2010. Databases for Estimating Health Insurance Coverage for Children: A Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/13024.
×
Page 125
Suggested Citation:"10 Income and Poverty Measurement in Surveys of Health Insurance Coverage--John L. Czajka." National Research Council. 2010. Databases for Estimating Health Insurance Coverage for Children: A Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/13024.
×
Page 126
Suggested Citation:"10 Income and Poverty Measurement in Surveys of Health Insurance Coverage--John L. Czajka." National Research Council. 2010. Databases for Estimating Health Insurance Coverage for Children: A Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/13024.
×
Page 127
Suggested Citation:"10 Income and Poverty Measurement in Surveys of Health Insurance Coverage--John L. Czajka." National Research Council. 2010. Databases for Estimating Health Insurance Coverage for Children: A Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/13024.
×
Page 128
Suggested Citation:"10 Income and Poverty Measurement in Surveys of Health Insurance Coverage--John L. Czajka." National Research Council. 2010. Databases for Estimating Health Insurance Coverage for Children: A Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/13024.
×
Page 129
Suggested Citation:"10 Income and Poverty Measurement in Surveys of Health Insurance Coverage--John L. Czajka." National Research Council. 2010. Databases for Estimating Health Insurance Coverage for Children: A Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/13024.
×
Page 130
Suggested Citation:"10 Income and Poverty Measurement in Surveys of Health Insurance Coverage--John L. Czajka." National Research Council. 2010. Databases for Estimating Health Insurance Coverage for Children: A Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/13024.
×
Page 131
Suggested Citation:"10 Income and Poverty Measurement in Surveys of Health Insurance Coverage--John L. Czajka." National Research Council. 2010. Databases for Estimating Health Insurance Coverage for Children: A Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/13024.
×
Page 132
Suggested Citation:"10 Income and Poverty Measurement in Surveys of Health Insurance Coverage--John L. Czajka." National Research Council. 2010. Databases for Estimating Health Insurance Coverage for Children: A Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/13024.
×
Page 133
Suggested Citation:"10 Income and Poverty Measurement in Surveys of Health Insurance Coverage--John L. Czajka." National Research Council. 2010. Databases for Estimating Health Insurance Coverage for Children: A Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/13024.
×
Page 134
Suggested Citation:"10 Income and Poverty Measurement in Surveys of Health Insurance Coverage--John L. Czajka." National Research Council. 2010. Databases for Estimating Health Insurance Coverage for Children: A Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/13024.
×
Page 135
Suggested Citation:"10 Income and Poverty Measurement in Surveys of Health Insurance Coverage--John L. Czajka." National Research Council. 2010. Databases for Estimating Health Insurance Coverage for Children: A Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/13024.
×
Page 136
Suggested Citation:"10 Income and Poverty Measurement in Surveys of Health Insurance Coverage--John L. Czajka." National Research Council. 2010. Databases for Estimating Health Insurance Coverage for Children: A Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/13024.
×
Page 137
Suggested Citation:"10 Income and Poverty Measurement in Surveys of Health Insurance Coverage--John L. Czajka." National Research Council. 2010. Databases for Estimating Health Insurance Coverage for Children: A Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/13024.
×
Page 138
Suggested Citation:"10 Income and Poverty Measurement in Surveys of Health Insurance Coverage--John L. Czajka." National Research Council. 2010. Databases for Estimating Health Insurance Coverage for Children: A Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/13024.
×
Page 139
Suggested Citation:"10 Income and Poverty Measurement in Surveys of Health Insurance Coverage--John L. Czajka." National Research Council. 2010. Databases for Estimating Health Insurance Coverage for Children: A Workshop Summary. Washington, DC: The National Academies Press. doi: 10.17226/13024.
×
Page 140
Next: 11 Using Uninsured Data to Track State CHIP Programs--John McInerney »
Databases for Estimating Health Insurance Coverage for Children: A Workshop Summary Get This Book
×
Buy Paperback | $62.00 Buy Ebook | $49.99
MyNAP members save 10% online.
Login or Register to save!
Download Free PDF

This report summarizes the proceedings of a workshop convened in June 2010 to critically examine the various databases that could provide national and state-level estimates of low-income uninsured children and could be effectively used as criteria for monitoring children's health insurance coverage.

  1. ×

    Welcome to OpenBook!

    You're looking at OpenBook, NAP.edu's online reading room since 1999. Based on feedback from you, our users, we've made some improvements that make it easier than ever to read thousands of publications on our website.

    Do you want to take a quick tour of the OpenBook's features?

    No Thanks Take a Tour »
  2. ×

    Show this book's table of contents, where you can jump to any chapter by name.

    « Back Next »
  3. ×

    ...or use these buttons to go back to the previous chapter or skip to the next one.

    « Back Next »
  4. ×

    Jump up to the previous page or down to the next one. Also, you can type in a page number and press Enter to go directly to that page in the book.

    « Back Next »
  5. ×

    Switch between the Original Pages, where you can read the report as it appeared in print, and Text Pages for the web version, where you can highlight and search the text.

    « Back Next »
  6. ×

    To search the entire text of this book, type in your search term here and press Enter.

    « Back Next »
  7. ×

    Share a link to this book page on your preferred social network or via email.

    « Back Next »
  8. ×

    View our suggested citation for this chapter.

    « Back Next »
  9. ×

    Ready to take your reading offline? Click here to buy this book in print or download it as a free PDF, if available.

    « Back Next »
Stay Connected!