Recipiency and Dependency Rates: 1987-95
[Chapter I in PDF format including all figures and tables]
The Welfare Indicators Act of 1994 (Pub. L. 103-432) directed the Secretary of Health and Human Services (HHS) to publish an annual report on welfare dependency. The purpose of this report is to address questions concerning the extent to which American families depend on income from welfare programs. HHS has been specifically directed to address the rate of welfare dependency, the degree and duration of welfare recipiency and dependence, and predictors of welfare dependence. The Act further specified that analyses of means-tested assistance should include benefits under the Aid to Families with Dependent Children (AFDC) program, now Temporary Assistance to Needy Families (TANF); the Food Stamp Program; and the Supplemental Security Income (SSI) program.
An Advisory Board on Welfare Indicators was established under the 1994 Act to assist the Secretary in defining welfare dependence, developing indicators of welfare dependence, and choosing appropriate data for inclusion in the first annual report. The Board consisted of a bipartisan group of experts appointed by the Senate, the House of Representatives and the President. Before its termination in October 1997, the Board developed a statistical definition of welfare dependence and oversaw the production of the first of these annual reports.
This March 2000 report, the third annual report, gives updated data on the measures of welfare recipiency, dependency, and predictors of welfare dependence developed for previous reports. It differs in two respects from earlier volumes. First, this report focuses on a smaller set of indicators and predictors of dependency, in keeping with Congressional intent. The reduction in length of the report also reflects the decision to move some of the more detailed data on poverty and deprivation to other Departmental publications.(1) A second change is that the date of publication has been moved from October to March, in conformance with the report's authorizing legislation, which requires the report to be released within sixty days of the start of the legislative session. A March release also allows the Department to present more timely data, as many important administrative and national survey figures are released at the end of the year.
This introductory chapter provides an overview of the specific summary measures of welfare dependence proposed by the Advisory Board. It also discusses summary measures of poverty, following the Board's recommendation that dependence measures not be assessed in isolation from measures of deprivation. Analysis of both measures is important because changes in dependence measures could result either from increases in work activity and other factors that would raise family incomes, or from sanctions or other changes in welfare programs that would reduce welfare program participation but might not improve the material circumstances of these families. The introduction concludes with a brief discussion of data sources used for the report.
Chapter II of the report, Indicators of Dependence, presents a broader group of indicators of welfare recipiency and dependence. These indicators include measures of the extent of recipiency for each of the three programs considered separately, as well as information on income from all three programs in combination. Interactions of AFDC/TANF, SSI and food stamp benefits with periods of employment and with benefits from other programs are also shown. The second chapter also includes data on movements on and off welfare programs.
Chapter III, Predictors and Risk Factors Associated with Welfare Receipt, focuses on predictors of welfare dependence — risk factors believed to be associated with welfare receipt in some way. These predictors are shown in three different groups:
Additional data are presented in two appendices. Appendix A provides basic program data on each of the main welfare programs and their recipients, while Appendix B includes additional data on non-marital childbearing. The main welfare programs included in Appendix A are:
Welfare dependence, like poverty, is a continuum, with variations in degree and in duration. Families may be more or less dependent if larger or smaller shares of their total resources are derived from welfare programs. The amount of time over which a family depends on welfare might also be considered in assessing its degree of dependency. Nevertheless, a summary measure of dependence to be used as an indicator for policy purposes must have some fixed parameters that allow one to determine which families should be counted as dependent, just as the poverty line defines who is poor under the official standard. The definition of dependence proposed by the Advisory Board for this purpose is as follows:
A family is dependent on welfare if more than 50 percent of its total income in a one-year period comes from AFDC, food stamps and/or SSI, and this welfare income is not associated with work activities. Welfare dependence is the proportion of all families who are dependent on welfare.
This measure is not without its limitations. The Advisory Board recognized that no single measure could fully capture all aspects of dependence and that the proposed measure should be examined in concert with other key indicators of dependence and deprivation. In addition, while the proposed definition would count unsubsidized and subsidized employment and work required to obtain benefits as work activities, existing data sources do not permit distinguishing between welfare income associated with work activities and non-work-related welfare benefits. As a result, the data shown in this report overstate the incidence of dependence (as defined above) because welfare income associated with work required to obtain benefits is classified as welfare and not as income from work. This issue may be growing in importance under the increased work requirements of the TANF program. In 1998, the percentage of welfare recipients who were working (including employment, work experience, and community service) reached an all-time high of 27 percent, compared to the 7 percent recorded in 1992.(2)
This proposed definition also represents an essentially arbitrary choice of a percentage (50 percent) of income from welfare beyond which families will be considered dependent. However, it is relatively easy to measure and to track over time, and is likely to be associated with any very large changes in total dependence, however defined. For example, as the recent changes in welfare law move more recipients into employment or work-related activities, dependence under this definition is expected to decline.
In 1995, the most recent year for which complete population data are available on monthly income and benefit recipiency, 17 percent of the population received means-tested assistance, as shown in Figure SUM 1. Less than one-third of this group, or about 5 percent of the total population, would be considered "dependent" on welfare under the above definition. Recipiency and dependency rates in 1995 were lower than in 1993 and 1994, but were still higher than they had been in 1987 and 1990. These numbers are consistent with administrative data showing a peak in AFDC caseloads in 1993 and in food stamp caseloads in 1994 and a decrease in both programs since that time. What is not apparent from administrative records, but is shown in these national survey data, is that the dependency rate peaked in 1993, declining over the next two years until it reached 5.1 percent in 1995, close to the same level as in 1992.
|
| Note: Recipiency is defined as receipt of any amount of AFDC, SSI, or food stamps during year. Dependency is defined as having more than 50 percent of annual income from AFDC, SSI and/or food stamps. While only affecting a small number of cases, general assistance income is included within AFDC income. Dependency rates would be lower if adjusted to exclude welfare assistance associated with working. Because full calendar year data for 1995 were not available for all SIPP respondents, 1995 estimates are based on a weighting adjustment to account for those who were not interviewed for the entire year. |
| Source: Unpublished data from the SIPP, 1987, 1990, 1992, and 1993 panels. |
Recipiency and dependency rates are higher for non-Hispanic blacks and Hispanics than for non-Hispanic whites, as shown in Table SUM 1, which shows these rates for various racial and age categories. Recipiency and dependency also are higher for young children than for adults.
| 1987 | 1990 | 1992 | 1993 | 1994 | 1995 | |
|---|---|---|---|---|---|---|
| Recipiency Rates (Receipt of Any Amount of AFDC, Food Stamps, or SSI) |
||||||
| All Persons | 14.9 | 14.1 | 16.9 | 17.0 | 18.0 | 17.0 |
| Racial Categories | ||||||
| Non-Hispanic White | 9.3 | 8.9 | 11.0 | 10.9 | 11.1 | 10.4 |
| Non-Hispanic Black | 40.9 | 36.6 | 41.0 | 41.8 | 43.2 | 40.9 |
| Hispanic | 28.3 | 29.5 | 33.3 | 33.9 | 37.1 | 34.6 |
| Age Categories | ||||||
| Children Ages 0 - 5 | 24.5 | 24.0 | 28.9 | 29.0 | 32.4 | 27.6 |
| Children Ages 6 - 10 | 23.2 | 20.2 | 23.8 | 24.0 | 28.6 | 28.7 |
| Children Ages 11 - 15 | 19.8 | 18.8 | 23.2 | 22.6 | 24.9 | 23.6 |
| Women Ages 16 - 64 | 14.4 | 14.1 | 17.0 | 17.3 | 17.5 | 16.8 |
| Men Ages 16 - 64 | 10.1 | 9.5 | 11.8 | 12.0 | 12.3 | 11.5 |
| Adults Age 65 and over | 13.6 | 12.1 | 12.6 | 12.2 | 12.3 | 12.2 |
| Dependency Rates (More than 50 Percent of Income from Means-Tested Assistance) |
||||||
| All Persons | 4.7 | 4.2 | 4.9 | 5.9 | 5.6 | 5.1 |
| Racial Categories | ||||||
| Non-Hispanic White | 2.2 | 2.1 | 2.4 | 2.8 | 2.6 | 2.3 |
| Non-Hispanic Black | 15.7 | 14.6 | 15.9 | 16.7 | 16.8 | 15.2 |
| Hispanic | 10.9 | 8.3 | 10.5 | 14.2 | 12.9 | 12.2 |
| Age Categories | ||||||
| Children Ages 0 - 5 | 10.0 | 10.3 | 12.2 | 13.3 | 12.5 | 10.6 |
| Children Ages 6 - 10 | 10.1 | 8.5 | 9.5 | 12.3 | 12.0 | 11.6 |
| Children Ages 11 - 15 | 8.0 | 6.4 | 7.5 | 10.5 | 9.3 | 9.1 |
| Women Ages 16 - 64 | 4.6 | 4.6 | 5.0 | 5.8 | 5.5 | 5.2 |
| Men Ages 16 - 64 | 2.0 | 1.5 | 1.9 | 2.7 | 2.5 | 2.3 |
| Adults Age 65 and over | 2.6 | 1.9 | 2.0 | 2.0 | 2.2 | 1.8 |
| Note: Means-tested assistance includes AFDC, SSI and food stamps. While only affecting a small number of cases, general assistance income is included within AFDC income. Dependency rates would be lower if adjusted to exclude welfare assistance associated with working. Because full calendar year data for 1995 were not available for all SIPP respondents, 1995 estimates are based on a weighting adjustment to account for those who were not interviewed for the entire year. | ||||||
| Source: Unpublished data from the SIPP, 1987, 1990, 1992, and 1993 panels. | ||||||
Dependency on assistance also varies depending upon which programs are counted as "welfare programs," as shown in Table SUM 2. Dependency is highest — 5.1 percent — when income from all three programs (AFDC, food stamps, and SSI) is counted, as in the first column of Table SUM 1 (and most of the report). Dependency is lower — 3.7 percent — when counting AFDC/TANF and food stamp benefits only, as in the second column of Table SUM 2. In general, 70 to 75 percent of individuals who are dependent under the standard definition also are dependent under the alternative definition that considers AFDC and food stamps alone (as is done in some measures in this report). Note, however, that the elderly depend more on SSI than on AFDC and food stamps; whereas 1.8 percent of elderly persons are dependent when counting the three major types of means-tested assistance, very few, 0.3 percent, are dependent when the definition is limited to AFDC and food stamps.
In general, non-whites and the very young were more likely to be dependent than other racial and age categories, and they are primarily dependent on AFDC and food stamps. Even in these populations, however, the vast majority of families do not meet the criteria for dependence.
| AFDC, SSI, & Food Stamps |
AFDC & Food Stamps |
SSI only | |
|---|---|---|---|
| All Persons | 5.1 | 3.7 | 0.9 |
| Racial Categories | |||
| Non-Hispanic | 2.3 | 1.6 | 0.5 |
| Non-Hispanic | 15.2 | 10.5 | 2.4 |
| Hispanic | 12.2 | 9.8 | 1.6 |
| Age Categories | |||
| Children Ages 0 - 5 | 10.6 | 10.4 | 0.5 |
| Children Ages 6 - 10 | 11.6 | 8.9 | 0.5 |
| Children Ages 11 - 15 | 9.1 | 6.9 | 0.8 |
| Women Ages 16 - 64 | 5.2 | 3.5 | 1.1 |
| Men Ages 16 - 64 | 2.3 | 1.1 | 0.7 |
| Adults Age 65 and over | 1.8 | 0.3 | 1.3 |
| Note: While only affecting a small number of cases, general assistance income is included within AFDC income. Because full calendar year data for 1995 were not available for all SIPP respondents, 1995 estimates are based on a weighting adjustment to account for those who were not interviewed for the entire year. | |||
| Source: Unpublished data from the SIPP, 1993 panel. | |||
The summary measure of dependence shown in Table SUM 1 focuses on the percentage of income received from means-tested assistance over a one-year time period. It also is important to look at dependency over a longer term perspective, as is done in Table SUM 3, which examines long-term recipiency and long-term dependency among AFDC recipients.
Half (50 percent) of all those who received welfare in 1982 did not receive more than 50 percent of their income from AFDC and food stamp benefits in any of the ten years between 1982 and 1991. About one-quarter (23 percent) were dependent for one to two years, 15 percent for three to five years, and 13 percent for six or more years.
Long-term dependence is rarer than long-term recipiency. Only 4 percent of those who were recipients in 1982, for example, received more than 50 percent of their income from AFDC and food stamps for nine to ten years. This is a smaller percentage than the proportion of recipients that received welfare of any amount for nine to ten years (11 percent). Child recipients have longer spells of welfare receipt and welfare dependence than do recipients in general, as shown in the table.
| All Recipients (in 1982) |
All Child Recipients (0-5 in 1982) |
|
|---|---|---|
| Years of Recipiency, 1982–1991 (Any AFDC Receipt) |
||
| 1 - 2 Years | 47 | 34 |
| 3 - 5 Years | 28 | 29 |
| 6 - 8 Years | 15 | 17 |
| 9 - 10 Years | 11 | 20 |
| Years of Dependency, 1982–1991 (AFDC & Food Stamps, >50% of Income) |
||
| 0 Years | 50 | 34 |
| 1 - 2 Years | 23 | 28 |
| 3 - 5 Years | 15 | 16 |
| 6 - 8 Years | 9 | 13 |
| 9 - 10 Years | 4 | 8 |
| Note: "Any AFDC Receipt" is defined as whether an individual has received any amount of AFDC at any time during the year. "AFDC & Food Stamps, >50% of Income" is defined as whether the sum of an individual's AFDC and food stamp benefits was more than 50% of their yearly income. "0 Years" means that while an individual received means-tested assistance, his or her benefits were not greater than 50 percent of his or her income for any years during the time period. Note that this table shows years of receipt and dependency between 1982 and 1991 and does not take into account years of receipt or dependency that may have occurred before 1982. | ||
| Source: Unpublished data from the PSID, 1983 – 1992. | ||
Changes in dependence may or may not be associated with changes in the level of deprivation, depending on the alternative sources of support found by families who might otherwise be dependent on welfare. To assess the social impacts of any change in dependence, changes in the level of poverty or deprivation also must be considered. One way of measuring deprivation is to look at changes in the level of need over time. Elsewhere in this report, for example, measures of food insecurity and lack of health insurance are presented.
The deprivation measure presented in this chapter, however, focuses directly on changes in the poverty rate, both under the official poverty rate and under expanded measures that take into account taxes and non-cash benefits. These measures also show the degree to which welfare and related programs are effective in moving people out of poverty. The data, shown in Table SUM 4 and its related figure, illustrate two primary points. First, cash welfare and non-cash welfare benefits such as food and housing reduce the number of poor families. Second, under any of the four alternate income measures presented in Table SUM 4, poverty rates have been decreasing since 1993, as economic conditions have improved and policies have promoted and rewarded work. Each of these points is discussed below.
Four different concepts of income are used in Table SUM 4, which shows alternative measures of poverty rates for all persons between 1979 and 1998. (A graph of these data is presented in Figure SUM 4, and a similar analysis is presented for the subset of the population that lives in families with related children under age 18 in Table SUM 5.) The four measures are as follows:
 |
| Source: Congressional Budget Office tabulations. Additional calculations by HHS. |
| 1979 | 1983 | 1986 | 1989 | 1993 | 1995 | 1996 | 1997 | 1998 | |
|---|---|---|---|---|---|---|---|---|---|
| Cash Income plus All Social Insurance | 12.8 | 16.0 | 14.5 | 13.7 | 16.3 | 14.9 | 14.8 | 14.2 | 13.5 |
|
11.6 | 15.2 | 13.6 | 12.8 | 15.1 | 13.8 | 13.7 | 13.3 | 12.7 |
|
9.7 | 13.7 | 12.2 | 11.2 | 13.4 | 12.0 | 12.1 | 11.9 | 11.3 |
|
10.0 | 14.7 | 13.1 | 11.7 | 13.3 | 11.5 | 11.5 | 11.1 | 10.5 |
| Reduction in Poverty Rate | 2.8 | 1.3 | 1.4 | 2.0 | 3.0 | 3.4 | 3.3 | 3.1 | 3.0 |
| Note: The first measure of poverty, labeled cash income plus all social insurance, includes social security but not meanstested cash transfers. Adding means-tested cash transfers yields the official census definition of poverty, the second line in the table. Food and housing benefits may be received either as cash or (more generally) as in-kind benefits, in which case the market value of food and housing benefits is added. EITC refers to the refundable Earned Income Tax Credit, which is always a positive adjustment to income whereas Federal payroll and income taxes are a negative adjustment. The fungible value of Medicare and Medicaid is not included. | |||||||||
| Source: Congressional Budget Office tabulations. Additional calculations by HHS. | |||||||||
The combined effect of means-tested assistance, food and housing benefits, and EITC and taxes was to reduce the poverty rate in 1998 by three percentage points, from 13.5 percent to 10.5 percent. The total reduction in the poverty rate is shown in the final row of Table SUM 4.
As economic conditions improved during the mid-1990s, poverty rates decreased under all four concepts of income. Of particular interest are the poverty rates in 1995, the same year as the dependence rates shown in Table SUM 1, and the poverty rates in 1998, the most recent year for which data are available. In 1995, the final poverty rate was 11.5 percent after adding in non-cash benefits and taxes, a decline from 13.3 percent in 1993. Over the same time period, the dependence measure also declined, from 5.9 percent to 5.1 percent.
More current data indicate that the poverty rate continued to fall between 1995 and 1998, falling to 12.7 percent under the official measure and 10.5 percent under the most comprehensive measure. Data are not yet available on dependence measures for 1998, although administrative data on caseloads indicate a continuing decline in overall receipt of AFDC/TANF and food stamps.
During most of the past two decades, means-tested benefits (including cash assistance, food and housing benefits, and the EITC and other taxes), have caused a net reduction in poverty rates for individuals of about three percentage points. The net effectiveness of these programs in reducing the poverty rate was somewhat lower during the recession of the early 1980s, and was somewhat higher in the mid 1990's, largely due to expansions in the EITC (see Figure SUM 4 and Table SUM 4).
The net effect of all sources of means-tested support (including cash assistance, food and housing benefits, and the EITC and taxes) on the reduction in poverty is higher for persons in families with related children under 18. The gap between poverty rates before and after public assistance has ranged from 3.5 to over 5 percentage points for these individuals in recent years, as shown in Table SUM 5. Again, the net effectiveness of means-tested programs was lower in the mid 1980s and highest in the mid 1990s.
Since the enactment of PRWORA in 1996 and the subsequent implementation of TANF, caseloads for AFDC/TANF and food stamps have fallen dramatically. Although dependency measures as defined in this report are not yet available for the period after PRWORA, available measures on recipiency rates suggest that the legislation has been successful in causing a noticeable fall in dependence on welfare programs. The deprivation measures presented in Tables SUM 4 and 5 suggest that these large caseload declines have been accomplished without observed increases in deprivation. In fact, under the strong economy of the late 1990s, poverty rates are at their lowest levels since 1989. It will be important to continue to track changes in these dependency and deprivation rates over the next several years, to see how they are affected by future changes in economic conditions.
| 1979 | 1981 | 1983 | 1986 | 1989 | 1991 | 1993 | 1995 | 1996 | 1997 | 1998 | |
|---|---|---|---|---|---|---|---|---|---|---|---|
| Cash Income plus All Social Insurance | 14.3 | 17.4 | 19.1 | 17.4 | 16.8 | 18.8 | 20.0 | 18.1 | 17.8 | 17.0 | 16.1 |
|
12.9 | 16.3 | 18.4 | 16.5 | 15.8 | 17.7 | 18.7 | 16.8 | 16.6 | 15.9 | 15.2 |
|
10.2 | 13.9 | 16.5 | 14.6 | 13.6 | 15.3 | 16.4 | 14.3 | 14.4 | 14.1 | 13.2 |
|
10.5 | 15.2 | 17.7 | 15.8 | 14.1 | 15.3 | 15.9 | 13.0 | 12.9 | 12.4 | 11.6 |
| Reduction in Poverty Rate | 3.8 | 2.2 | 1.4 | 1.6 | 3.5 | 3.5 | 4.1 | 5.1 | 4.9 | 4.6 | 4.5 |
| Note: The first measure of poverty, labeled cash income plus all social insurance, includes social security but not meanstested cash transfers. Adding means-tested cash transfers yields the official census definition of poverty, the second line in the table. Food and housing benefits may be received either as cash or (more generally) as in-kind benefits, in which case the market value of food and housing benefits is added. EITC refers to the refundable Earned Income Tax Credit, which is always a positive adjustment to income whereas Federal payroll and income taxes are a negative adjustment. The fungible value of Medicare and Medicaid is not included. | |||||||||||
| Source: Congressional Budget Office tabulations. Additional calculations by HHS. | |||||||||||
For purposes of this report, the Survey of Income and Program Participation (SIPP) has been used the most extensively and is considered the most useful national survey. Its strengths are its longitudinal design, system of monthly accounting, and detail concerning employment, income and participation in federal income-support and related programs. These features make the SIPP particularly effective for capturing the complexities of program dynamics and many of the indicators and predictors, or risk factors, associated with welfare receipt.
The SIPP does not, however, follow families for more than three years. Therefore, the Panel Study of Income Dynamics (PSID) also is used in this report. The PSID collects annual income data, including transfer income, over a long time-period, providing vital data for indicators of long-term welfare receipt, dependence, and deprivation.
Some indicators in this report are based upon the annual March Current Population Survey (CPS), which is available on a more timely basis than the SIPP. The March CPS measures income and poverty over a single annual accounting period, and provides important information regarding child poverty. Finally, the report also draws upon administrative data for the AFDC/TANF, Food Stamp, and SSI programs.
One of the difficulties in preparing this year's annual report has been the challenge of obtaining recent data from the SIPP and the PSID, the two data sources used for most of the report. The most recent SIPP data available at the time of preparation of this year's annual report were 1995 data, collected from the third year of the 1993 SIPP three-year panel. Data from 1995, however, do not reflect many of the dramatic changes in welfare programs that have occurred since enactment of the welfare reform legislation in August 1996. Two more years of SIPP data are expected to be available next year, allowing an update of many indicators through early implementation of the TANF program.
PSID data for the mid- to late-1990s also were not available at the time of updating this report. Instead, the indicators that are based on PSID data cover the same ten-year period (1982-1991) as in last year's volume. Updated PSID analyses will be published in next year's report.
The most recent data are from the CPS and administrative sources. The CPS data are available for calendar year 1998 (and in some cases, March 1999), while administrative data are generally available through fiscal year 1998 (or, for some aggregate caseload statistics, fiscal year 1999). To the extent possible, TANF administrative data are reported in a consistent manner with data from the earlier AFDC program, as noted in the footnotes to the tables in Appendix A. The fact remains that assistance under locally designed TANF programs encompasses a diverse set of cash and non-cash services designed to support families in making a transition to work, and so direct comparisons between AFDC receipt and TANF receipt must be made with caution. This issue also will affect reported data on TANF receipt in national data sets such as the SIPP, once these are available.
Most of the data sources allow analysis of the indicators and predictors of welfare dependence across several age and race/ethnic categories. Where the data are available, statistics are shown for three racial/ethnic groups — non-Hispanic whites, non-Hispanic blacks, and Hispanics. In some instances, however, there are not sufficient data on individuals of Hispanic origin, and so the measures are shown for only two racial/ethnic categories.
A final technical note concerns the unit of analysis and the measurement of benefit receipt. The individual, rather than the family or household, is the unit of analysis for most of the statistics in this report. An individual is considered a recipient of AFDC/TANF or food stamps if he or she lives in a family receiving such benefits. In contrast, the SSI program provides benefits to individuals and couples, and so an individual is only considered an SSI recipient if he or she directly receives such an SSI benefit. All means-tested benefits — AFDC/TANF, food stamps, and SSI — are summed together with earnings and other sources of income for all individuals in a family unit to determine total family income, which is used to determine the poverty status, dependency status, and income levels for all individuals in the family.
1. Further data on poverty and income, as well as current and past annual reports on Indicators of Welfare Dependence, are available online at http://aspe.hhs.gov/hsp/index.htm. This same web page provides access to the annual Trends in the Well-Being of America's Children and Youth, another important data source for indicators of economic, health, and social well-being.
2. The earnings of those in unsubsidized employment would be correctly captured as income from work in national surveys. Any welfare benefits associated with work experience, community service programs or other work activities, however, would be counted as income from welfare in most national surveys, an incorrect classification according to the proposed definition.
[Chapter I in PDF format including all figures and tables]
Top of Page
Contents of Chapter
Main Page of Report
Executive Summary
Introduction
Indicators of Dependence
Predictors and Risk Factors Associated with Welfare
Receipt
Appendix A: Program Data
Appendix B: Additional Non-Marital Birth Data
Home Pages:
Human Services Policy
(HSP)
Assistant Secretary for Planning and Evaluation
(ASPE)
U.S. Department of Health and Human Services
(HHS)
Last updated February 3, 2004