Key Themes: Reflections from the Child Indicators Projects
The principal speaker at the session was Cynthia Taeuber, Program Policy Advisor of the University of Baltimore and the U.S. Bureau of the Census. The session was coordinated by Allen Harden of Chapin Hall. The purpose of this presentation was to give an overview of Census 2000 and the American Community Survey and review how they can be used to build child and family well-being indicators. Ms. Taeuber's Powerpoint presentation is attached following text summary and the slide numbers mentioned in the text refer to that presentation.
The speaker began by comparing and contrasting the Census 2000 and the American Community Survey. While the main purpose of Census 2000 is to count the population every 10 years, the main purpose of the American Community Survey is to provide yearly updated information on the characteristics of the population in small geographic areas and also, for relatively small population groups in larger geographic regions. The questions on the American Community Survey provide indicators that are similar to those of the Census 2000 long form (described below).
The Census 2000 short form asks seven questions (figure 11 of handout) of every person and housing unit in the U.S. about age, race, Hispanic origin, gender, household relationship, and housing tenure (owner or rented). Field staff determine characteristics of vacant housing units. Additional questions are asked in the long form of a sample of every person in 1 in 6 housing units and of 1 in 6 people living in group quarters (a national average). Population statistics are provided on a range of topics (see figures 4-13 of handout) including marital status, place of birth/citizenship, disability, ancestry, migration, language spoken at home and ability to speak English, school enrollment and educational attainment, grandparents as caregivers, place of work and journey to work, occupation, industry and class of worker, work status in the week before the census or the last year in which the person worked, and income in 1999. Housing statistics based on the long form include number of rooms and bedrooms, plumbing and kitchen facilities, the age and value of the housing unit, and questions to indicate housing affordability including the cost and type of utilities, mortgage/rent paid, and taxes and insurance. Results are available for geographic levels, including the Block (short form information only), Block Group, Census Tract, County, Metropolitan area, state, and national levels.
Ms. Taeuber provided handouts that described options for comparing racial categories from the 1990 and 2000 censuses (see slides 13-15, 18-21) and detailed Census 2000 products that are available currently or in the near future (slides 16-17, 22-37). Information about the products, documentation, and the product release schedule are found on the Census Bureau's website: www.census.gov.
Selected helpful sites for Census 2000:
The American Community Survey, once the sample is fully implemented in every county (planned to start in 2003, slide 49), will provide annual-average estimates of demographic, housing, social, and economic characteristics updated every year for the nation, all states, and as well as for all jurisdictions of 65,000 or more people such as cities, counties, metropolitan areas, and for large population groups. Statistics for smaller jurisdictions, geographic sub-areas, and smaller population groups will be updated for multi-year averages (3-year averages for areas of 20,000 -64,999 and 5-year averages for areas of less than 20,000 people).
Ms. Taeuber outlined several new opportunities for information that the American Community Survey will provide. First, because the survey is updated every year, it allows measurement of the level and direction of change for small areas and population groups (slides 38-43 of handout) on topics such as unemployment and poverty. Second, migration patterns can be better analyzed through this survey data (slide 44). Also, the American Community Survey will improve the ability to develop performance measures for local programs (slide 45). Finally, informed strategic decision making will be made possible by providing the community context for the assessment of needs and resources (slides 46-47).
The speaker emphasized that the American Community Survey is a bridge between Census 2000 and the future (slide 48). The Census Bureau plans to replace the Census long form with the American Community Survey for the 2010 census. She noted, however, that researchers will need to pay attention to the range of error in the estimates (that is, the confidence intervals - slide 51) because, like the decennial census long form, the data are from a sample of the population. For smaller areas, the sample will be accumulated over multiple years to achieve sufficient sample to approximate the sampling error of the decennial census long form. For example, areas of 20,000 to 64,999 can use data averaged over three years starting in 2006, and every year thereafter. For rural areas and city neighborhoods or population groups of less than 20,000 people, starting in 2008 and every year thereafter, a 5-year accumulation of sample will provide estimates similar to those of the decennial census long form. These averages will be updated every year, so that eventually, it will be possible to measure changes over time for small areas and population groups.
The American Community Survey is currently in its development stage. The Census Bureau plans that, beginning in 2003, the American Community Survey will be implemented in all counties across the country of the United States if Congress allocates the necessary funding (slide 49). The fully implemented survey would include three million addresses (households and group quarters). Data are collected by mail with follow-up calls and visits from Census Bureau staff if a household does not respond.
Selected Helpful Sites for ACS (slide 50):
Census 2000: Changes between 1990 and 2000: Several changes were made in the Census 2000 that have implications for indicator development. One new question asked about grandparents as caregivers for dependent children (slide 5). Grandparents who had grandchildren living in their households were asked if they are responsible for the basic needs of the children in the household and for what time period. Another change involved a revision of the question on disability status (slide 6). In Census 2000, the question specifically asks about vision or hearing impairments as well as conditions that limit learning or remembering. And, for the first time, the respondent can select one or more races (slide 13). As in past Censuses, there is a separate question on Hispanic origin (slide 12). In Census 2000, this question was asked before the revised question on race.
The additional choices that the revised question on race allowed meant that tabulations of race and Hispanic origin are more complicated for Census 2000 than for past Censuses. According to Ms. Taeuber, less than 2 percent of the total US population marked two or more races (slides 18-19) although the percentage is higher among children. There are 126 race and Hispanic origin categories in some Census products (slide 15). Most products, however, show only the counts of those who reported six single racial groups and "two or more races" (slide 14). The speaker also provided a paper by Sharon M. Lee entitled Using the New Racial Categories in the 2000 Census for further discussion of the implications of the new categories. This paper was funded by the Annie E. Casey Foundation and the Population Reference Bureau March 2000 (http://www.aecf.org/kidscount/categories.htm).
Ms. Taeuber cautioned the audience about overly simplified direct comparisons of survey data, such as from the American Community Survey or the decennial census, with administrative data. There are crucial differences in concepts and data collection methods among data sets. As such, estimates of population characteristics from surveys such as the decennial census and the American Community Survey will differ. For example, Ms. Taeuber headed a study of differences in 1990 Census long-form estimates for Maryland of poor children and poor families, and counts of AFDC recipients and cases. The most important factor in the differences was the sampling error. Sampling error is present in all surveys but not in administrative records (this accounted for most of the apparent discrepancy among children except in the city of Baltimore). The other two factors of significance were undercounting people in the 1990 census and differences in the definition and reporting of "income" between the two data sets.
The remainder of the discussion focused on some of the challenges that the Census Bureau faces. The Congress approves questions on the decennial census and the American Community Survey. They have approved only those questions mandated or required by federal legislation or court cases. That presents considerable challenges to adding new questions to the American Community Survey or the next Census. If data are collected but not presented in a Census product that meets the researcher's needs, it is possible to request special tabulations of Census data. The speaker cautioned that this usually requires considerable time and money (slide 35). There are other alternatives such as the Public Use Microdata files if the limitations on geographic area and sample size are not a problem (slide 33). The speaker encourages researchers to report their needs to the Census Bureau so they can consider these for future American Community Survey or Census products (slides 67-68).
Slide 1
Slide 2
Slide 3
Slide 4
Slide 5
Slide 6
Slide 7
Slide 8
Slide 9
Slide 10
Slide 11
Slide 12
Changes in Census 2000 question from 1990-same question on American
Community Survey: "Hispanic or Latino" asked before race
Every respondent to Census 2000 was asked to respond to the Hispanic
origin question.
Those who were notof Hispanic origin marked the box "No, not Spanish/
Hispanic/Latino."
People who were of Hispanic origin mark the box indicating the specific
group they belong to: Mexican, Puerto Rican, Cuban, or other Spanish, Hispanic,
Latino, such as Spanish, Honduran, or Venezuelan.
People of Hispanic origin may be of any race.
Slide 13
Changes in Census 2000 question from 1990-same question on American
Community Survey: Respondents may select one or more races
Asian and Pacific Islander category split:
Asian
Native Hawaiian and Other Pacific Islander
Slide 14
For the first time ever Respondents to the decennial census were
allowed to mark more than 1 race category. Race tabulations include six "Alone"
categories.
White
Black or African American
American Indian or Alaska Native
Asian
Native Hawaiian and Other Pacific Islander, and
Some other race
There are also 57 possibilities of "Two or more races."
15 combinations of 2 races
20 combinations of 3 races
15 combinations of 4 races
6 combinations of 5 races
1 combination of 6 races
The race question will also supply information on 36 American Indian groups, 6 Native Alaskan groups, 17 Asian groups, and 13 Pacific Islander groups.
Slide 15
Slide 16
Slide 17
The Redistricting Summary File was the first Census 2000 product released.
It contains the data from the Census short form that is needed for redistricting.
Redistricting is the process of revising the geographic boundaries within
a state from which people elected to the U.S. House of Representatives, state
legislatures, county and city political subdivisions, school boards, and
other voting districts.
This file is available on the Internet and CD-ROM.
The statistical summaries contain population totals and the population
18 years and older. You can subtract to obtain counts of children under 18
for every block in the country. There are summaries for geographic areas,
including states, counties, voting districts, county subdivisions, American
Indian/Alaska Native/Native Hawaiian areas, census tracts, block groups,
and blocks.
There are summaries for the total population, but not for age groups,
by race, Hispanic origin, and voting age for geographic areas down to blocks.
Because Census 2000 allowed respondents to check one or more race categories,
the race tabulations are verydetailed.
This file contains block level data showing 63 race categories and
Hispanic origin.
Slide 18
In Census 2000, nearly all respondents reported only one race. White alone, accounted for 75 percent of all people living in the United States. The African American alone represented 12 percent. American Indian and Alaska Natives alone represented just under 1 percent of the total. Approximately 4 percent of respondents indicated Asian only. The smallest race group was the Native Hawaiian and other Pacific Islander population alone which represented 0.1 percent of the population.
Almost 6 percent of all respondents indicated that they were Some other race. And about 2 percent of all respondents reported two or more races.
Slide 19
Of the nearly 7 million respondents in this category, 93 percent reported
exactly two races.
16 percent were White AND American Indian and Alaska Native.
13 percent were White AND Asian.
11 percent White AND African American.
Of all respondents reporting exactly two races, 47 percent included
some other race as one of the two races.
Slide 20
About 13 percent of the population (35 million people) are Latino,
according to Census 2000.
About 59 percent of Hispanics were of Mexican origin.
The next largest group was Puerto Rican, accounting for about 10 percent
of all Hispanics.
Cubans were the third largest group, making up just 4 percent of the
total Hispanic population.
Half of all Hispanics live in just two states: California and
Texas.
There are seven states with Hispanic populations of more than one
million: California, Texas, New York, Florida, Illinois, Arizona, & New
Jersey
New Mexico had the highest share of residents who were Hispanic, 42
percent.
Hispanics may be of any race.
Slide 21
About 48 percent of the Latino population reported that they were
White alone.
Forty-two percent said that they were some other race alone. Only
2 percent reported Black only. One 1 percent were American Indian and Alaska
Native only. About 6 percent were two or more races.
Slide 22
Slide 23
Slide 24
Slide 25
Slide 26
Slide 27
Slide 28
Slide 29
Slide 30
Slide 31
Slide 32
Slide 33
Slide 34
Slide 35
Slide 36
Slide 37
Slide 38
Slide 39
Slide 40
Slide 41
Slide 42
Slide 43
Slide 44
Slide 45
Slide 46
Slide 47
Slide 48
Slide 49
Slide 50
Slide 51
Slide 52
Slide 53
Slide 54
Slide 55
Slide 56
Slide 57
Slide 58
Slide 59
Slide 60
Slide 61
Slide 62
Slide 63
Slide 64
Slide 65
Slide 66
Slide 67
Slide 68
The moderator was Robert Goerge of Chapin Hall. The session focused on the ethical and attendant legal barriers to sharing data. The first speaker was Loretta Fuddy of Hawaii. Slides from a Powerpoint used by Ms. Fuddy follow the text.
Ms. Fuddy discussed some of the difficulties of sharing data, using as an example Hawaii's long-term plan to link birth certificate data to other data as part of a perinatal data system. Barriers to the creation of the system include these:
Participants from other states, and Chapin Hall staff members, noted similar difficulties with HIPAA data, and, in some states, with TANF and Medicaid data.
Debbykay Peterson and John Oswald sketched Minnesota's data linkage initiative. Minnesota has an extremely comprehensive privacy law. The state attorney general, in particular, favors limiting use of health data in response, in part, to the activities of health maintenance organizations. Key questions include:
Strategies for protecting privacy among shared data include
A memorandum of understanding among users can be a powerful protector of all parties.
Slide 1.
Slide 2
Slide 3
Slide 4
Slide 5
Slide 6
Slide 7
Slide 8.
Slide 9
Slide 10
Slide 11
Slide 12
Slide 13
Slide 14
Slide 15
Slide 16
Slide 17
Slide 18
Slide 19
Top of Page |
Contents
Back to Key Themes main pages
Home Pages:
Advancing States' Child Indicators Initiatives
Human Services Policy (HSP)
Assistant Secretary for Planning and Evaluation
(ASPE)
U.S. Department of Health and Human Services
(HHS)