Fox, Burns, The Supplemental Poverty Measure: 2020 (P60-275) (Current Population Reports P60-275, 2021-09)
Summary
Current Population Reports P60-275, "The Supplemental Poverty Measure: 2020," by Liana E. Fox and Kalee Burns, dated September 2021. The report presents estimates of poverty using the official measure and the Supplemental Poverty Measure (SPM), based on "Current Population Survey" supplements. Its highlights state that the overall SPM rate in 2020 was 9.1 percent, the lowest rate since estimates were initially published for 2009, and that this was the first time in the history of the SPM that poverty was lower using the SPM than the official poverty rate of 11.4 percent. They also state that Social Security moved 26.5 million individuals out of poverty and that stimulus payments moved 11.7 million.
Summary drafted by a model from the document's text below and checked by script against that text before publication. It is a navigation aid, not a reading of what the document proves. Where AI is used
Cite this document
“Fox, Burns, The Supplemental Poverty Measure: 2020 (P60-275) (Current Population Reports P60-275, 2021-09).”
Original at www.census.gov. Archived at Pandemic Darlings, https://pandemicdarlings.com/source-documents/fox-burns-the-supplemental-poverty-measure-2020-p60-275-current-popula-a42ca0021e/
, SHA-256 43d7a9f4f67fd67d08114df4a503c92c672fdf0fa40a882f806a03e1e3333a0d.Full text
The Supplemental Poverty Measure: 2020
Current Population Reports
By Liana E. Fox and Kalee Burns
September 2021
P60-275
INTRODUCTION
Figure 1.
Since the publication of the first Supplemental Poverty Measure Rates for Total Population
official U.S. poverty estimates and by Age Group: 2019 and 2020
in the 1960s, researchers and (In percent)
20191 2020
policymakers have continued to
discuss the best approach to mea-
12.6 12.8
sure income and poverty in the 11.8
11.2
United States. Beginning in 2011, 9.7 9.5
9.1 8.8
the U.S. Census Bureau began
publishing the Supplemental
Poverty Measure (SPM), which
extends the official poverty
measure by taking account of
many of the government pro-
grams designed to assist low- All people Under 18 18 to 64 65 years
years years and older
income families and individuals
that are not included in the official
1
The data for 2019 reflect the implementation of revised Supplemental Poverty Measure
methodology. More information is available in the report appendix.
poverty measure. The SPM is pro- Note: Information on confidentiality protection, sampling error, nonsampling error, and
definitions is available at <https://www2.census.gov/programs-surveys/cps
duced with the support of the /techdocs/cpsmar21.pdf>.
U.S. Bureau of Labor Statistics Source: U.S. Census Bureau, Current Population Survey, 2020 and 2021 Annual Social
and Economic Supplements (CPS ASEC).
(BLS), and this report is the
eleventh in the series. This report
presents estimates of the preva-
2.6 percentage points lower to 64, and adults aged 65 and
lence of poverty in the United than the 2019 SPM rate of 11.8 older between 2019 and 2020
States using the official measure (Figure 1).1, 2 (Figures 1 and 2).
and the SPM based on informa-
• SPM rates were down for all • The SPM rate for 2020 was 2.3
tion collected in 2021 and earlier major age categories: children percentage points lower than
Current Population Survey Annual under age 18, adults aged 18 the official poverty rate of 11.4
Social and Economic Supplements percent (Figure 3). This is the
1
Calculated differences here and
(CPS ASEC). throughout this report may differ due to first time in the history of the
rounding. SPM that poverty was lower
HIGHLIGHTS 2
SPM rate reflects methodological
changes from published results in Fox using the SPM than the official
• In 2020, the overall SPM rate (2020). More information is contained in poverty rate.
the appendix of this report.
was 9.1 percent. This was
• The 2020 SPM rate of 9.1 per- This report presents estimates SPM, compares the distribution
cent was the lowest rate since of the prevalence of poverty in of income-to-poverty threshold
estimates were initially pub- the United States, overall and for ratios between the two, and pres-
lished for 2009 (Figure 4). selected demographic groups, ents poverty rates by state. In the
• There were 11 states plus the using the official poverty measure third section, individual compo-
District of Columbia for which and the SPM. 3, 4 The first section nents of the SPM are subtracted
SPM rates were higher than provides detailed information from resources to assess the
official poverty rates, 30 states about changes in SPM rates from marginal impacts of taxes, trans-
with lower rates, and 9 states
2019 to 2020. The second section fers, and necessary expenses on
for which the differences were
presents differences between the poverty rates.
not statistically significant
official poverty measure and the
(Figure 7).
DATA
• Social Security continued to 3
The estimates in this report (which
be the most important anti-
may be shown in text, figures, and tables) This report provides estimates for
are based on responses from a sample of
poverty program, moving the population and may differ from actual calendar year 2020, which coin-
26.5 million individuals out of values because of sampling variability cided with the COVID-19 pan-
or other factors. As a result, apparent
poverty. differences between the estimates for two demic, the end of the economic
or more groups may not be statistically expansion in February 2020, and
• Stimulus payments, enacted as significant. All comparative statements
part of economic relief legisla- have undergone statistical testing and are the 2-month recession that began
tion related to the COVID-19 significant at the 90 percent confidence in March 2020. The data collec-
level, unless otherwise noted. Standard
pandemic, moved 11.7 million errors were calculated using replicate tion period for the 2021 CPS ASEC
individuals out of poverty. weights. Further information about the
occurred about 1 year into the
source and accuracy of the estimates is
Unemployment insurance available at <https://www2.census.gov COVID-19 pandemic and the asso-
benefits, also expanded dur- /programs-surveys/cps/techdocs
ciated public health response. For
/cpsmar21.pdf>.
ing 2020, prevented 5.5 million 4
The Census Bureau reviewed this data details on the effect of COVID-
individuals from falling into product for unauthorized disclosure of
confidential information and approved the
19 on CPS ASEC data collection
poverty (Figure 8).
disclosure avoidance practices applied to in 2021, refer to the text box
this release: CBDRB-FY21-POP001-0206.
THE IMPACT OF THE CORONAVIRUS (COVID-19) PANDEMIC ON THE CURRENT POPULATION
SURVEY ANNUAL SOCIAL AND ECONOMIC SUPPLEMENT (CPS ASEC)
The U.S. Census Bureau administers the CPS ASEC quite returning to the prepandemic trend. While the
each year between February and April by tele- response rate improved, it is important to exam-
phone and in-person interviews, with the majority ine how respondents differ from nonrespondents,
of data collected in March. In 2020, data collection as this difference could affect income and pov-
faced extraordinary circumstances due to the onset erty estimates. Using administrative data, Census
of the COVID-19 pandemic as the Census Bureau Bureau researchers have documented that the
suspended in-person interviews and closed both nonrespondents in both 2020 and 2021 are less
telephone contact centers. The response rate for similar to respondents than in earlier years. Of
the CPS basic household survey was 73 percent in particular interest for the estimates in this report,
March 2020, about 10 percentage points lower than are the differences in median income and educa-
preceding months and the same period in 2019, tional attainment, indicating that respondents in
which were regularly above 80 percent. 2020 and 2021 had relatively higher income and
were more educated than nonrespondents. For
During collection of the 2021 CPS ASEC, for the
more details on how these sample differences
safety of both interviewers and respondents, in-
and the associated nonresponse bias impact
person interviews were only conducted when tele-
income and official poverty estimates, refer to
phone interviews could not be done. In March 2021,
<www.census.gov/newsroom/blogs/research
the response rate for the CPS basic household
-matters/2021/09/pandemic-affect-survey
survey improved to about 76 percent, though not
-response.html>.
2 U.S. Census Bureau
“The Impact of the Coronavirus the SPM, poverty is estimated to differences between the official
(COVID-19) Pandemic on the have been lower using the SPM and supplemental measures. 5
Current Population Survey Annual than using the official poverty
The SPM does not replace the
Social and Economic Supplement definition.
official poverty measure and is not
(CPS ASEC).”
BACKGROUND designed to be used for program
In response to the COVID-19 eligibility or funding distribution.
pandemic, Congress passed After many years of research, The SPM is designed to provide
legislation to aid individuals and analysis, and debate, the information on aggregate levels of
families. This legislation included Interagency Technical Working economic need at a national level
the Coronavirus Aid, Relief, and Group (ITWG) on Developing a or within large subpopulations or
Economic Security Act (CARES Supplemental Poverty Measure areas. As such, the SPM provides
Act), the Coronavirus Response reviewed methods and data an additional macroeconomic
and Relief Supplemental needed for poverty measurement. statistic for further understanding
Appropriations Act (CRRSA Act), The group listed suggestions for a economic well-being, conditions,
and the Families First Coronavirus new measure that would supple- and trends.
Response Act (FFCRA). The ment the current official measure
CARES and CRRSA Acts provided of poverty (ITWG, 2010). The CHANGES IN SPM RATES
households with additional income appendix to this report includes BETWEEN 2019 AND 2020
in the form of stimulus payments detailed descriptions of how these
Figure 2 shows SPM rates for 2019
and tax credits. FFCRA autho- suggestions have been applied to
and 2020.6 In 2020, the percent-
rized expansions in Supplemental the SPM. In 2020, several meth-
age of people who were poor as
Nutritional Assistance Program odological improvements were
estimated using the SPM was 9.1
(SNAP) benefits. The SPM is a applied to both the resources and
percent, compared to 11.8 percent
post-tax and transfer poverty threshold estimation to the SPM.
in 2019, a decline of 2.6 percent-
measure, so it captures expansion Details of those improvements are
age points. The poverty rate
of unemployment insurance (like available in the report appendix.
declined for all groups shown in
the official poverty measure), but All 2019 and 2020 estimates in
Figure 2 between 2019 and 2020.
also includes stimulus payments this report reflect implementation
and expansions to SNAP that are of the revised SPM methodology. 5
Thresholds for the SPM are produced
not included in the official poverty The “Poverty Measure Concepts: by the BLS Division of Price and Index
Number Research and presented for 2019
definition. As a result, in 2020, Official and Supplemental” table
and 2020 in Appendix Table 3.
for the first time in the history of summarizes the most important 6
Appendix Table 1 contains rates for a
more extensive list of demographic groups.
POVERTY MEASURE CONCEPTS: OFFICIAL AND SUPPLEMENTAL
Concept Official Poverty Measure Supplemental Poverty Measure
Measurement Families (individuals related by Resource units (official family definition plus any coresident unrelated
Units birth, marriage, or adoption) or children, foster children, and unmarried partners and their relatives)
unrelated individuals. or unrelated individuals (who are not otherwise included in the family
definition).
Poverty Three times the cost of a Based on expenditures of food, clothing, shelter, and utilities (FCSU).
Threshold minimum food diet in 1963.
Threshold Vary by family size, composition, Vary by family size, composition, and tenure, with geographic
Adjustments and age of householder. adjustments for differences in housing costs.
Updating Consumer Price Index for All Five-year moving average of expenditures on FCSU, lagged 1 year.
Thresholds Urban Consumers: all items.
Resource Gross before-tax cash income. Sum of cash income, plus noncash benefits that resource units
Measure can use to meet their FCSU needs, minus taxes (or plus tax credits),
work expenses, medical expenses, and child support paid to another
household.
U.S. Census Bureau 3
Figure 2.
Change in Percentage of People in Poverty Using the Supplemental Poverty Measure:
2019 to 2020
Decrease Increase
( Statistically different from zero)
20191 2020
All people 11.8 9.1 -2.6
Sex
Male 11.1 8.6 -2.4
Female 12.4 9.6 -2.8
Age
Under 18 years 12.6 9.7 -2.9
18 to 64 years 11.2 8.8 -2.4
65 years and older 12.8 9.5 -3.3
Type of Unit
Married couple 6.8 5.0 -1.8
Cohabiting partners 12.1 8.5 -3.5
Female reference person 23.8 18.2 -5.6
Male reference person 14.2 11.7 -2.5
Unrelated individuals 21.6 17.5 -4.1
Race and Hispanic Origin
White 10.5 8.1 -2.4
White, not Hispanic 8.2 6.5 -1.7
Black 18.9 14.6 -4.3
Asian 11.3 8.8 -2.5
Hispanic (any race) 18.8 14.0 -4.9
Educational Attainment2
No high school diploma 27.7 20.3 -7.4
High school, no college 13.9 11.1 -2.9
Some college 9.5 7.3 -2.1
Bachelor’s degree or higher 5.6 4.3 -1.2
Work Experience3
All workers 6.8 4.6 -2.2
Worked full-time, year-round 3.9 1.9 -2.0
Less than full-time, year-round 14.6 9.8 -4.7
Did not work at least 1 week 27.0 22.9 -4.1
0 5 10 15 20 25 30
Percent
1
The data for 2019 reflect the implementation of revised Supplemental Poverty Measure (SPM) methodology. More information is available in
the report appendix.
2
Population limited to individuals aged 25 and older. In 2020, the overall SPM rate for this group was 8.4 percent.
3
Population limited to individuals aged 18 to 64. In 2020, the overall SPM rate for this group was 8.8 percent.
Notes: Statistically significant indicates the change is statistically different from zero at the 90 percent confidence level. Details may not sum
to totals due to rounding. More details are available in Appendix Table 1. More information on confidentiality protection, sampling error,
nonsampling error, and definitions is avaiable at <https://www2.census.gov/programs-surveys/cps/techdocs/cpsmar21.pdf>.
Source: U.S. Census Bureau, Current Population Survey, 2020 and 2021 Annual Social and Economic Supplements (CPS ASEC).
4 U.S. Census Bureau
POVERTY ESTIMATES FOR Census Bureau estimates for the Comparisons over time should be
2020: OFFICIAL AND SPM SPM are available back to 2009.9 made with caution.11
Since the SPM’s initial produc-
Using the SPM definition of pov-
tion, the SPM rate has been higher 11
This report provides SPM and official
erty, Figure 3 shows that 9.1 per- poverty estimates from 2009 to 2020.
than the official poverty rate. This
cent of people were poor, lower However, it is important to be aware that
marks the first year that the offi- the CPS ASEC is updated periodically to
than the 11.4 percent using the improve data quality. These improvements
cial poverty rate is higher than the include changes to survey design such as
official definition of poverty with
SPM rate. Figures 4 and 5 present sampling and survey instrument changes,
the comparable universe.7, 8 While as well as changes to data processing
estimates for the official measure such as weighting and data imputation
the SPM rates were lower than
and the SPM from 2009 to 2020. methods. When feasible, the Census
official poverty rates for most Bureau provides data users with resources
The charts show two values for that allow them to evaluate the impact of
groups, the SPM shows higher
2013, one using the traditional these survey changes across years. Most
than official poverty rates for indi- recently, the 2014 CPS ASEC introduced
income questions comparable to new income questions, new relationship
viduals 65 years and older, individ-
estimates from 2009 to 2012, and categories were phased in over the 2015
uals living in a male reference unit, and 2016 CPS ASEC, and the 2019 CPS
the second using the redesigned ASEC reflects the implementation of
those with a bachelor’s degree or an updated data processing system.
income questions used for this
higher, and people who worked Given these changes over time, historical
report and comparable to the comparisons should be made with caution.
full-time, year-round (Figure 3). In this report, 2020 SPM estimates are
2014 to 2017 estimates presented
Official and SPM poverty rates for compared to published estimates for
here. Additionally, there are two earlier years when the questionnaire and
Asian individuals were not statisti- processing system changes did not result
sets of SPM numbers for 2017,
cally different. in statistically significant differences.
with one set using the legacy data When survey changes did have statistically
significant impacts on income or poverty
7
Since the CPS ASEC does not ask processing system and the other estimates, comparisons are made by
income questions for individuals under the using the updated processing sys- adjusting historical published estimates
age of 15, all unrelated individuals under the to approximate the magnitude of these
age of 15 are excluded from the universe tem. Finally, there are two sets for impacts. More details on the adjustment
for official poverty calculations in Shrider, SPM rates for 2019 to reflect the used for these comparisons is available at
Kollar, Chen, and Semega (2021). However, <www.census.gov/library/stories/2019/09
these individuals are included in the official revised SPM methodology.10 /us-median-household-income-not
poverty universe for this report and are -significantly-different-from-2017.html>.
assigned the official poverty status of the
householder. The appendix contains more
9
SPM estimates from 1967 to 2012 are
details. available in Fox et al. (2015).
8
Appendix Table 2 contains rates for a
10
Research at BLS and the
more extensive list of demographic groups. Census Bureau is ongoing to extend
the methodological improvements
implemented this year to historical
estimates. Revised estimates and public-
use datasets will be available in the
future to allow for historically consistent
comparisons.
U.S. Census Bureau 5
Figure 3.
Percentage of People in Poverty by Different Poverty Measures: 2020
Official1 SPM Difference Official1 SPM
All people 11.4 9.1 *-2.3
Sex
Male 10.2 8.6 *-1.6
Female 12.6 9.6 *-3.0
Age
Under 18 years 16.0 9.7 *-6.3
18 to 64 years 10.4 8.8 *-1.6
65 years and older 9.0 9.5 *0.5
Type of Unit
Married couple 5.3 5.0 *-0.2
Cohabiting partners 23.1 8.5 *-14.5
Female reference person 23.9 18.2 *-5.7
Male reference person 10.5 11.7 *1.2
Unrelated individuals 18.9 17.5 *-1.4
Race and Hispanic Origin
White 10.1 8.1 *-2.0
White, not Hispanic 8.2 6.5 *-1.7
Black 19.5 14.6 *-4.9
Asian 8.1 8.8 0.7
Hispanic (any race) 17.0 14.0 *-3.0
Educational Attainment2
No high school diploma 24.7 20.3 *-4.4
High school, no college 13.2 11.1 *-2.1
Some college 8.4 7.3 *-1.1
Bachelor's degree or higher 4.0 4.3 *0.3
Work Experience3
All workers 5.0 4.6 *-0.4
Worked full-time, year-round 1.6 1.9 *0.3
Less than full-time, year-round 11.3 9.8 *-1.5
Did not work at least 1 week 28.8 22.9 *-5.9
0 5 10 15 20 25 30 35
Percent
* An asterisk preceding an estimate indicates change is statistically different from zero at the 90 percent confidence level.
1
Includes unrelated individuals under the age of 15.
2
Population limited to individuals aged 25 and older. In 2020, the overall Supplemental Poverty Measure (SPM) rate for this
group was 8.4 percent.
3
Population limited to individuals aged 18 to 64. In 2020, the overall SPM rate for this group was 8.8 percent.
Notes: Details may not sum to totals due to rounding. More details are available in Appendix Table 2. More information on
confidentiality protection, sampling error, nonsampling error, and definitions is avaiable at <https://www2.census.gov
/programs-surveys/cps/techdocs/cpsmar21.pdf>.
Source: U.S. Census Bureau, Current Population Survey, 2021 Annual Social and Economic Supplement (CPS ASEC).
6 U.S. Census Bureau
Figure 4.
Poverty Rates Using the Official and Supplemental Poverty Measures: 2009 to 2020
Recession
Percent
18
16 15.1 SPM
14 14.5
Official1
12
11.4
10
9.1
8
6
4
2
0
2009 '10 '11 '12 '13 '14 '15 '16 '17 '18 '19 2020
1
Includes unrelated individuals under the age of 15.
Notes: The Supplemental Poverty Measure (SPM) estimates for 2019 and 2020 reflect the implementation of revised SPM methodology. More
information is available in the report appendix. The data for 2017 and beyond reflect the implementation of an updated processing system. The
data for 2013 and beyond reflect the implementation of the redesigned income questions. More information on confidentiality protection,
sampling error, nonsampling error, and definitions is available at <https://www2.census.gov/programs-surveys/cps/techdocs/cpsmar21.pdf>.
Source: U.S. Census Bureau, Current Population Survey, 2010 to 2021 Annual Social and Economic Supplements (CPS ASEC).
Figure 4 shows the official mea- Figure 5 shows the poverty rate the first time in the history of the
sure (with the comparable uni- using both measures for three SPM. The gap between official and
verse) and the SPM since 2009. major age groups. While the SPM SPM rates for individuals 65 and
The SPM ranged from 0.6 to 1.6 rate for children declined between older declined from 3.9 percent-
percentage points higher than 2019 and 2020, the official pov- age points in 2019 to 0.5 percent-
the official measure through 2019. erty rate increased over the same age points in 2020. Even after
In contrast, the 2020 SPM was period, widening the gap between accounting for breaks in series,
2.3 percentage points lower than the two measures from 1.8 per- SPM rates for each major age
the official measure. SPM rates in centage points to 6.3 percentage group in 2020 were at their lowest
2020 were at their lowest level points. SPM rates for individuals level since 2009.
since the first year for which the aged 18 to 64 were also lower than
Census Bureau published SPM official poverty rates in 2020, for
estimates, even after adjusting for
the breaks in series.
U.S. Census Bureau 7
Figure 5.
Poverty Rates Using the Official and Supplemental Poverty Measures by Age Group:
2009 to 2020
Recession
Under 18 years
Percent
25
21.2 Official1 children
20
16.0
17.0
15
SPM children
10
9.7
5
0
2009 '10 '11 '12 '13 '14 '15 '16 '17 '18 '19 2020
18 to 64 years
Percent
25
20
SPM 18–64
14.4
15
13.0 10.4
Official1 18–64
10
8.8
5
0
2009 '10 '11 '12 '13 '14 '15 '16 '17 '18 '19 2020
Percent
65 years and older
25
20
14.9 SPM 65+
15
9.5
10
8.9 Official1 65+ 9.0
5
0
2009 '10 '11 '12 '13 '14 '15 '16 '17 '18 '19 2020
1
Includes unrelated individuals under the age of 15.
Notes: The Supplemental Poverty Mesaure (SPM) estimates for 2019 and 2020 reflect the implementation of revised SPM methodology.
More information is available in the report appendix. The data for 2017 and beyond reflect the implementation of an updated processing system.
The data for 2013 and beyond reflect the implementation of the redesigned income questions. More information on confidentiality protection,
sampling error, nonsampling error, and definitions is available at <https://www2.census.gov/programs-surveys/cps/techdocs/cpsmar21.pdf>.
Source: U.S. Census Bureau, Current Population Survey, 2010 to 2021 Annual Social and Economic Supplements (CPS ASEC).
8 U.S. Census Bureau
DISTRIBUTION OF INCOME- their poverty threshold using the children. When these policies and
TO-POVERTY RATIOS: SPM compared to the official mea- programs are taken into account
OFFICIAL AND SPM sure. Including targeted noncash in the SPM, the share of children
Comparing the distribution of benefits and tax credits/stimulus with income below 50 percent
pretax cash income with that of payments, and subtracting neces- of their poverty thresholds is
SPM resources also allows an sary expenses, reduced the per- lower using the SPM definition of
examination of the effect of taxes centage of the population in the resources (2.9 percent) than with
and noncash transfers across lowest category for children under the official measure (7.6 percent).
the income/resource distribu- the age of 18 and adults aged 18 to
At the other end of the distribu-
tion. Figure 6 shows the percent 64. The share of individuals aged
tion, relative to the official mea-
distribution of income-to-poverty 65 and older with income below
sure, the SPM shows a smaller
ratio categories for all people and half their poverty threshold was
percentage of the population
by major age category. Dividing not statistically different between
with income four or more times
income by the respective poverty SPM and the official measure.
the poverty threshold relative
threshold controls income by unit Many of the noncash benefits to the official measure. The SPM
size and composition. Appendix included in the SPM, such as resource measure subtracts
Table 4 shows the distribution of SNAP, Program for Women, taxes—compared with the
income-to-poverty ratios for vari- Infants, and Children (WIC), official measure, that does not—
ous groups in 2019 and 2020. and school meals, as well as tax bringing down the percentage of
Overall, the comparison shows credits and stimulus payments are people with income in the highest
that a smaller share of the popu- targeted at families with children category.
lation had incomes below half of or increase in benefit value with
Figure 6.
Distribution of People by Income-to-Poverty Ratios: 2020
(In percent)
Less than 0.50 0.50 to 0.99 1.00 to 1.99 2.00 to 3.99 4.00 or more
Total population
Official1 5.5 5.9 16.1 28.3 44.1
SPM 3.3 5.9 26.4 38.4 26.0
Under 18 years
Official1 7.6 8.4 19.7 28.7 35.5
SPM 2.9 6.8 32.1 39.0 19.1
18 to 64 years
Official1 5.2 5.3 13.8 27.8 48.0
SPM 3.3 5.6 24.3 39.2 27.8
65 years and older
Official1 3.9 5.1 19.7 29.7 41.6
SPM 3.8 5.7 26.8 35.1 28.6
1
Includes unrelated individuals under the age of 15.
Notes: Details may not sum to totals due to rounding. More information on confidentiality protection, sampling error, nonsampling error,
and definitions is available at <https://www2.census.gov/programs-surveys/cps/techdocs/cpsmar21.pdf>.
Source: U.S. Census Bureau, Current Population Survey, 2021 Annual Social and Economic Supplement (CPS ASEC).
U.S. Census Bureau 9
Another notable difference the United States from 2018–2020 Dakota, Tennessee, Texas, West
between the distributions using was 11.2 percent with the offi- Virginia, and Wisconsin. Lower
these two measures was the cial measure and 11.2 percent SPM rates could occur due to
larger number of individuals with using the SPM. The difference in lower thresholds reflecting lower
income-to-threshold ratios in the these rates was not statistically housing costs, a different mix of
middle categories, 1.00 to 3.99, significant. housing tenure, or more generous
using the SPM. Since the effect noncash benefits.
While the 3-year average national
of taxes and transfers is often to
SPM rate was not statistically dif- The nine states that were not
move income from the extremes
ferent from the official rate, there statistically different under
of the distribution to the center of
was variation by geographic area. the two measures include
the distribution, that is, from the
Figure 7 shows the United States Alaska, Connecticut, Illinois,
very bottom with targeted trans-
divided into three categories Nevada, Oregon, Utah, Vermont,
fers or from the very top via taxes
by state. States where the SPM Washington, and Wyoming.
and other expenses, the increase
rates were higher than official are Details are provided in Appendix
in the size of these middle catego-
shaded blue; states where SPM Table 5.
ries is to be expected.
was lower than official are shaded
Appendix Table 4 shows similar orange; and states where the THE SPM AND THE EFFECT
calculations by race and ethnic- differences in the rates were not OF CASH AND NONCASH
ity. For all groups, except Asians, statistically significant are gray. TRANSFERS, TAXES, AND
OTHER NONDISCRETIONARY
smaller percentages had income
The 11 states for which the SPM EXPENSES
below half of their poverty thresh-
rates were higher than the offi-
olds when using the SPM com- This section moves away from
cial poverty rates were California,
pared with the official measure. comparing the SPM with the
Colorado, Delaware, Florida,
The share of Asians with income official measure and looks only
Hawaii, Maryland, Massachusetts,
below half of their poverty thresh- at the SPM. This analysis allows
New Hampshire, New Jersey, New
olds in the SPM was not statisti- one to gauge the effects of taxes
York, and Virginia. The SPM rate
cally different than the share and transfers and other necessary
for the District of Columbia was
below half in the official measure. expenses using the SPM as a mea-
also higher. Higher SPM rates by
sure of economic well-being.
POVERTY RATES BY STATE: state may occur for many rea-
OFFICIAL AND SPM sons. Geographic adjustments for Income used for estimating the
housing costs, as well as differ- official poverty measure includes
To create state-level estimates cash benefits from the govern-
ent mixes of housing tenure, may
using the CPS ASEC, the Census ment (e.g., Social Security, unem-
result in higher SPM thresholds.
Bureau recommends using 3-year ployment insurance benefits,
Higher nondiscretionary expenses,
averages for additional statistical public assistance benefits, and
such as taxes or medical
reliability.12, 13 Appendix Table 5 workers’ compensation benefits),
expenses, may also drive higher
shows 3-year averages of poverty but does not take into account
SPM rates.
rates by poverty measure for the taxes or noncash benefits aimed
United States and each state. The The 30 states for which SPM rates
at improving the economic situ-
3-year average poverty rate for were lower than the official pov-
ation of the population. The SPM
erty rates were Alabama, Arizona,
incorporates all of these elements,
12
The Census Bureau recommends using Arkansas, Georgia, Idaho, Indiana,
the American Community Survey (ACS) for
adding cash benefits, noncash
Iowa, Kansas, Kentucky, Louisiana,
state-level poverty estimates. In 2020, a transfers, and stimulus payments,
working paper detailing a methodology for Maine, Michigan, Minnesota,
implementing the SPM in the ACS, as well
while subtracting necessary
Mississippi, Missouri, Montana,
as research data extracts and tables for expenses such as taxes, medical
2009–2018 were released. More information Nebraska, New Mexico, North
is available in Fox, Glassman, and Pacas expenses, and expenses related
Carolina, North Dakota, Ohio,
(2020). to work. An important contribu-
13
The data for 2019 and 2020 reflect Oklahoma, Pennsylvania, Rhode
the implementation of a revised SPM tion of the SPM is that it allows us
Island, South Carolina, South
methodology. The estimates for 2018 to gauge the potential magnitude
reflect the previous methodology.
10 U.S. Census Bureau
AK
Figure 7.
Difference in Poverty Rates by State Using the Official and Supplemental
Poverty Measures: 3-Year Average 2018 to 2020
0 500 Miles
WA
ME
MT ND
OR MN VT
ID NH
SD WI NY MA
MI
WY CT
RI
IA PA
NE
NV OH NJ
MD
UT IL IN !!
! !
!
!
DE
WV
!
CA
! !
CO
! !
!
!
!
! !
! !
VA
!
!
!
KS
! !
MO
!
!
KY
!
DC
!
!
!
NC
TN
OK
AZ AR
NM SC
MS AL GA SPM higher than official1
Not statistically different
TX
LA SPM lower than official1
FL
0 100 Miles
HI
0 100 Miles
1
Includes unrelated individuals under the age of 15.
Notes: The data for 2019 and 2020 reflect the implementation of revised SPM methodology. More information is available in the report
appendix. Information on confidentiality protection, sampling error, nonsampling error, and definitions is available at <https://www2.census
.gov/programs-surveys/cps/techdocs/cpsmar21.pdf>.
Source: U.S. Census Bureau, Current Population Survey, 2019 to 2021 Annual Social and Economic Supplements (CPS ASEC).
of the effect of tax credits and included only in the SPM. This payments had the largest impacts,
transfers in alleviating poverty. We allows us to examine the effects of preventing 26.5 million and 11.7
can also examine the effects of government transfers on poverty million individuals, respectively,
nondiscretionary expenses such estimates. Since child support from falling into poverty. Medical
as work and medical expenses. paid is subtracted from income, expenses were the largest con-
we also examine the effect of child tributor to increasing the number
Figure 8 shows the effect that
support received on alleviating of individuals in poverty.
various additions and subtrac-
poverty. Child support payments
tions had on the number of people Appendix Table 6 shows the effect
received are counted as income
who would have been consid- that various additions and sub-
in both the official measure and
ered poor in 2020, holding all tractions had on the SPM rate in
the SPM (but child support paid is
else the same and assuming no 2019 and 2020, holding all else the
only deducted in the SPM).
behavioral changes. Additions same and assuming no behavioral
and subtractions are shown for Figure 8 allows us to compare the changes. Appendix Table 7 shows
the total population and for three effect of transfers, both cash and the same set of additions and sub-
age groups. Additions shown in noncash, and nondiscretionary tractions but shows the number of
the figure include cash benefits, expenses on numbers of individu- people affected by removing each
also included in the official mea- als in poverty, all else equal. Social element from the SPM, rather than
sure, as well as noncash benefits Security transfers and stimulus the change in the SPM rate.
U.S. Census Bureau 11
Removing one item from the resources, 11.7 million fewer peo- In comparison to 2019, the 2020
calculation of SPM resources and ple were considered poor, all else antipoverty impacts of refundable
recalculating poverty rates shows, constant. On the other hand, when tax credits, school lunch, and WIC
for example, that Social Security the SPM subtracts amounts paid all decreased (Appendix Table
benefits decreased the SPM rate for child support, income and pay- 7).15 The antipoverty impacts of
by 8.1 percentage points, from 17.3 roll taxes, work-related expenses, unemployment insurance and
percent to 9.1 percent (Appendix and medical expenses, the num- Temporary Assistance for
Table 6).14 This means that with ber and percentage in poverty Needy Families (TANF)/general
Social Security benefits, 26.5 were higher. When subtracting 15
In 2020, pandemic electronic benefits
million fewer people were living medical expenses from income, transfer (P-EBT) benefits were distributed
to free and reduced-price school lunch
below the poverty line (Figure the SPM rate was 1.5 percentage recipients. SNAP recipients had this benefit
8 and Appendix Table 7). When points higher. In numbers, 5.0 mil- added directly to their SNAP EBT cards.
Due to likely comingled reporting of P-EBT
including stimulus payments in lion more people were classified benefits as part of SNAP values, these
as poor. programs are reported both jointly and
14
Details do not sum to totals due to separately this year. See report appendix
rounding. for details.
Figure 8.
Change in Number of People in Poverty After Including Each Element: 2020
(In millions)
Under 18 years 18 to 64 years 65 years and older
Social Security –26.5
Economic Impact/stimulus1 –11.7
Unemployment insurance –5.5
Refundable tax credits2 –5.3
SNAP + school lunch –3.2
SNAP –2.9
SSI –2.7
Housing subsidies –2.4
Child support received –0.5
TANF/general assistance –0.5
School lunch –0.3
Workers' compensation –0.2
Energy assistance –0.1
WIC –0.1
Child support paid 0.2
Federal income tax 0.5
FICA 2.0
Work expenses 2.5
Medical expenses 5.0
1
Includes the first two rounds of stimulus payments. Additional details available in the report appendix.
2
Refundable tax credits do not include stimulus payments.
Notes: SNAP: Supplemental Nutrition Assistance Program; SSI: Supplemental Security Income; TANF: Temporary Assistance
for Needy Families; WIC: Special Supplemental Nutrition Program for Women, Infants, and Children; FICA: Federal Insurance
Contributions Act. More information on confidentiality protection, sampling error, nonsampling error, and definitions is available at
<https://www2.census.gov/programs-surveys/cps/techdocs/cpsmar21.pdf>.
Source: U.S. Census Bureau, Current Population Survey, 2021 Annual Social and Economic Supplement (CPS ASEC).
12 U.S. Census Bureau
assistance increased in 2020 income, while Social Security As such, there are lower percent-
compared with 2019. Conversely, benefits lowered poverty rates ages of the SPM poverty popula-
federal income taxes, Federal by 33.1 percentage points for this tions in the very high and very
Insurance Contributions Act group, lifting 18.5 million individu- low resource categories than we
(FICA), work expenses, and medi- als above the poverty line. find using the official measure.
cal expenses pushed fewer indi- Since noncash benefits help those
viduals into poverty in 2020 than SUMMARY in extreme poverty, there were
in 2019. In the absence of stimulus This report provides estimates lower percentages of individu-
payments, which were authorized of poverty using the SPM for the als with resources below half the
in 2020 in response to the COVID- United States. The results illustrate SPM threshold for most groups.
19 pandemic and the resulting differences between the official In addition, the effect of benefits
economic disruption, the poverty measure of poverty and a pov- received from each program and
rate in 2020 would have increased erty measure that takes account taxes and other nondiscretion-
from 11.8 percent in 2019 to 12.7 of noncash benefits received ary expenses on SPM rates were
percent in 2020.16 Instead, poverty by families and nondiscretion- examined.
declined to 9.1 percent in 2020. ary expenses that they must pay.
COMMENTS
Appendix Tables 6 and 7 also The SPM also employs a poverty
show effects of individual ele- threshold that is updated by the The Census Bureau welcomes the
ments for different age groups. BLS with information on expen- comments and advice of data and
In 2020, accounting for stimu- ditures for food, clothing, shelter, report users. If you have sugges-
lus payments resulted in a 4.5 and utilities. In 2020, multiple tions or comments on this report,
percentage-point decrease in the pieces of legislation (CARES Act please write to:
child poverty rate, representing and CRRSA Act) were passed in
Liana E. Fox
3.2 million children prevented response to the COVID-19 pan-
Chief, Poverty Statistics Branch
from falling into poverty by the demic that provided households
Social, Economic, and Housing
inclusion of these payments. with additional income in the
Statistics Division
Subtracting medical expenses, form of stimulus payments and
U.S. Census Bureau
such as contributions toward the expanded unemployment, SNAP,
and pandemic electronic benefits Washington, DC 20233–8500
cost of medical care and health
transfer (P-EBT) benefits. As a Or e-mail
insurance premiums, from the
result, in 2020, for the first time <Liana.E.Fox@census.gov>.
income of families with children
resulted in a child poverty rate in the history of the SPM, poverty
IN MEMORIAM
1.2 percentage points higher. For is estimated to be lower using the
SPM than using the official pov- In memory of Mark Levitan,
the group aged 65 and older, SPM
erty definition. a champion of poverty
rates increased by 2.7 percent-
measurement.
age points with the inclusion of The SPM allows us to examine the
medical expense deductions from effect of taxes, noncash transfers,
and necessary expenses on the
16
Additional analysis of the effect population in and near poverty.
of stimulus payments on poverty rates
is available at <www.census.gov
/america-counts/impact-of-stimulus>.
U.S. Census Bureau 13
REFERENCES Citro, Constance F. and Robert , “The Supplemental
T. Michael (eds.), “Measuring Poverty Measure: 2017,”
Many of the working papers
Poverty: A New Approach,” Current Population Reports,
listed below are available at
National Academy Press, P60-265, U.S. Census Bureau,
<www.census.gov/topics
Washington, DC, 1995, available Washington, DC, September
/income-poverty/supplemental
at <www.census.gov/library 2018, available at
-poverty-measure/library
/publications/1995/demo <www.census.gov/content
/working-papers.html> or
/citro-01.html>. /dam/Census/library
<https://stats.bls.gov/pir
Edwards, Ashley, Brian McKenzie, /publications/2018/demo
/spmhome.htm>.
and Kathleen Short, /p60-265.pdf>.
Betson, David, “Is Everything “Work-Related Expenses in , “The Supplemental
Relative? The Role of the Supplemental Poverty Poverty Measure: 2016,”
Equivalence Scales in Poverty Measure,” U.S. Census Bureau, Current Population Reports,
Measurement,” University Washington, DC, January 2014, P60-261, U.S. Census Bureau,
of Notre Dame, Poverty available at <www.census.gov Washington, DC, September
Measurement Working /content/dam/Census/library 2017, available at
Paper, 1996, available at /working-papers/2014/demo <www.census.gov/content
<https://aspe.hhs.gov /sgeworkexpense.pdf>. /dam/Census/library
/system/files/pdf/106776 /publications/2017/demo
Fox, Liana, “Revising Poverty
/escale.pdf>. /p60-261.pdf>.
Assignments of Unrelated
Bridges, Benjamin and Robert V. Children under Age 15 in the , Brian Glassman, and Jose
Gesumaria, “The Supplemental Supplemental Poverty Measure Pacas, “The Supplemental
Poverty Measure and the Report,” SEHSD Working Paper Poverty Measure using
Aged: How and Why the SPM Number 2017–42, U.S. Census the American Community
and Official Poverty Estimates Bureau, Washington, DC, 2017, Survey,” SEHSD Working
Differ,” Social Security Bulletin, available at <www.census.gov Paper Number 2020-09,
Vol. 73, No. 4, 2013, available at /content/dam/Census/library U.S. Census Bureau,
<www.ssa.gov/policy/docs/ssb /working-papers/2017/demo Washington, DC, 2020, avail-
/v73n4/v73n4p49.html>. /SEHSD-WP2017-42.pdf>. able at <www.census.gov
Bureau of Labor Statistics , “The Supplemental /library/working-papers/2020
(BLS), “2012 Supplemental Poverty Measure: 2019,” /demo/SEHSD-WP2020-09
Poverty Measure Thresholds,” Current Population Reports, .html>.
Washington, DC, September 10, P60-272, U.S. Census Bureau, , Christopher Wimer, Irwin
2013, available at <www.bls.gov Washington, DC, September Garfinkel, Neeraj Kaushal, and
/pir/spm/spm_thresholds_2012 2020, available at Jane Waldfogel, “Waging War
.htm>. <www.census.gov/library on Poverty: Poverty Trends
Caswell, Kyle and Kathleen /publications/2020/demo Using a Historical Supplemental
Short, “Medical Out-of-Pocket /p60-272.html>. Poverty Measure,” Journal
Spending of the Uninsured: , “The Supplemental of Policy Analysis and
Differential Spending and Poverty Measure: 2018,” Management, Vol. 34, 2015,
the Supplemental Poverty Current Population Reports, pp. 567–592, available at
Measure,” SEHSD working P60-268, U.S. Census Bureau, <www.ncbi.nlm.nih.gov/pmc
paper Number 2011-24, U.S. Washington, DC, September /articles/PMC4559281/>.
Census Bureau, Washington, 2019, available at
DC, 2011, presented at the Joint <www.census.gov/content
Statistical Meetings, Miami, FL, /dam/Census/library
August 2011. /publications/2019/demo
/p60-268.pdf>.
14 U.S. Census Bureau
and Thesia Garner, Johnson, Paul, Trudi Renwick, , Eric Figueroa, and Bettina
“Moving to the Median and and Kathleen Short, Aten, “Supplemental Poverty
Expanding the Estimation “Estimating the Value of Measure: A Comparison of
Sample: The Case for Changing Federal Housing Assistance Geographic Adjustments
the Expenditures Underlying for the Supplemental Poverty with Regional Price Parities
SPM Thresholds,” SEHSD Measure,” SEHSD Working vs. Median Rents From the
Working Paper Number Paper 2010–13, U.S. Census American Community Survey:
2018–02, U.S. Census Bureau, Bureau, Washington, DC, An Update,” SEHSD Working
Washington, DC, 2018, available July 2011, available at Paper 2017-36, U.S. Census
at <www.census.gov <www.census.gov/library Bureau, Washington, DC,
/library/working-papers /working-papers/2010 July 2017, available at
/2018/demo/SEHSD /demo/SEHSD-WP2010-13 <www.census.gov/library
-WP2018-02.html>. .html>. /working-papers/2017/demo
Garner, Thesia I., Marisa Mohanty, Abinash, Ashley /SEHSD-WP2017-36.html>.
Gudrais, and Kathleen Short, Edwards, and Liana Fox, and Liana Fox, “The
“Supplemental Poverty “Measuring the Cost of Supplemental Poverty Measure:
Measure Thresholds and Employment: Work-Related 2015,” Current Population
Noncash Benefits,” presented Expenses in the Supplemental Reports, P60-258, U.S. Census
at The Supplemental Poverty Poverty Measure,” SEHSD Bureau, Washington, DC,
Measure Workshop, U.S. Bureau Working Paper Number September 2016, available
of Labor Statistics, Washington, 2017–43, U.S. Census Bureau, at <www.census.gov
DC, April 2016, available at Washington, DC, 2017, available /content/dam/Census/library
<www.bls.gov/pir/spm at <www.census.gov/library /publications/2016/demo
/smp-thresholds-and-noncash /working-papers/2017/demo /p60-258.pdf>.
-benefits-brookings-paper-4-16 /SEHSD-WP2017-43.html>. Short, Kathleen, “Experimental
.pdf>. Renwick, Trudi, “Geographic Poverty Measures: 1999,”
ITWG, “Observations From Adjustments of Supplemental Current Population Reports,
the Interagency Technical Poverty Measure Thresholds: P60-216, U.S. Census Bureau,
Working Group on Developing Using the American Community Washington, DC, 2001, available
a Supplemental Poverty Survey 5-Year Data on Housing at <www.census.gov/library
Measure,” March 2010, Costs,” SEHSD Working Paper /publications/2001/demo
available at <www.census.gov Number 2011–21, U.S. Census /p60-216.html>.
/content/dam/Census/topics Bureau, Washington, DC, 2011, , “The Research Supplemental
/income/supplemental available at <www.census.gov Poverty Measure: 2012,”
-poverty-measure/spm /library/working-papers/2011 Current Population Reports,
-twgobservations.pdf>. /demo/SEHSD-WP2011-21 P60-247, U.S. Census Bureau,
Janicki, Hubert, “Medical Out-of- .html>. Washington, DC, November
Pocket Expenses in the 2013 , “Estimating the Value of 2013, available at
and 2014 CPS ASEC,” SEHSD Federal Housing Assistance <www.census.gov/library
Working Paper, U.S. Census for the Supplemental Poverty /publications/2013/demo
Bureau, Washington, DC, 2015, Measure: Eliminating the /p60-247.html>.
available at <www.census.gov Public Housing Adjustment,” , “The Research Supplemental
/library/working-papers/2015 SEHSD Working Paper Number Poverty Measure: 2011,”
/demo/Medical-Out-of-pocket 2017–38, U.S. Census Bureau, Current Population Reports,
-Expenses-CPSASEC-2013-2014 Washington, DC, 2017, available P60-244, U.S. Census Bureau,
.html>. at <www.census.gov/library Washington, DC, November
/working-papers/2017/demo 2012, available at
/SEHSD-WP2017-38.html>. <www.census.gov/library
/publications/2012/demo
/p60-244.html>.
U.S. Census Bureau 15
, “The Research Supplemental , “The Supplemental Poverty Smith, Jessica C. and Carla
Poverty Measure: 2010,” Measure: 2013,” Current Medalia, “Health Insurance
Current Population Reports, Population Reports, P60- Coverage in the United States:
P60-241, U.S. Census Bureau, 251, U.S. Census Bureau, 2013,” Current Population
Washington, DC, November Washington, DC, October Reports, P60-250, U.S. Census
2011, available at 2014b, available at Bureau, Washington, DC, 2014,
<www.census.gov/library <www.census.gov/library available at <www.census.gov
/publications/2011/demo /publications/2014/demo /library/publications/2014
/p60-241.html>. /p60-251.html>. /demo/p60-250.html>.
, “The Supplemental Poverty , Thesia Garner, David Wheaton, Laura and Kathryn
Measure in the Survey Johnson, and Patricia Doyle, Stevens, “The Effect of
of Income and Program “Experimental Poverty Different Tax Calculators on
Participation,” U.S. Census Measures: 1990 to 1997,” the Supplemental Poverty
Bureau, Washington, DC, Current Population Reports, Measure,” Urban Institute,
October 2014a, available at P60-205, U.S. Census Bureau, April 2016, available at
<www.census.gov/library Washington, DC, June 1999, <www.census.gov/content
/working-papers/2014 available at <www.census.gov /dam/Census/library/working
/demo/short-01.html>. /library/publications/1999 -papers/2016/demo/Effect-of
, “The Supplemental Poverty /demo/p60-205.html>. -Different-Tax-Calculators
Measure: 2014,” Current Shrider, Emily A., Melissa Kollar, -on-the-SPM.pdf>.
Population Reports, P60- Frances Chen, and Jessica
254, U.S. Census Bureau, Semega, “Income and Poverty
Washington, DC, September in the United States: 2020,”
2015, available at Current Population Reports,
<www.census.gov/library P60-273, U.S. Census Bureau,
/publications/2015/demo Washington, DC, September
/p60-254.html>. 2021, available at
<www.census.gov/library
/publications/2021/demo
/p60-273.html>.
16 U.S. Census Bureau
APPENDIX To estimate a household’s stimu- who reported that they received
lus amount, the following ques- a stimulus payment. Full details
Data Challenges for 2021 tion was added: “Since April 1, of the methodology for valuing
The COVID-19 pandemic created 2020, have you or anyone in your stimulus payments, as well as
challenges to data collection, as household received a ‘stimulus upper and lower-bound estimates
well as changes to safety-net pro- payment,’ that is the coronavirus of stimulus payment values and
grams included in Supplemental (COVID-19) related Economic impacts on SPM rates can be
Poverty Measure (SPM) resources. Impact Payment from the Federal found at <www.census.gov/library
Data collection challenges are dis- Government?” However, due to /working-papers/2021/demo
cussed in <www.census.gov the timing of EIP payments, the /SEHSD-WP2021-18.html>.
/newsroom/blogs/research stimulus values reported on the
In addition to stimulus payments,
-matters/2021/08/how-did-the survey could have included one–
school lunch programs and SNAP
-pandemic-affect-survey three rounds of EIP payments. The
benefits changed during calendar
-response.html>. For calen- second round of payments began
year 2020. The COVID-19 pan-
dar year 2020, several pieces disbursement on December 28,
demic resulted in nationwide tran-
of legislation were passed that 2020. As such, some households
sitions from in-person to virtual
affected resources available received one stimulus payment in
or hybrid schooling, with consid-
to households. The Families calendar year 2020, while other
erable variation throughout the
First Coronavirus Response households received two pay-
country. The U.S. Census Bureau’s
Act (FFCRA); Coronavirus Aid, ments. Additionally, the third
traditional valuation of school
Relief, and Economic Security round of stimulus payments were
lunch benefits, which assumes 179
Act (CARES); and H.R. 133— disbursed during 2021 ASEC data
in-person days (for free, reduced-
Consolidated Appropriations collection. The yes/no indicators
price, and paid lunches) was not
Act, 2021 all contained legislative for receipt of payments were used
able to capture the value of bene-
changes to ensure households to estimate recipiency among
fits received by families, especially
had economic security amid presumed nonfilers, but the value
given the substantial variation
the ongoing coronavirus pan- of EIP payments were estimated in
in in-person school attendance
demic. These legislative changes the CPS ASEC Tax Model. The CPS
across states. Additionally, states
included increased Supplemental ASEC Tax Model does not assume
rolled out pandemic electronic
Nutritional Assistance Program full take-up of stimulus payment
benefits transfer (P-EBT) benefits
(SNAP) benefits, three rounds of receipt. Instead, it follows logical
to individuals who would normally
Economic Impact Payment (EIP)/ assignment of potential receipt,
receive free or reduced-price
stimulus payments, and expanded assigning stimulus payments to
meals. For households receiving
unemployment benefits among all tax units who were assumed
SNAP, the value of P-EBT ben-
other provisions. To fully capture to be filers absent stimulus pay-
efits was added directly to their
these changes in the data, the ments, all tax units who received
SNAP EBT cards. For households
2021 Current Population Survey Supplemental Security Income
not receiving SNAP, but nor-
Annual Social and Economic (SSI), Social Security, or Veterans
mally receiving free or reduced-
Supplement (CPS ASEC) added Affairs (VA) benefits (and for
price lunches, P-EBT cards were
several new questions related to whom stimulus checks were sent
mailed to their homes or added
receipt of stimulus and school out automatically) and all pre-
to another benefit card such as
lunch benefits. sumed nonfilers in our tax model
Medicaid.
U.S. Census Bureau 17
For 2020, an alternative approach The SPM was developed fol- public.18 In November 2011, the
to school lunch valuation was lowing decades of research on Census Bureau released the first
taken that accounts for state-level poverty measurement. Concerns SPM report, providing SPM esti-
variation in the average number about the adequacy of the offi- mates for 2009 and 2010.
of in-person school days, as well cial measure culminated in a
In 2016, OMB convened a new
as the possibility of receiving congressional appropriation in
ITWG on improving the SPM to
P-EBT benefits either directly 1990 for an independent scientific
provide advice on challenges and
or on their SNAP EBT card. For study of the concepts, measure-
opportunities brought before it
individuals who reported receiv- ment methods, and information
by the Census Bureau and the
ing SNAP benefits, the value of needed for a poverty measure. In
BLS concerning data sources,
P-EBT is assumed to be collected response, the National Academy
estimation, survey production,
in their reported SNAP value. For of Sciences (NAS) convened a
and processing activities for
individuals not reporting SNAP Panel on Poverty and Family
development, implementation,
but reporting that they normally Assistance, which released its
publication, and improvement
received free or reduced-price report, “Measuring Poverty: A
of the SPM. With OMB as chair,
school lunch, a value of P-EBT is New Approach,” in 1995 (Citro
the SPM working group is com-
added to their school lunch value. and Michael, 1995).
prised of career federal employ-
Full details of the 2020 school
The Interagency Technical ees representing their respective
lunch valuation methodology can
Working Group (ITWG) on agencies. The agencies currently
be found at <www.census.gov
Developing a Supplemental represented include the Bureau
/library/working-papers/2021
Poverty Measure was formed in of Economic Analysis, the BLS,
/demo/SEHSD-WP2021-20.html>.
2009 and charged with develop- the Council of Economic Advisors,
The introduction of P-EBT bene- ing a set of initial starting points the Census Bureau, the Economic
fits and distribution through SNAP to permit the Census Bureau, Research Service, the Food and
EBT cards poses a measurement in cooperation with the BLS, Nutrition Service, the Department
challenge to SNAP benefits values to produce the SPM. In 2010, of Health and Human Services,
as well. The Food and Nutrition this ITWG (which included rep- the Department of Housing and
Service of the U.S. Department resentatives from the BLS, the Urban Development (HUD), the
of Agriculture (USDA) has indi- Census Bureau, the Economics Internal Revenue Service (IRS),
cated that it is unlikely a house- and Statistics Administration, the the National Center for Education
hold would be able to distinguish Council of Economic Advisers, the Statistics, the National Center for
increases in SNAP benefits due to U.S. Department of Health and Health Statistics, OMB, and the
benefit expansions authorized by Human Services, and the Office of Social Security Administration. As
FFCRA from benefit increases due Management and Budget [OMB]) discussed in the next section, in
to the inclusion of P-EBT benefits. issued a series of suggestions to September 2020, the ITWG rec-
In this year’s report, we show the Census Bureau and the BLS on ommended changes to be imple-
the combined marginal impact how to develop the SPM.17 Their mented in the 2021 SPM report.
of SNAP and school lunches in suggestions drew on the recom- Additionally in 2020, a new NAS
Figure 8, in addition to the indi- mendations of the 1995 NAS Committee on National Statistics
vidual impacts, as reporting of report and the subsequent exten- expert panel was convened to
benefits across the two programs sive research on poverty measure- further evaluate and improve the
is likely commingled. ment. These suggestions were SPM. Recommendations from the
published in the Federal Register, panel are expected in 2022.
SPM HISTORY and the Census Bureau and the
This is the eleventh report BLS reviewed comments from the
describing the SPM that has been
released by the Census Bureau,
18
Federal Register notice Volume 75,
Number 101, page 29513 was issued on
17
Refer to <www.census.gov/content
with support from the May 26, 2010, soliciting public comments
/dam/Census/topics/income/supplemental
U.S. Bureau of Labor Statistics regarding specific methods and data
-poverty-measure/spm-twgobservations
sources in developing the SPM.
.pdf>.
(BLS).
18 U.S. Census Bureau
SPM Changes Implemented in e. Telephone expenditures are A microdata extract with revised
2021 no longer geographically 2019 SPM variables is available at
adjusted.19 <www.census.gov/topics/income
Since the first publication of
SPM estimates in 2011, no major f. Add home internet expendi- -poverty/supplemental-poverty
changes have been made to the tures to thresholds. -measure/data/datasets.html>.
SPM, but research has been ongo- g. Use a composite food, Additional extracts extending the
ing at the BLS and Census Bureau clothing, shelter, and utili- adjustments historically will be
on potential improvements and ties (FCSU) consumer price available soon.
index as opposed to the All
validation of prior assumptions.
In 2018, the ITWG on improving
Items, All Urban index to SPM METHODOLOGY
adjust components.
the SPM announced a process Poverty Thresholds
and timeline for considering These changes have been imple-
mented in all 2019 and 2020 Consistent with the NAS panel
changes to be made to the SPM.
estimates in this report. The com- recommendations and the sug-
In September 2020, the ITWG
bined threshold changes did not gestions of the ITWG, the SPM
convened to review proposals by
change any of the thresholds by thresholds are based on out-of-
the Census Bureau and BLS on
a statistically significant amount, pocket spending on a basic set of
improvements that could feasibly
but did reduce the portion of the goods and services that includes
be implemented by September
thresholds that are geographically food, clothing, shelter, utilities,
2021. Methodological improve-
adjusted. 20 The impact of these and telecommunications (FCSU),
ments were approved for the esti-
changes on 2019 SPM rates can be and a small additional amount
mation of both resources and the
found in Appendix Table 8 in this to allow for other needs (e.g.,
thresholds. Below are changes to
report and in this working paper household supplies, personal care,
the SPM methodology that were
<www.census.gov/library nonwork-related transportation).
voted on during the September
/working-papers/2021/demo SPM thresholds are produced by
30, 2020, SPM ITWG meeting.
/SEHSD-WP2021-17.html>. the Bureau of Labor Statistics
1. Resources (BLS) Division of Price and Index
For more details on the research
a. Move from a national value Number Research (BLS DPINR),
to state-varying values for behind these changes, please
using 5 years of quarterly CE
the Special Supplemental refer to <www.census.gov/topics
interview data for all consumer
Nutrition Program for /income-poverty/supplemental
units with children, lagged 1
Women, Infants and -poverty-measure/library/working
year. 21, 22 All individuals who share
Children (WIC). -papers/topics/potential-changes
expenses with others in the house-
2. Threshold .html> and <www.bls.gov/pir
hold are included in the consumer
a. Move the base of thresholds /spmhome.htm>.
unit. 23 FCSU expenditures are con-
from the average between verted to equivalized values using
19
Since the inception of the SPM,
the 30th–36th percentile to telephone expenditures were categorized a three-parameter equivalence
83 percent of the average of as utilities and, therefore, included in the
housing portion of the thresholds and scale (refer to the “Equivalence
the 47th–53rd percentiles. subject to geographic adjustment. Going Scales” section for more detail).
forward, telephone expenditures will be
b. Expand the estimation sam-
taken out of the shelter component and
ple from all consumer units included with home internet as a separate 21
For information on the CE, refer to
with exactly two children threshold component. <https://stats.bls.gov/cex/>.
20
The portion adjusted is restricted 22
Changes to the threshold estimation
to all consumer units with to housing; housing is defined to include were implemented in 2021. Refer to
children. shelter and utilities. The housing share for <www.census.gov/library/working
the 2019 published thresholds included -papers/2021/demo/SEHSD-WP2021-17
c. Lag Consumer Expenditure telephone in utilities, while for the 2019 .html> and <www.bls.gov/pir/spmhome
Survey (CE) data used to reestimated thresholds utilities are limited .htm> for details.
to energy (i.e., electricity, natural gas, and 23
This includes unmarried partners and
estimate the thresholds by other fuels) and water and related public others making joint expenditure decisions.
1 year. utilities. In the 2019 reestimated thresholds, For full definition, refer to <www.bls.gov
telephone joined food and clothing and the /cex/csxfaqs.htm>.
d. Add imputed in-kind ben- introduction of internet as the parts of the
efits to the thresholds. thresholds not geographically adjusted.
This change has been implemented for
future years.
U.S. Census Bureau 19
The three-parameter equivalence levels. This method produces based on ACS 5-year estimates of
scale is used to convert the esti- thresholds that are consistent with median gross rents for two-bed-
mation sample FCSU expenditures the resource measure. 25 room units with complete kitchen
to those of a reference consumer and plumbing facilities. Separate
unit composed of two adults with Equivalence Scales medians were estimated for each
two children. The ITWG guidelines state that of the 260 metropolitan statistical
the “three-parameter equivalence areas large enough to be identi-
SPM thresholds are produced for
scale” is to be used to adjust SPM fied on the public-use version of
three housing tenure groups to
reference thresholds for the num- the CPS ASEC file. For each state,
account for differences in hous-
ber of adults and children. 26 The a median is estimated for all non-
ing costs. The three groups are
three-parameter scale allows for metropolitan areas (47 areas) and
owners with mortgages, owners
a different adjustment for single for a combination of all smaller
without mortgages, and renters.
parents (Betson, 1996). This scale metropolitan areas within a state
Thresholds reflect 83 percent
has been used in several BLS and (35 areas). This results in 342
of the median of FCSU expendi-
Census Bureau studies (Short, et adjustment factors. For details,
tures for the estimation sample,
al., 1999; Short, 2001). The three- refer to Renwick (2011). 27 The
multiplied by 1.2 to account for
parameter scale is calculated in movement of telephone expendi-
additional basic needs, with
the following way: tures from the utility category to
adjustments for shelter and
the telecommunications category
utilities for each housing group. One and two adults: scale = reduced the share of the thresh-
Refer to the BLS DPINR Research (adults)0.5 olds subject to geographic adjust-
Experimental Poverty Measures
Single parents: scale = (adults ment for all tenure categories.
Web page for specifics regarding
the production of the SPM thresh- + 0.8 * first child + 0.5 * other
Unit of Analysis
olds and related statistics. 24 children)0.7
The ITWG suggested that the
The thresholds used here include All other families: scale = (adults +
resource unit in the SPM include
the value of all noncash benefits 0.5 * children)0.7
all related individuals who live at
included in the resources. CE data In the calculation used to produce the same address, any coresident
used to produce the thresholds thresholds for two adults, the unrelated children who are cared
reflect the use of SNAP benefits scale is set to 1.41. The economy of for by the family (such as foster
for food since these benefits are scale factor is set at 0.70 for other children), and any cohabiters and
considered equivalent to cash. family types, which is within the their children. 28, 29 This definition
However, the CE data as collected 0.65 to 0.75 range recommended corresponds broadly with the unit
does not account for the value by the NAS panel. of data collection (the consumer
of in-kind benefits from other unit) that is employed for the CE
food programs, rent, and energy Geographic Adjustments data that are used to calculate
assistance. The value of these poverty thresholds. They are
The American Community Survey
other in-kind benefits are imputed referred to as SPM Resource Units.
(ACS) is used to adjust the hous-
to the CE data using data col-
ing portion of the FCSU thresh-
lected in the CPS ASEC to assign 27
Renwick, Figueroa, and Aten (2017)
olds for differences in prices
recipiency of school lunch, WIC examined an alternative method of
across geographic areas. The calculation for the geographic indexes
and energy assistance, CE data using Regional Price Parities from the
geographic adjustments are U.S. Bureau of Economic Analysis.
on rental assistance receipt, and 28
Foster children up to the age of 22
program data on average benefit 25
Additional information available at are included in the new unit.
<www.bls.gov/pir/spm/smp-thresholds 29
The official measure of poverty uses
24
These are referred to as BLS DPINR -and-missing-data-problem-6-16.pdf>. the Census Bureau-defined family that
Research Experimental Supplemental 26
The official measure adjusts includes all individuals residing together
Poverty Measure (SPM) Thresholds. thresholds based on family size, number of who are related by birth, marriage, or
Additional information is available at children and adults, as well as whether or adoption and treats all unrelated individuals
<https://stats.bls.gov/pir/spmhome.htm>. not the householder is aged 65 or older. aged 15 and older independently.
20 U.S. Census Bureau
For all resource units that contain of their countable monthly cash National School Lunch Program
a set of male/female unmarried income to the purchase of food, This program offers children free
partners, the female partner’s and SNAP benefits make up the school lunches if family income
weight is used as the SPM family remaining cost of an adequate is below 130 percent of federal
weight. For all other units, there is low-cost diet. This amount is set poverty guidelines, reduced-price
no change in family weight. 30 at the level of the USDA’s Thrifty school meals if family income is
Food Plan. In the CPS ASEC, between 130 and 185 percent of
Official Poverty Treatment of respondents report whether the federal poverty guidelines,
Unrelated Individuals Under the
anyone in the household received and a subsidized school meal for
Age of 15
SNAP benefits in the previ- all other children. 31 In the CPS
Unrelated children under the age ous calendar year and, if so, the ASEC, the reference person
of 15 are excluded from the official face value of those benefits. The is asked how many children
poverty measure universe but annual household amount is pro- “usually” ate a complete lunch
included in the SPM universe. To rated to the SPM Resource Units at school, and if so, if it was a
compare the two measures in the within each household. free or reduced-price school
SPM report, unrelated individuals lunch. The value of school meals
under the age of 15 are assigned The FFCRA authorized states to
is assigned based on the assump-
an official poverty status to match distribute P-EBT payments to
tion that the children received the
that of the reference person of the households with children who
lunches every day during the last
household in which they reside. would have received free or
school year. Note that this method
The official poverty status is not reduced-price school lunches
may overestimate the benefits
recalculated for anyone else in under the National School Lunch
received by each family. To value
the household. A comparison of Act, if not for a reduction in in-
benefits, we obtain amounts
official poverty estimates using person learning due to the pan-
on the cost per lunch from the
different methods is available demic. These temporary food
USDA Food and Nutrition Service,
at Fox (2017a). Prior to the 2016 benefits were provided to help
that administers the school
SPM report, all unrelated children cover the cost of meals children
lunch program. There is no value
under the age of 15 were consid- would have otherwise received
included for school breakfast.
ered poor in the official poverty at school. For SNAP recipients,
estimates used in the SPM reports. the value of P-EBT was added Due to the COVID-19 pandemic,
Since these children were not to existing SNAP EBT cards. As which resulted in a transition from
asked any income questions, they such, CPS ASEC respondents in-person to virtual schooling and
were assigned income of $0 and who reported SNAP values likely disbursement of school lunch
a poverty threshold for a single- included the value of P-EBT in benefits for some free/reduced-
person unit. their SNAP amount as expan- price school lunch recipients via
sions to SNAP values happened P-EBT cards, the methodology for
Noncash Benefits at the same time and respondents valuing school lunch changed for
were likely unable (and were not 2020. 32 The revised school lunch
Supplemental Nutrition Assistance
asked) to separate out SNAP
Program (SNAP)
benefits from P-EBT benefits in 31
The poverty guidelines are issued
SNAP benefits (formerly known each year by the Department of Health
their response. Due to the poten- and Human Services. The guidelines are a
as food stamps) are designed to
tial of commingled response, this simplified version of the Census Bureau’s
allow eligible low-income house- poverty thresholds used for administrative
report also shows the joint mar- purposes—for instance, determining
holds to afford a nutritionally
ginal impacts of SNAP and school financial eligibility for certain federal
adequate diet. Households that programs. For more details and guidelines,
lunch, as well as the individual refer to <https://aspe.hhs.gov/poverty
participate in the SNAP program
impacts. -guidelines>.
are assumed to devote 30 percent 32
The traditional school lunch valuation
is still available on the public-use CPS ASEC
30
Appropriate weighting of these new file at the family level as F_MV_SL, while
units is an area of additional research at the the new valuation is available at the SPM
Census Bureau. unit level as SPM_SCHLUNCH.
U.S. Census Bureau 21
methodology for 2020 considers someone reports receiving WIC and mortgage-interest subsidies
state-level variation in average are also assigned receipt of WIC. targeted to very-low-income rent-
number of in-person school days, If the child is aged 0 or 1 year, ers and are either project-based
potential reporting of P-EBT bene- then we assume that the mother (public housing) or tenant-based
fits in SNAP values, and responses also gets WIC. If there is no child (vouchers). The value of hous-
to a new ASEC survey question in the family, but the household ing subsidies is estimated as the
regarding receipt of meals during reference person said “yes” to the difference between the “market
virtual schooling. 33 WIC question, we assume this is a rent” for the housing unit and the
pregnant woman receiving WIC. total tenant payment. The “market
Supplementary Nutrition Program rent” for the household is esti-
for Women, Infants, and Children Energy Assistance mated using a statistical match
(WIC) Energy assistance typically falls with HUD administrative data from
This program is designed to pro- into three categories. Under the the Public and Indian Housing
vide food assistance and nutri- federal Low-Income Home Energy Information Center and the Tenant
tional screening to low-income Assistance Program (LIHEAP) Rental Assistance Certification
pregnant and postpartum women or similar state/local programs, System. For each household
and their infants and to low- states may help pay heating or identified in the CPS ASEC as
income children up to the age of cooling bills, provide allotments receiving help with rent or living
5. Incomes must be at or below for low-cost weatherization, or in public housing, an attempt was
185 percent of the poverty guide- provide assistance during energy- made to match on state, core-
lines and participants must be related emergencies. States based statistical area (CBSA), and
nutritionally at-risk (having determine eligibility and can pro- household size. 35 The total tenant
abnormal nutritional conditions, vide assistance in various ways, payment is estimated by apply-
nutrition-related medical condi- including cash payments, vendor ing HUD program rules to total
tions, or dietary deficiencies). payments, two-party checks, household income reported in the
Benefits include supplemental vouchers/coupons, and payments CPS ASEC. Generally, participants
foods in the form of food items or directly to landlords. In the CPS in either public housing or tenant-
vouchers for purchases of specific ASEC, the question on energy based subsidy programs admin-
food items. There are questions on assistance asks for information istered by HUD are expected to
current receipt of WIC in the CPS about the entire previous year. contribute the greater of one-third
ASEC. Lacking additional infor- Many households receive both a of their “adjusted” income or 10
mation, we assume 12 months of “regular” benefit and one or more percent of their gross income
participation and value the benefit crisis or emergency benefits.
using state-level average monthly Since energy assistance payments
WIC values obtained from the are often made directly to a utility
USDA. 34 As with school lunch, company or fuel oil vendor, many
assuming yearlong participation households may have difficulty
may overestimate the value of reporting the precise amount of
WIC benefits received by a given the payment made on their behalf.
SPM unit. In these estimates, we
35
HUD operates two major housing
assistance programs: public housing
assume that all children less than Housing Assistance and tenant-based or voucher programs.
5 years old in a household where Previous research has found that
Households can receive hous-
households misreport whether they receive
ing assistance from a plethora public housing or rental assistance in the
33
The details of the revised 2020 school CPS ASEC and that the value of public
lunch valuation methodology can be found
of federal, state, and local pro- housing is not unambiguously worth
at <www.census.gov/library/working grams. Federal housing assistance less than the value of rental assistance
-papers/2021/demo/SEHSD-WP2021-20 (Renwick, 2017). Given these ambiguities
.html>.
consists of a number of programs and increasing challenges in the reporting
34
Details of changing from national administered primarily by the of housing subsidy values across various
average WIC benefit values to state-varying types of housing assistance, beginning in
average WIC values can be found at HUD. These programs traditionally the 2016 SPM report, we have eliminated
<www.census.gov/library/working-papers take the form of rental subsidies the adjustment factor previously applied to
/2020/demo/SEHSD-WP2020-16.html>. public housing subsidy values.
22 U.S. Census Bureau
towards housing costs. 36 More paid by the family. The measure Work-Related Expenses
details on this method are avail- subtracts federal, state, and local Going to work and earning a wage
able in Johnson, et al. (2010). income taxes and Social Security often entails incurring expenses
Initially, subsidies are estimated payroll taxes (FICA) before such as travel to work and pur-
at the household level. If there assessing the ability of a family chase of uniforms or tools. For
is more than one SPM unit in a to obtain basic necessities such work-related expenses (other than
household, then the value of the as FCSU. Taking account of taxes child care), the NAS panel and
subsidy is prorated based on the allows us to account for receipt original SPM ITWG recommended
number of people in the SPM unit of the federal or state Earned subtracting a fixed amount for
relative to the total number of Income Tax Credit (EITC) and each earner 18 years or older.
people in the household. other tax credits. The CPS ASEC Their calculation was based on
does not collect information on the 1987 Survey of Income and
Housing subsidies help families
taxes paid, but instead relies on Program Participation (SIPP)
pay their rent and, as such, are
a tax calculator to simulate taxes data that collected information
added to income for the SPM.
paid. These simulations include on work expenses in a set of
However, there is general agree-
federal and state income taxes supplementary questions. They
ment that, while the value of
and FICA taxes. 37 These simula- calculated 85 percent of median
a housing subsidy can free up
tions also use a statistical match weekly expenses—$14.42 per
a family’s income to purchase
to the IRS Statistics of Income week worked for anyone aged
food and other basic items, it will
public-use microdata file of tax 18 or older in the family in 1992.
do so only to the extent that it
returns. Total expenses were obtained by
meets the need for shelter. Thus,
multiplying this fixed amount by
the values for housing subsidies The first two EIP are included
the number of weeks respondents
included as income are limited to in the CPS ASEC Tax Model and
reported working in the year.
the proportion of the threshold are modeled based on 2020
Each person in the SIPP reports
that is allocated to housing costs. adjusted gross income, number
their own expenditures on work-
The subsidy is capped at the of dependents under age 17, and
related items in a given week. The
housing portion of the appropriate filing status. Presumed nonfilers
most recent available data are
threshold minus the total tenant were assigned EIP values if they
used to calculate median weekly
payment. either reported receiving Social
expenses. 38 The number of weeks
Security, SSI, or VA payments, or
Necessary Expenses Subtracted worked, reported in the CPS
responded affirmatively to the
From Resources ASEC, is multiplied by 85 percent
new receipt of stimulus payment
of median weekly work-related
Taxes question asked in the 2021 CPS
expenses for each person to arrive
The NAS panel and the ITWG ASEC. This
at annual work-related expenses. 39
recommended that the calculation methodology assumes less than
of family resources for poverty 100 percent take-up among
Child Care Expenses
measurement should subtract potentially eligible recipients.
Another important part of work-
necessary expenses that must be Additional details, as well as
related expenses is paying some-
sensitivity tests, are available
36
HUD regulations define “adjusted one to care for children while par-
household income” as cash income, at <www.census.gov/library
excluding income from certain sources
ents work. These expenses have
/working-papers/2021/demo
minus numerous deductions. Three of become important for families
the income exclusions can be identified /SEHSD-WP2021-18.html>.
from the CPS ASEC: income from the
with young children in which both
employment of children, student financial parents (or a single parent) work.
assistance, and earnings in excess of
$480 for each full-time student 18 years
or older. Deductions that can be modeled 37
Wheaton and Stevens (2016) compare 38
Median weekly work expenses were
from the CPS ASEC include $480 for each the Census Bureau’s tax calculator to $46.60 for 2020 using the 2018 SIPP.
dependent, and $400 for any elderly or TAXSIM and the Bakija tax model and find 39
Edwards, et al. (2014) examined an
disabled family member, child care, and consistency in tax estimates across the alternative method of valuing work-related
medical expenses. models. expenses using the ACS.
U.S. Census Bureau 23
To account for child care expenses Medical Expenses Medicare, Medicare Part B premi-
while parents worked, the CPS The ITWG recommended sub- ums are simulated using the rules
ASEC asks parents whether they tracting medical expenses from for income and tax filing status for
pay for child care and how much income, following the NAS panel. people aged 65 and older (refer to
they spent. The amounts paid for The NAS panel was aware that <www.medicare.gov/>).41 For mar-
any type of child care while par- expenditures for health care are a ried respondents with a “spouse
ents are at work are summed over significant portion of a family bud- present,” combined reported
all children. The ITWG, following get and have become an increas- income is used to determine the
the recommendations of the NAS ingly larger budget item since the appropriate Medicare Part B pre-
report, suggested capping the 1960s. These expenses include mium assuming that these couples
amount subtracted from income, the payment of health insurance filed married, joint returns. Finally,
when combined with other work- premiums plus other medically the simulation model assumes two
related expenses, so that these do necessary items such as prescrip- groups paid zero Part B premi-
not exceed total reported earn- tion drugs and doctor copayments ums: (1) respondents enrolled in
ings of the lowest earning refer- that are not covered or reim- Medicare and Medicaid, and (2)
ence person or spouse/partner of bursed by insurance. Subtracting those with a family income less
the reference person in the family. these amounts from income, like than 135 percent of the federal
This capping procedure is applied taxes and work expenses, leaves poverty level.42 This strategy for
before determining poverty the amount of income that the estimating Medicare Part B premi-
status.40 family has available to purchase ums largely follows the method-
the basic bundle of goods. ology developed by Caswell and
Child Support Paid Short (2011). Estimates for 2017
When reporting medical
The NAS panel recommended and beyond reflect the implemen-
expenses, respondents are asked
that since child support received tation of an updated processing
not to report Medicare Part B
from other households is counted system.43
premiums. Instead, Medicare Part
as income, child support paid
B premiums are estimated using
out to those households should 41
We make the simplifying assumption
other information collected in that respondents were insured by Medicare
be deducted from the resources
the CPS ASEC. If respondents for the entire year.
of those households that paid it. 42
The family income assumption is
received Social Security benefits, based on a rough estimate of eligibility and
Without this subtraction, all child participation in at least one of the following
they may have reported Medicare
support is double counted in over- programs: Qualified Medicare Beneficiary,
premiums, and the reported Specified Low-Income Medicare
all income statistics. Questions Beneficiary, or Qualified Individual or
amount is taken. For respondents
ascertaining amounts paid in Qualified Disabled and Working Individuals.
aged 65 and older who reported We do not take into account the possibility
child support are included in the of (state-specific) asset requirements.
that their Social Security pay-
CPS ASEC, and these reported 43
For more details on changes to the
ment was after deductions but did medical expenditures estimation, reference
amounts are subtracted in the Edward R. Berchick and Heide M. Jackson,
not report a deduction amount “Health Insurance Coverage in the 2017
estimates presented here.
greater than $0, the Medicare Part CPS ASEC Research File,” SEHSD
Working Paper Number 2019-01, 2019,
B premium is set at the standard <www.census.gov/library/working-papers
amount per month and added to /2019/demo/SEHSD-WP2019-01.html>,
and U.S. Census Bureau, “Updates to
40
Some analysts have suggested that income and medical expenditures. the Processing of Out of Pocket Medical
this cap may be inappropriate in certain For the remaining respondents Expenditures and Medicare Premiums,”
cases such as if the parent is in school, SEHSD Working Paper 2019-31,
looking for work, or receiving types of who reported being covered by <www.census.gov/library/working-papers
compensation other than earnings. /2019/demo/SEHSD-WP2019-31.html>.
24 U.S. Census Bureau
Appendix Table 1.
Number and Percentage of People in Poverty Using the Supplemental Poverty Measure:
2019 and 2020—Con.
(Numbers in thousands. Margin of error in thousands or percentage points as appropriate. People as of March of the
following year. Information on confidentiality protection, sampling error, nonsampling error, and definitions is available
at <https://www2.census.gov/programs-surveys/cps/techdocs/cpsmar21.pdf>)
SPM 2020 SPM 20191
Difference
Number Percent Number Percent
Characteristic
Margin of Margin of Margin of Margin of
Estimate error2 (±) Estimate error2 (±) Estimate error2 (±) Estimate error2 (±) Number Percent
All people. . . . . . . . . 29,805 766 9.1 0.2 38,300 876 11.8 0.3 *–8,496 *–2.6
Sex
Male. . . . . . . . . . . . . . . . . . . . . . 13,837 418 8.6 0.3 17,666 485 11.1 0.3 *–3,829 *–2.4
Female. . . . . . . . . . . . . . . . . . . . 15,968 438 9.6 0.3 20,634 498 12.4 0.3 *–4,667 *–2.8
Age
Under 18 years. . . . . . . . . . . . . 7,079 325 9.7 0.4 9,253 358 12.6 0.5 *–2,174 *–2.9
18 to 64 years. . . . . . . . . . . . . . 17,433 521 8.8 0.3 22,073 603 11.2 0.3 *–4,640 *–2.4
65 years and older. . . . . . . . . 5,293 246 9.5 0.4 6,975 251 12.8 0.5 *–1,682 *–3.3
Type of Unit
Married couple. . . . . . . . . . . . . 9,728 457 5.0 0.2 13,479 620 6.8 0.3 *–3,751 *–1.8
Cohabiting partners. . . . . . . . 2,466 233 8.5 0.8 3,147 330 12.1 1.2 *–681 *–3.5
Female reference person . . . 7,668 429 18.2 1.0 9,732 474 23.8 1.0 *–2,064 *–5.6
Male reference person. . . . . . 1,818 204 11.7 1.2 2,141 236 14.2 1.5 *–323 *–2.5
Unrelated individuals. . . . . . . 8,125 265 17.5 0.5 9,801 319 21.6 0.6 *–1,676 *–4.1
Race3 and Hispanic Origin
White. . . . . . . . . . . . . . . . . . . . . 20,156 587 8.1 0.2 26,029 645 10.5 0.3 *–5,872 *–2.4
White, not Hispanic . . . . . . 12,646 481 6.5 0.2 15,921 477 8.2 0.2 *–3,275 *–1.7
Black . . . . . . . . . . . . . . . . . . . . . 6,369 348 14.6 0.8 8,144 409 18.9 1.0 *–1,775 *–4.3
Asian . . . . . . . . . . . . . . . . . . . . . 1,770 200 8.8 1.0 2,257 189 11.3 0.9 *–487 *–2.5
Hispanic (any race). . . . . . . . . 8,570 436 14.0 0.7 11,437 478 18.8 0.8 *–2,867 *–4.9
Nativity
Native-born . . . . . . . . . . . . . . . 23,657 678 8.4 0.2 30,460 733 10.9 0.3 *–6,803 *–2.5
Foreign-born. . . . . . . . . . . . . . 6,148 296 13.7 0.6 7,840 322 17.5 0.7 *–1,693 *–3.7
Naturalized citizen . . . . . . . 2,335 163 10.3 0.7 3,076 200 13.5 0.9 *–741 *–3.2
Not a citizen. . . . . . . . . . . . . 3,813 243 17.2 1.0 4,764 259 21.5 1.2 *–951 *–4.3
Educational Attainment
Total, aged 25 and
older. . . . . . . . . . . . . . . 18,866 505 8.4 0.2 24,312 547 10.9 0.2 *–5,446 *–2.5
No high school diploma . . . . 4,068 202 20.3 0.9 5,600 261 27.7 1.1 *–1,532 *–7.4
High school, no college. . . . . 6,929 280 11.1 0.4 8,588 305 13.9 0.5 *–1,659 *–2.9
Some college. . . . . . . . . . . . . . 4,177 207 7.3 0.3 5,449 231 9.5 0.4 *–1,273 *–2.1
Bachelor’s degree or
higher . . . . . . . . . . . . . . . . . . . 3,692 212 4.3 0.2 4,674 214 5.6 0.2 *–982 *–1.2
Tenure
Owner/mortgage . . . . . . . . . . 5,283 344 3.9 0.2 7,035 389 5.1 0.3 *–1,752 *–1.2
Owner/no mortgage/rent-
free. . . . . . . . . . . . . . . . . . . . . . 7,313 390 8.2 0.4 9,726 424 10.9 0.4 *–2,413 *–2.7
Renter . . . . . . . . . . . . . . . . . . . . 17,210 634 17.1 0.6 21,540 760 22.1 0.7 *–4,331 *–5.0
Residence4
Inside metropolitan
statistical areas. . . . . . . . . . . 26,286 719 9.3 0.2 33,359 871 11.8 0.3 *–7,074 *–2.5
Inside principal cities. . . . . 12,573 553 11.9 0.5 15,654 673 14.9 0.6 *–3,081 *–3.1
Outside principal cities . . . 13,712 464 7.7 0.3 17,705 631 10.0 0.3 *–3,992 *–2.2
Outside metropolitan
statistical areas. . . . . . . . . . . 3,519 376 8.3 0.7 4,941 458 11.6 0.7 *–1,422 *–3.3
Region
Northeast. . . . . . . . . . . . . . . . . 4,686 337 8.5 0.6 6,341 366 11.5 0.7 *–1,655 *–3.0
Midwest. . . . . . . . . . . . . . . . . . . 4,498 299 6.7 0.4 6,081 355 9.0 0.5 *–1,584 *–2.3
South. . . . . . . . . . . . . . . . . . . . . 12,704 550 10.1 0.4 15,662 644 12.6 0.5 *–2,958 *–2.5
West. . . . . . . . . . . . . . . . . . . . . . 7,918 415 10.1 0.5 10,216 405 13.1 0.5 *–2,298 *–3.0
Footnotes provided at end of table.
U.S. Census Bureau 25
Appendix Table 1.
Number and Percentage of People in Poverty Using the Supplemental Poverty Measure:
2019 and 2020—Con.
(Numbers in thousands. Margin of error in thousands or percentage points as appropriate. People as of March of the
following year. Information on confidentiality protection, sampling error, nonsampling error, and definitions is available
at <https://www2.census.gov/programs-surveys/cps/techdocs/cpsmar21.pdf>)
SPM 2020 SPM 20191
Difference
Number Percent Number Percent
Characteristic
Margin of Margin of Margin of Margin of
Estimate error2 (±) Estimate error2 (±) Estimate error2 (±) Estimate error2 (±) Number Percent
Health Insurance Coverage
With private insurance. . . . . . 8,976 376 4.1 0.2 11,933 500 5.4 0.2 *–2,956 *–1.3
With public, no private
insurance . . . . . . . . . . . . . . . . 15,290 558 18.8 0.6 19,993 534 25.8 0.6 *–4,703 *–6.9
Not insured. . . . . . . . . . . . . . . . 5,538 319 19.4 1.0 6,375 323 23.8 1.1 *–836 *–4.3
Work Experience
Total, 18 to 64 years.. . 17,433 521 8.8 0.3 22,073 603 11.2 0.3 *–4,640 *–2.4
All workers. . . . . . . . . . . . . . . . 7,056 275 4.6 0.2 10,491 365 6.8 0.2 *–3,435 *–2.2
Worked full-time,
year-round. . . . . . . . . . . . . . . 1,864 118 1.9 0.1 4,371 225 3.9 0.2 *–2,508 *–2.0
Less than full-time,
year-round. . . . . . . . . . . . . . . 5,192 234 9.8 0.4 6,120 255 14.6 0.5 *–928 *–4.7
Did not work at least
1 week . . . . . . . . . . . . . . . . . . . 10,377 369 22.9 0.7 11,582 388 27.0 0.8 *–1,205 *–4.1
Disability Status5
Total, 18 to 64 years.. . 17,433 521 8.8 0.3 22,073 603 11.2 0.3 *–4,640 *–2.4
With a disability . . . . . . . . . . . 2,562 161 17.6 1.0 3,203 164 22.2 1.1 *–641 *–4.6
With no disability. . . . . . . . . . 14,842 469 8.2 0.3 18,804 554 10.3 0.3 *–3,962 *–2.2
* An asterisk preceding an estimate indicates change is statistically different from zero at the 90 percent confidence level.
1
The data for 2019 reflect the implementation of revised Supplemental Poverty Measure methodology. More information is available in the
report appendix.
2
A margin of error (MOE) is a measure of an estimate's variability. The larger the MOE in relation to the size of the estimate, the less reli-
able the estimate. This number, when added to and subtracted from the estimate, forms the 90 percent confidence interval. MOEs shown in
this table are based on standard errors calculated using replicate weights.
3
Federal surveys give respondents the option of reporting more than one race. Therefore, two basic ways of defining a race group are
possible. A group, such as Asian, may be defined as those who reported Asian and no other race (the race-alone or single-race concept) or
as those who reported Asian regardless of whether they also reported another race (the race-alone-or-in-combination concept). This table
shows data using the first approach (race alone). The use of the single-race population does not imply that it is the preferred method of pre-
senting or analyzing data. The Census Bureau uses a variety of approaches. Data for American Indians and Alaska Natives, Native Hawaiians
and Other Pacific Islanders, and those reporting two or more races are not shown separately.
4
Information on metropolitan statistical areas and principal cities is available at <www.census.gov/programs-surveys/metro-micro/about
/glossary.html>.
5
The sum of those with and without a disability does not equal the total because disability status is not defined for individuals in the
U.S. armed forces.
Note: Details may not sum to totals due to rounding.
Source: U.S. Census Bureau, Current Population Survey, 2020 and 2021 Annual Social and Economic Supplements (CPS ASEC).
26 U.S. Census Bureau
Appendix Table 2.
Number and Percentage of People in Poverty by Different Poverty Measures: 2020—Con.
(Numbers in thousands. Margin of error in thousands or percentage points as appropriate. People as of March of the
following year. Information on confidentiality protection, sampling error, nonsampling error, and definitions is available
at <https://www2.census.gov/programs-surveys/cps/techdocs/cpsmar21.pdf>)
Official1 SPM Difference
Number Percent Number Percent
Characteristic
Margin of Margin of Margin of Margin of
Number1 Estimate error2 (±) Estimate error2 (±) Estimate error2 (±) Estimate error2 (±) Number Percent
All people. . . . . . . . 326,195 37,314 891 11.4 0.3 29,805 766 9.1 0.2 *–7,509 *–2.3
Sex
Male. . . . . . . . . . . . . . . . . . . . . 159,977 16,370 461 10.2 0.3 13,837 418 8.6 0.3 *–2,532 *–1.6
Female. . . . . . . . . . . . . . . . . . . 166,219 20,944 540 12.6 0.3 15,968 438 9.6 0.3 *–4,976 *–3.0
Age
Under 18 years. . . . . . . . . . . . 72,777 11,674 416 16.0 0.6 7,079 325 9.7 0.4 *–4,595 *–6.3
18 to 64 years. . . . . . . . . . . . . 197,582 20,640 524 10.4 0.3 17,433 521 8.8 0.3 *–3,207 *–1.6
65 years and older. . . . . . . . 55,836 5,000 243 9.0 0.4 5,293 246 9.5 0.4 *293 *0.5
Type of Unit
Married couple. . . . . . . . . . . . 193,316 10,193 508 5.3 0.3 9,728 457 5.0 0.2 *–464 *–0.2
Cohabiting partners. . . . . . . 28,856 6,661 331 23.1 1.0 2,466 233 8.5 0.8 *–4,194 *–14.5
Female reference person . . 42,090 10,073 492 23.9 1.1 7,668 429 18.2 1.0 *–2,405 *–5.7
Male reference person. . . . . 15,571 1,636 189 10.5 1.2 1,818 204 11.7 1.2 *182 *1.2
Unrelated individuals. . . . . . 46,362 8,752 265 18.9 0.5 8,125 265 17.5 0.5 *–627 *–1.4
Race3 and Hispanic Origin
White. . . . . . . . . . . . . . . . . . . . 248,163 25,052 666 10.1 0.3 20,156 587 8.1 0.2 *–4,896 *–2.0
White, not Hispanic . . . . . 194,524 15,974 514 8.2 0.3 12,646 481 6.5 0.2 *–3,327 *–1.7
Black . . . . . . . . . . . . . . . . . . . . 43,490 8,493 416 19.5 1.0 6,369 348 14.6 0.8 *–2,124 *–4.9
Asian . . . . . . . . . . . . . . . . . . . . 20,157 1,629 173 8.1 0.8 1,770 200 8.8 1.0 141 0.7
Hispanic (any race). . . . . . . . 61,304 10,422 474 17.0 0.8 8,570 436 14.0 0.7 *–1,852 *–3.0
Nativity
Native-born . . . . . . . . . . . . . . 281,396 31,292 786 11.1 0.3 23,657 678 8.4 0.2 *–7,635 *–2.7
Foreign-born. . . . . . . . . . . . . 44,799 6,022 314 13.4 0.7 6,148 296 13.7 0.6 126 0.3
Naturalized citizen . . . . . . 22,667 2,080 153 9.2 0.6 2,335 163 10.3 0.7 *255 *1.1
Not a citizen. . . . . . . . . . . . 22,132 3,942 260 17.8 1.1 3,813 243 17.2 1.0 –129 –0.6
Educational Attainment
Total, aged 25 and
older. . . . . . . . . . . . . . 224,580 21,443 540 9.5 0.2 18,866 505 8.4 0.2 *–2,578 *–1.1
No high school diploma . . . 20,054 4,953 219 24.7 1.0 4,068 202 20.3 0.9 *–886 *–4.4
High school, no college. . . . 62,547 8,273 290 13.2 0.4 6,929 280 11.1 0.4 *–1,343 *–2.1
Some college. . . . . . . . . . . . . 56,942 4,781 210 8.4 0.4 4,177 207 7.3 0.3 *–604 *–1.1
Bachelor’s degree or
higher . . . . . . . . . . . . . . . . . . 85,037 3,436 214 4.0 0.2 3,692 212 4.3 0.2 *256 *0.3
Tenure
Owner/mortgage . . . . . . . . . 136,077 5,570 336 4.1 0.2 5,283 344 3.9 0.2 –288 –0.2
Owner/no mortgage/rent-
free. . . . . . . . . . . . . . . . . . . . . 89,492 9,707 484 10.8 0.5 7,313 390 8.2 0.4 *–2,394 *–2.7
Renter . . . . . . . . . . . . . . . . . . . 100,627 22,036 729 21.9 0.7 17,210 634 17.1 0.6 *–4,827 *–4.8
Residence4
Inside metropolitan
statistical areas. . . . . . . . . . 283,834 31,337 851 11.0 0.3 26,286 719 9.3 0.2 *–5,051 *–1.8
Inside principal cities. . . . 106,022 15,128 631 14.3 0.5 12,573 553 11.9 0.5 *–2,554 *–2.4
Outside principal cities . . 177,812 16,209 615 9.1 0.3 13,712 464 7.7 0.3 *–2,497 *–1.4
Outside metropolitan
statistical areas. . . . . . . . . . 42,362 5,977 580 14.1 0.9 3,519 376 8.3 0.7 *–2,458 *–5.8
Region
Northeast. . . . . . . . . . . . . . . . 54,850 5,558 340 10.1 0.6 4,686 337 8.5 0.6 *–872 *–1.6
Midwest. . . . . . . . . . . . . . . . . . 67,561 6,840 377 10.1 0.6 4,498 299 6.7 0.4 *–2,342 *–3.5
South. . . . . . . . . . . . . . . . . . . . 125,612 16,651 621 13.3 0.5 12,704 550 10.1 0.4 *–3,947 *–3.1
West. . . . . . . . . . . . . . . . . . . . . 78,173 8,265 410 10.6 0.5 7,918 415 10.1 0.5 *–347 *–0.4
Footnotes provided at end of table.
U.S. Census Bureau 27
Appendix Table 2.
Number and Percentage of People in Poverty by Different Poverty Measures: 2020—Con.
(Numbers in thousands. Margin of error in thousands or percentage points as appropriate. People as of March of the
following year. Information on confidentiality protection, sampling error, nonsampling error, and definitions is available
at <https://www2.census.gov/programs-surveys/cps/techdocs/cpsmar21.pdf>)
Official1 SPM Difference
Number Percent Number Percent
Characteristic
Margin of Margin of Margin of Esti- Margin of
Number1 Estimate error2 (±) Estimate error2 (±) Estimate error2 (±) mate error2 (±) Number Percent
Health Insurance Coverage
With private insurance. . . . . . 216,532 8,642 384 4.0 0.2 8,976 376 4.1 0.2 *335 *0.2
With public, no private
insurance . . . . . . . . . . . . . . . . 81,149 22,182 628 27.3 0.6 15,290 558 18.8 0.6 *–6,892 *–8.5
Not insured. . . . . . . . . . . . . . . . 28,515 6,490 355 22.8 1.1 5,538 319 19.4 1.0 *–952 *–3.3
Work Experience
Total, 18 to 64 years . . 197,582 20,640 524 10.4 0.3 17,433 521 8.8 0.3 *–3,207 *–1.6
All workers. . . . . . . . . . . . . . . . 152,246 7,593 266 5.0 0.2 7,056 275 4.6 0.2 *–537 *–0.4
Worked full-time,
year-round. . . . . . . . . . . . . . . 99,404 1,609 121 1.6 0.1 1,864 118 1.9 0.1 *255 *0.3
Less than full-time,
year-round. . . . . . . . . . . . . . . 52,842 5,984 231 11.3 0.4 5,192 234 9.8 0.4 *–792 *–1.5
Did not work at least
1 week . . . . . . . . . . . . . . . . . . . 45,336 13,047 392 28.8 0.7 10,377 369 22.9 0.7 *–2,670 *–5.9
Disability Status 5
Total, 18 to 64 years . . 197,582 20,640 524 10.4 0.3 17,433 521 8.8 0.3 *–3,207 *–1.6
With a disability . . . . . . . . . . . 14,559 3,643 183 25.0 1.1 2,562 161 17.6 1.0 *–1,081 *–7.4
With no disability. . . . . . . . . . 181,934 16,966 465 9.3 0.3 14,842 469 8.2 0.3 *–2,124 *–1.2
* An asterisk preceding an estimate indicates change is statistically different from zero at the 90 percent confidence level.
1
Includes unrelated individuals under the age of 15.
2
A margin of error (MOE) is a measure of an estimate's variability. The larger the MOE in relation to the size of the estimate, the less reliable
the estimate. This number, when added to and subtracted from the estimate, forms the 90 percent confidence interval. MOEs shown in this
table are based on standard errors calculated using replicate weights.
3
Federal surveys give respondents the option of reporting more than one race. Therefore, two basic ways of defining a race group are
possible. A group, such as Asian, may be defined as those who reported Asian and no other race (the race-alone or single-race concept) or as
those who reported Asian regardless of whether they also reported another race (the race-alone-or-in-combination concept). This table shows
data using the first approach (race alone). The use of the single-race population does not imply that it is the preferred method of presenting or
analyzing data. The Census Bureau uses a variety of approaches. Data for American Indians and Alaska Natives, Native Hawaiians and Other
Pacific Islanders, and those reporting two or more races are not shown separately.
4
Information on metropolitan statistical areas and principal cities is available at <www.census.gov/programs-surveys/metro-micro/about
/glossary.html>.
5
The sum of those with and without a disability does not equal the total because disability status is not defined for individuals in the
U.S. armed forces.
Note: Details may not sum to totals due to rounding.
Source: U.S. Census Bureau, Current Population Survey, 2021 Annual Social and Economic Supplement (CPS ASEC).
28 U.S. Census Bureau
Appendix Table 3.
Two-Adult, Two-Child Poverty Thresholds: 2019 and 2020
(In nominal dollars)
Measure 20191 Standard error 2020 Standard error
Official Poverty Measure
Official poverty measure . . . . . . . . . . . . . . . . . . . . 25,926 N 26,246 N
Research Supplemental Poverty Measure
Owners with mortgages. . . . . . . . . . . . . . . . . . . . . 29,080 210 29,959 241
Owners without mortgages. . . . . . . . . . . . . . . . . . 24,413 344 25,222 402
Renters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29,194 179 30,150 255
N Not available.
1
The data for 2019 reflect the implementation of revised Supplemental Poverty Measure (SPM) methodology. More information is avail-
able in the report appendix.
Source: The thresholds were produced by Juan D. Munoz under the guidance of Thesia I. Garner. Munoz and Garner work in the Division of
Price and Index Number Research, Bureau of Labor Statistics (BLS). The SPM thresholds are based on consumer unit weighted data; standard
errors of the SPM thresholds are derived using replicate weights available on the Consumer Expenditure Survey (CE) data files (information
regarding the estimation of standard errors based on CE replicate weights is available at <www.bls.gov/cex/pumd-getting-started-guide.
htm>). The SPM thresholds and statistics are produced for research purposes only using the U.S. Consumer Expenditure Interview Survey.
The thresholds are not BLS production quality. Methodological details and related research regarding the SPM thresholds are available at
<https://stats.bls.gov/pir/spmhome.htm>. The thresholds and related statistics were finalized as of July 28, 2021.
U.S. Census Bureau 29
Appendix Table 4.
Percentage of People by Ratio of Income/Resources to Poverty Threshold: 2019 and 2020—Con.
(Margin of error in percentage points. People as of March of the following year. Information on confidentiality protection,
sampling error, nonsampling error, and definitions is available at <https://www2.census.gov/programs-surveys/cps/techdocs
/cpsmar21.pdf>)
Less Margin 0.50 Margin 1.00 Margin 1.50 Margin 2.00 Margin 4.00 Margin
Characteristic than of error1 to of error1 to of error1 to of error1 to of error1 or of error1
0.50 (±) 0.99 (±) 1.49 (±) 1.99 (±) 3.99 (±) more (±)
2020
OFFICIAL2
All people. . . . . . . 5.5 0.2 5.9 0.2 7.9 0.2 8.2 0.2 28.3 0.4 44.1 0.4
Age
Under 18 years. . . . . . . . . . . 7.6 0.4 8.4 0.4 9.8 0.4 9.9 0.4 28.7 0.7 35.5 0.6
18 to 64 years. . . . . . . . . . . . 5.2 0.2 5.3 0.2 6.8 0.2 7.0 0.2 27.8 0.4 48.0 0.5
65 years and older. . . . . . . 3.9 0.3 5.1 0.3 9.6 0.4 10.1 0.4 29.7 0.6 41.6 0.8
Race3 and Hispanic
Origin
White. . . . . . . . . . . . . . . . . . . 4.8 0.2 5.3 0.2 7.3 0.2 8.0 0.2 28.2 0.4 46.4 0.5
White, not Hispanic . . . . 4.1 0.2 4.1 0.2 5.9 0.2 6.8 0.3 27.0 0.5 52.1 0.6
Black . . . . . . . . . . . . . . . . . . . 9.9 0.7 9.7 0.7 11.3 0.8 9.9 0.7 30.7 1.0 28.5 1.1
Asian . . . . . . . . . . . . . . . . . . . 3.8 0.6 4.3 0.7 6.2 0.8 6.1 0.8 23.3 1.4 56.3 1.5
Hispanic (any race). . . . . . . 7.3 0.5 9.7 0.6 12.4 0.6 12.0 0.6 32.5 0.9 26.1 0.7
SPM
All people. . . . . . . 3.3 0.1 5.9 0.2 12.5 0.3 13.9 0.3 38.4 0.4 26.0 0.4
Age
Under 18 years. . . . . . . . . . . 2.9 0.3 6.8 0.4 15.1 0.5 17.0 0.6 39.0 0.6 19.1 0.6
18 to 64 years. . . . . . . . . . . . 3.3 0.1 5.6 0.2 11.3 0.3 13.0 0.3 39.2 0.4 27.8 0.5
65 years and older. . . . . . . 3.8 0.3 5.7 0.3 13.6 0.5 13.2 0.5 35.1 0.7 28.6 0.8
Race3 and Hispanic
Origin
White. . . . . . . . . . . . . . . . . . . 3.0 0.1 5.1 0.2 11.4 0.3 13.1 0.3 39.1 0.4 28.3 0.5
White, not Hispanic . . . . 2.8 0.2 3.7 0.2 8.7 0.3 11.2 0.3 40.4 0.5 33.2 0.6
Black . . . . . . . . . . . . . . . . . . . 4.7 0.4 9.9 0.7 18.6 0.9 18.2 1.0 34.9 1.1 13.7 0.8
Asian . . . . . . . . . . . . . . . . . . . 3.4 0.5 5.4 0.8 11.3 1.1 12.9 1.1 36.9 1.6 30.2 1.5
Hispanic (any race). . . . . . . 4.0 0.4 10.0 0.6 21.2 0.8 20.0 0.8 34.7 0.9 10.2 0.5
20194
OFFICIAL2
All people. . . . . . . 4.7 0.2 5.7 0.2 7.7 0.2 8.2 0.2 28.4 0.4 45.3 0.5
Age
Under 18 years. . . . . . . . . . . 6.2 0.4 8.2 0.4 10.3 0.5 9.8 0.4 29.0 0.6 36.6 0.7
18 to 64 years. . . . . . . . . . . . 4.5 0.2 5.0 0.2 6.4 0.2 7.4 0.2 27.8 0.4 49.0 0.5
65 years and older. . . . . . . 3.7 0.3 5.2 0.3 8.8 0.4 9.1 0.4 29.6 0.7 43.6 0.8
Race3 and Hispanic
Origin
White. . . . . . . . . . . . . . . . . . . 4.1 0.2 5.0 0.2 7.1 0.2 7.8 0.2 28.3 0.5 47.7 0.5
White, not Hispanic . . . . 3.5 0.2 3.8 0.2 5.7 0.2 6.7 0.2 27.0 0.5 53.4 0.6
Black . . . . . . . . . . . . . . . . . . . 8.2 0.6 10.6 0.7 10.8 0.7 10.5 0.8 30.4 1.2 29.5 1.1
Asian . . . . . . . . . . . . . . . . . . . 4.1 0.6 3.2 0.6 6.0 0.8 6.3 0.9 23.7 1.5 56.7 1.6
Hispanic (any race). . . . . . . 6.4 0.5 9.4 0.6 12.3 0.7 12.2 0.6 32.6 1.0 27.0 0.9
Footnotes provided at end of table.
30 U.S. Census Bureau
Appendix Table 4.
Percentage of People by Ratio of Income/Resources to Poverty Threshold: 2019 and 2020—Con.
(Margin of error in percentage points. People as of March of the following year. Information on confidentiality protection,
sampling error, nonsampling error, and definitions is available at <https://www2.census.gov/programs-surveys/cps/techdocs
/cpsmar21.pdf>)
Less Margin 0.50 Margin 1.00 Margin 1.50 Margin 2.00 Margin 4.00 Margin
Characteristic than of error1 to of error1 to of error1 to of error1 to of error1 or of error1
0.50 (±) 0.99 (±) 1.49 (±) 1.99 (±) 3.99 (±) more (±)
SPM
All people. . . . . . . 4.0 0.2 7.8 0.2 14.0 0.3 12.7 0.3 36.4 0.4 25.1 0.4
Age
Under 18 years. . . . . . . . . . . 3.5 0.3 9.1 0.5 17.9 0.6 15.0 0.5 36.3 0.7 18.1 0.5
18 to 64 years. . . . . . . . . . . . 4.0 0.2 7.2 0.3 12.6 0.3 12.2 0.3 37.3 0.5 26.7 0.5
65 years and older. . . . . . . 4.7 0.3 8.1 0.4 13.8 0.5 11.4 0.5 33.4 0.7 28.6 0.8
Race3 and Hispanic
Origin
White. . . . . . . . . . . . . . . . . . . 3.6 0.2 6.9 0.2 12.7 0.3 12.4 0.3 37.1 0.5 27.4 0.5
White, not Hispanic . . . . 3.2 0.2 5.0 0.2 9.9 0.3 11.1 0.3 38.8 0.5 32.0 0.6
Black . . . . . . . . . . . . . . . . . . . 6.2 0.6 12.7 0.8 21.0 1.1 14.6 0.8 33.0 1.2 12.6 0.7
Asian . . . . . . . . . . . . . . . . . . . 4.5 0.6 6.8 0.8 13.3 1.2 10.8 1.1 35.2 1.5 29.4 1.4
Hispanic (any race). . . . . . . 4.9 0.4 13.9 0.7 22.9 0.9 16.8 0.8 31.0 1.0 10.5 0.6
1
A margin of error (MOE) is a measure of an estimate's variability. The larger the MOE in relation to the size of the estimate, the less reli-
able the estimate. This number, when added to and subtracted from the estimate, forms the 90 percent confidence interval. MOEs shown in
this table are based on standard errors calculated using replicate weights.
2
Includes unrelated individuals under the age of 15.
3
Federal surveys give respondents the option of reporting more than one race. Therefore, two basic ways of defining a race group are
possible. A group, such as Asian, may be defined as those who reported Asian and no other race (the race-alone or single-race concept) or
as those who reported Asian regardless of whether they also reported another race (the race-alone-or-in-combination concept). This table
shows data using the first approach (race alone). The use of the single-race population does not imply that it is the preferred method of pre-
senting or analyzing data. The Census Bureau uses a variety of approaches. Data for American Indians and Alaska Natives, Native Hawaiians
and Other Pacific Islanders, and those reporting two or more races are not shown separately.
4
The data for 2019 reflect the implementation of revised Supplemental Poverty Measure methodology. More information is available in
the report appendix.
Note: Details may not sum to totals due to rounding.
Source: U.S. Census Bureau, Current Population Survey, 2020 and 2021 Annual Social and Economic Supplements (CPS ASEC).
U.S. Census Bureau 31
Appendix Table 5.
Number and Percentage of People in Poverty by State: 3-Year Average of 2018–2020—Con.
(Numbers in thousands. Margin of error in thousands or percentage points as appropriate. People as of March of the
following year. Information on confidentiality protection, sampling error, nonsampling error, and definitions is available
at, see <https://www2.census.gov/programs-surveys/cps/techdocs/cpsmar21.pdf>)
Official1 SPM Difference
Number Percent Number Percent
State
Margin of Margin of Margin of Margin of
Estimate error2 (±) Estimate error2 (±) Estimate error2 (±) Estimate error2 (±) Number Percent
United States . . . . . 36,525 536 11.2 0.2 36,508 500 11.2 0.2 –16 Z
Alabama. . . . . . . . . . . . . . . . 712 95 14.6 2.0 588 88 12.0 1.8 *–125 *–2.6
Alaska . . . . . . . . . . . . . . . . . . 87 8 12.2 1.1 86 8 12.0 1.1 –1 –0.1
Arizona . . . . . . . . . . . . . . . . . 822 105 11.2 1.5 767 97 10.4 1.3 *–54 *–0.7
Arkansas. . . . . . . . . . . . . . . . 432 40 14.7 1.4 336 35 11.4 1.2 *–96 *–3.3
California. . . . . . . . . . . . . . . . 4,328 190 11.0 0.5 6,041 223 15.4 0.6 *1,713 *4.4
Colorado. . . . . . . . . . . . . . . . 535 88 9.3 1.6 644 83 11.2 1.5 *109 *1.9
Connecticut. . . . . . . . . . . . . 343 47 9.9 1.4 370 50 10.7 1.4 28 0.8
Delaware. . . . . . . . . . . . . . . . 80 10 8.1 1.0 95 11 9.8 1.1 *16 *1.6
District of Columbia. . . . . . 104 8 14.7 1.1 116 10 16.5 1.4 *13 *1.8
Florida. . . . . . . . . . . . . . . . . . 2,740 185 12.8 0.9 2,992 161 14.0 0.8 *252 *1.2
Georgia. . . . . . . . . . . . . . . . . 1,399 145 13.4 1.4 1,264 112 12.1 1.1 *–135 *–1.3
Hawaii . . . . . . . . . . . . . . . . . . 131 18 9.5 1.3 166 21 12.0 1.5 *35 *2.5
Idaho. . . . . . . . . . . . . . . . . . . 166 17 9.2 0.9 135 17 7.5 0.9 *–31 *–1.7
Illinois. . . . . . . . . . . . . . . . . . . 1,150 103 9.2 0.8 1,195 109 9.5 0.9 46 0.4
Indiana. . . . . . . . . . . . . . . . . . 751 68 11.3 1.0 576 62 8.7 0.9 *–175 *–2.6
Iowa. . . . . . . . . . . . . . . . . . . . 286 48 9.1 1.5 206 37 6.6 1.2 *–79 *–2.5
Kansas. . . . . . . . . . . . . . . . . . 246 31 8.7 1.1 182 24 6.4 0.8 *–64 *–2.2
Kentucky. . . . . . . . . . . . . . . . 637 84 14.4 1.9 481 65 10.9 1.5 *–156 *–3.5
Louisiana. . . . . . . . . . . . . . . . 792 51 17.4 1.1 634 52 13.9 1.1 *–158 *–3.5
Maine. . . . . . . . . . . . . . . . . . . 134 19 10.0 1.5 98 17 7.3 1.3 *–35 *–2.6
Maryland. . . . . . . . . . . . . . . . 488 65 8.1 1.1 652 71 10.8 1.2 *163 *2.7
Massachusetts. . . . . . . . . . . 565 59 8.2 0.9 649 66 9.4 1.0 *84 *1.2
Michigan. . . . . . . . . . . . . . . . 1,044 97 10.6 1.0 865 85 8.8 0.9 *–178 *–1.8
Minnesota. . . . . . . . . . . . . . . 415 49 7.3 0.9 332 48 5.9 0.8 *–84 *–1.5
Mississippi. . . . . . . . . . . . . . . 549 61 18.8 2.1 422 35 14.5 1.2 *–126 *–4.3
Missouri. . . . . . . . . . . . . . . . . 653 96 10.8 1.6 489 68 8.1 1.1 *–164 *–2.7
Montana . . . . . . . . . . . . . . . . 109 11 10.3 1.1 96 10 9.1 1.0 *–13 *–1.3
Nebraska. . . . . . . . . . . . . . . . 175 21 9.2 1.1 154 23 8.1 1.2 *–21 *–1.1
Nevada . . . . . . . . . . . . . . . . . 373 37 12.1 1.2 357 36 11.5 1.2 –16 –0.5
New Hampshire. . . . . . . . . . 72 11 5.3 0.8 89 13 6.5 0.9 *17 *1.2
New Jersey. . . . . . . . . . . . . . 664 67 7.6 0.8 871 74 10.0 0.8 *207 *2.4
New Mexico . . . . . . . . . . . . . 332 25 16.1 1.3 252 21 12.2 1.1 *–81 *–3.9
New York. . . . . . . . . . . . . . . . 2,252 125 11.8 0.7 2,549 142 13.3 0.7 *298 *1.6
North Carolina. . . . . . . . . . . 1,379 100 13.2 1.0 1,247 101 11.9 1.0 *–131 *–1.3
North Dakota. . . . . . . . . . . . 71 8 9.5 1.1 65 8 8.6 1.1 *–7 *–0.9
Ohio. . . . . . . . . . . . . . . . . . . . 1,428 138 12.4 1.2 1,070 110 9.3 1.0 *–358 *–3.1
Oklahoma. . . . . . . . . . . . . . . 516 61 13.2 1.5 403 52 10.3 1.3 *–113 *–2.9
Oregon . . . . . . . . . . . . . . . . . 379 53 9.1 1.3 389 36 9.4 0.9 10 0.2
Pennsylvania . . . . . . . . . . . . 1,299 119 10.4 1.0 1,188 96 9.5 0.8 *–111 *–0.9
Rhode Island . . . . . . . . . . . . 92 13 8.8 1.3 65 10 6.2 1.0 *–27 *–2.6
South Carolina. . . . . . . . . . . 704 62 13.7 1.2 620 61 12.1 1.2 *–84 *–1.6
South Dakota. . . . . . . . . . . . 94 20 10.9 2.4 73 9 8.5 1.1 *–21 *–2.4
Tennessee. . . . . . . . . . . . . . . 863 93 12.7 1.3 729 73 10.8 1.1 *–134 *–2.0
Texas . . . . . . . . . . . . . . . . . . . 3,718 191 12.9 0.7 3,587 193 12.5 0.7 *–132 *–0.5
Utah. . . . . . . . . . . . . . . . . . . . 231 56 7.2 1.8 228 50 7.1 1.6 –3 –0.1
Footnotes provided at end of table.
32 U.S. Census Bureau
Appendix Table 5.
Number and Percentage of People in Poverty by State: 3-Year Average of 2018–2020—Con.
(Numbers in thousands. Margin of error in thousands or percentage points as appropriate. People as of March of the
following year. Information on confidentiality protection, sampling error, nonsampling error, and definitions is available
at, see <https://www2.census.gov/programs-surveys/cps/techdocs/cpsmar21.pdf>)
Official1 SPM Difference
Number Percent Number Percent
State
Margin of Margin of Margin of Margin of
Estimate error2 (±) Estimate error2 (±) Estimate error2 (±) Estimate error2 (±) Number Percent
Vermont . . . . . . . . . . . . . . . . 55 7 8.9 1.2 51 7 8.3 1.1 –4 –0.7
Virginia . . . . . . . . . . . . . . . . . 741 95 8.8 1.1 846 106 10.1 1.3 *106 *1.3
Washington . . . . . . . . . . . . . 601 108 7.9 1.4 571 86 7.5 1.1 –30 –0.4
West Virginia. . . . . . . . . . . . 257 29 14.6 1.6 184 22 10.5 1.3 *–72 *–4.1
Wisconsin. . . . . . . . . . . . . . . 480 65 8.3 1.1 393 53 6.8 0.9 *–88 *–1.5
Wyoming. . . . . . . . . . . . . . . . 54 8 9.5 1.5 50 7 8.8 1.3 –3 –0.6
* An asterisk preceding an estimate indicates change is statistically different from zero at the 90 percent confidence level.
Z Rounds to zero.
1
Includes unrelated individuals under the age of 15.
2
A margin of error (MOE) is a measure of an estimate's variability. The larger the MOE in relation to the size of the estimate, the less reli-
able the estimate. This number, when added to and subtracted from the estimate, forms the 90 percent confidence interval. MOEs shown in
this table are based on standard errors calculated using replicate weights.
Notes: Details may not sum to totals due to rounding. The data for 2019 and 2020 reflect the implementation of revised Supplemental
Poverty Measure methodology. More information is available in the report appendix.
Source: U.S. Census Bureau, Current Population Survey, 2019 to 2021 Annual Social and Economic Supplements (CPS ASEC).
U.S. Census Bureau 33
Appendix Table 6.
Effect of Individual Elements on Supplemental Poverty Measure Rates: 2019 and 2020
(Margin of error in percentage points. People as of March of the following year. Information on confidentiality protection,
sampling error, nonsampling error, and definitions is available at <https://www2.census.gov/programs–surveys/cps
/techdocs/cpsmar21.pdf>)
All people Under 18 years 18 to 64 years 65 years and older
Element Margin of Margin of Margin of Margin of
Estimate error1 (±) Estimate error1 (±) Estimate error1 (±) Estimate error1 (±)
2020
All people . . . . . . . . . . . . . . . . . . . . 9.14 0.23 9.73 0.45 8.82 0.26 9.48 0.44
ADDITIONS
Social Security. . . . . . . . . . . . . . . . . . –8.12 0.20 –1.53 0.17 –3.50 0.18 –33.07 0.77
Economic Impact/stimulus2. . . . . . –3.58 0.17 –4.46 0.30 –3.23 0.17 –3.70 0.28
Unemployment insurance. . . . . . . . –1.70 0.11 –1.97 0.22 –1.95 0.12 –0.48 0.10
Refundable tax credits3. . . . . . . . . . –1.62 0.12 –3.75 0.30 –1.27 0.10 –0.09 0.04
SNAP4 and school lunch. . . . . . . . . –0.98 0.09 –1.79 0.21 –0.81 0.09 –0.55 0.09
SNAP4 . . . . . . . . . . . . . . . . . . . . . . . . . –0.89 0.08 –1.55 0.20 –0.74 0.08 –0.53 0.08
SSI4. . . . . . . . . . . . . . . . . . . . . . . . . . . . –0.83 0.07 –0.49 0.11 –0.93 0.08 –0.90 0.13
Housing subsidies. . . . . . . . . . . . . . . –0.74 0.07 –1.08 0.16 –0.58 0.06 –0.84 0.12
Child support received . . . . . . . . . . –0.16 0.03 –0.42 0.09 –0.10 0.02 –0.01 0.01
TANF/general assistance 4. . . . . . . . –0.14 0.03 –0.29 0.08 –0.12 0.03 –0.03 0.02
School lunch. . . . . . . . . . . . . . . . . . . . –0.09 0.03 –0.22 0.07 –0.07 0.02 –0.01 0.02
Workers’ compensation . . . . . . . . . –0.06 0.02 –0.06 0.04 –0.07 0.03 –0.02 0.02
Energy assistance. . . . . . . . . . . . . . . –0.04 0.02 –0.04 0.03 –0.05 0.02 –0.05 0.03
WIC4. . . . . . . . . . . . . . . . . . . . . . . . . . . –0.03 0.02 –0.09 0.05 –0.02 0.01 0.00 0.00
SUBTRACTIONS
Child support paid. . . . . . . . . . . . . . 0.06 0.02 0.08 0.03 0.06 0.02 0.04 0.03
Federal income tax. . . . . . . . . . . . . . 0.15 0.03 0.10 0.04 0.19 0.04 0.11 0.04
FICA4 . . . . . . . . . . . . . . . . . . . . . . . . . . 0.63 0.07 0.72 0.12 0.72 0.08 0.17 0.05
Work expenses . . . . . . . . . . . . . . . . . 0.78 0.08 0.97 0.15 0.86 0.08 0.23 0.05
Medical expenses. . . . . . . . . . . . . . . 1.53 0.09 1.17 0.14 1.35 0.09 2.65 0.22
20195
All people . . . . . . . . . . . . . . . . . . . . 11.78 0.27 12.65 0.49 11.18 0.31 12.76 0.46
ADDITIONS
Social Security. . . . . . . . . . . . . . . . . . –8.13 0.20 –1.99 0.21 –3.78 0.18 –32.07 0.78
Unemployment insurance. . . . . . . . –0.16 0.04 –0.19 0.06 –0.17 0.04 –0.10 0.05
Refundable tax credits . . . . . . . . . . –2.40 0.14 –5.69 0.36 –1.78 0.11 –0.21 0.06
SNAP4 and school lunch. . . . . . . . . –1.11 0.10 –2.13 0.23 –0.87 0.09 –0.60 0.10
SNAP4 . . . . . . . . . . . . . . . . . . . . . . . . . –0.80 0.08 –1.41 0.19 –0.65 0.08 –0.54 0.09
SSI4. . . . . . . . . . . . . . . . . . . . . . . . . . . . –0.85 0.07 –0.72 0.13 –0.89 0.08 –0.87 0.12
Housing subsidies. . . . . . . . . . . . . . . –0.67 0.06 –0.84 0.14 –0.53 0.06 –0.95 0.13
Child support received . . . . . . . . . . –0.20 0.04 –0.52 0.11 –0.14 0.03 –0.02 0.02
TANF/general assistance 4. . . . . . . . –0.08 0.03 –0.21 0.07 –0.06 0.02 –0.02 0.02
School lunch. . . . . . . . . . . . . . . . . . . . –0.34 0.06 –0.80 0.13 –0.25 0.05 –0.07 0.04
Workers’ compensation . . . . . . . . . –0.04 0.02 –0.02 0.02 –0.05 0.02 –0.05 0.04
Energy assistance. . . . . . . . . . . . . . . –0.06 0.02 –0.06 0.04 –0.05 0.02 –0.10 0.04
WIC4. . . . . . . . . . . . . . . . . . . . . . . . . . . –0.07 0.03 –0.16 0.07 –0.05 0.03 –0.01 0.01
SUBTRACTIONS
Child support paid. . . . . . . . . . . . . . 0.08 0.03 0.11 0.07 0.09 0.03 0.02 0.02
Federal income tax. . . . . . . . . . . . . . 0.28 0.04 0.18 0.06 0.36 0.06 0.09 0.03
FICA4 . . . . . . . . . . . . . . . . . . . . . . . . . . 1.21 0.11 1.77 0.22 1.26 0.11 0.29 0.07
Work expenses . . . . . . . . . . . . . . . . . 1.52 0.12 2.16 0.23 1.57 0.12 0.46 0.09
Medical expenses. . . . . . . . . . . . . . . 2.28 0.12 1.81 0.19 1.96 0.13 4.04 0.29
1
A margin of error (MOE) is a measure of an estimate's variability. The larger the MOE in relation to the size of the estimate, the less reli-
able the estimate. This number, when added to and subtracted from the estimate, forms the 90 percent confidence interval. MOEs shown in
this table are based on standard errors calculated using replicate weights.
2
Includes the first two rounds of stimulus payments. Additional details available in the report appendix.
3
Refundable tax credits do not include stimulus payments.
4
SNAP: Supplemental Nutrition Assistance Program; SSI: Supplemental Security Income; TANF: Temporary Assistance for Needy
Families; WIC: Special Supplemental Nutrition Program for Women, Infants, and Children; FICA: Federal Insurance Contributions Act.
5
The data for 2019 reflect the implementation of revised Supplemental Poverty Measure methodology. More information is available in
the report appendix.
Note: Details may not sum to totals due to rounding.
Source: U.S. Census Bureau, Current Population Survey, 2020 and 2021 Annual Social and Economic Supplements (CPS ASEC).
34 U.S. Census Bureau
Appendix Table 7.
Effect of Individual Elements on the Number of Individuals in Poverty: 2019 and 2020
(Numbers and margin of error in thousands. People as of March of the following year. Information on confidentiality
protection, sampling error, nonsampling error, and definitions are available at <https://www2.census.gov/programs
-surveys/cps/techdocs/cpsmar21.pdf>)
All people Under 18 years 18 to 64 years 65 years and older
Element Margin of Margin of Margin of Margin of
Number error1 (±) Number error1 (±) Number error1 (±) Number error1 (±)
2020
All people . . . . . . . . . . . . . . . . . . . . 29,805 766 7,079 325 17,433 521 5,293 246
ADDITIONS
Social Security. . . . . . . . . . . . . . . . . . –26,502 657 –1,115 123 –6,923 350 –18,464 435
Economic Impact/stimulus2. . . . . . –11,684 543 –3,246 221 –6,373 328 –2,065 156
Unemployment insurance. . . . . . . . –5,545 359 –1,430 157 –3,845 243 –270 58
Refundable tax credits3. . . . . . . . . . –5,281 391 –2,729 220 –2,505 192 –48 20
SNAP4 and school lunch. . . . . . . . . –3,209 301 –1,302 156 –1,601 170 –306 48
SNAP4 . . . . . . . . . . . . . . . . . . . . . . . . . –2,888 276 –1,131 144 –1,460 158 –298 46
SSI4. . . . . . . . . . . . . . . . . . . . . . . . . . . . –2,698 222 –354 80 –1,839 161 –505 71
Housing subsidies. . . . . . . . . . . . . . . –2,407 228 –785 119 –1,154 127 –468 67
Child support received . . . . . . . . . . –514 108 –305 69 –205 43 –4 5
TANF/general assistance 4. . . . . . . . –472 105 –211 58 –243 60 –18 10
School lunch. . . . . . . . . . . . . . . . . . . . –298 89 –157 48 –136 46 –5 9
Workers’ compensation . . . . . . . . . –193 75 –41 27 –142 55 –10 9
Energy assistance. . . . . . . . . . . . . . . –142 54 –27 18 –90 42 –26 14
WIC4. . . . . . . . . . . . . . . . . . . . . . . . . . . –102 53 –66 34 –35 20 0 0
SUBTRACTIONS
Child support paid. . . . . . . . . . . . . . 197 57 58 25 119 32 20 15
Federal income tax. . . . . . . . . . . . . . 504 95 75 30 370 74 59 24
FICA4 . . . . . . . . . . . . . . . . . . . . . . . . . . 2,039 225 525 88 1,417 166 96 26
Work expenses . . . . . . . . . . . . . . . . . 2,537 246 704 111 1,707 168 127 30
Medical expenses. . . . . . . . . . . . . . . 5,002 292 854 102 2,667 186 1,482 122
20195
All people . . . . . . . . . . . . . . . . . . . . 38,300 876 9,253 358 22,073 603 6,975 251
ADDITIONS
Social Security. . . . . . . . . . . . . . . . . . –26,454 652 –1,454 153 –7,474 352 –17,525 432
Unemployment insurance. . . . . . . . –524 119 –142 42 –329 78 –53 26
Refundable tax credits . . . . . . . . . . –7,792 446 –4,161 263 –3,514 212 –117 31
SNAP4 and school lunch. . . . . . . . . –3,600 317 –1,561 169 –1,710 171 –328 55
SNAP4 . . . . . . . . . . . . . . . . . . . . . . . . . –2,603 267 –1,030 140 –1,280 151 –293 51
SSI4. . . . . . . . . . . . . . . . . . . . . . . . . . . . –2,760 242 –529 94 –1,758 167 –473 68
Housing subsidies. . . . . . . . . . . . . . . –2,189 198 –614 101 –1,053 115 –522 73
Child support received . . . . . . . . . . –665 132 –382 79 –271 59 –12 9
TANF/general assistance 4. . . . . . . . –276 83 –151 51 –112 36 –13 11
School lunch. . . . . . . . . . . . . . . . . . . . –1,116 184 –582 94 –494 94 –39 21
Workers’ compensation . . . . . . . . . –142 52 –18 13 –95 38 –30 20
Energy assistance. . . . . . . . . . . . . . . –189 64 –43 27 –90 38 –56 23
WIC4. . . . . . . . . . . . . . . . . . . . . . . . . . . –227 106 –120 54 –102 51 –5 7
SUBTRACTIONS
Child support paid. . . . . . . . . . . . . . 273 100 79 50 181 56 13 10
Federal income tax. . . . . . . . . . . . . . 896 135 130 41 719 110 47 18
FICA4 . . . . . . . . . . . . . . . . . . . . . . . . . . 3,929 341 1,292 157 2,479 220 158 38
Work expenses . . . . . . . . . . . . . . . . . 4,935 376 1,577 168 3,109 238 249 50
Medical expenses. . . . . . . . . . . . . . . 7,400 402 1,326 136 3,867 253 2,206 158
1
A margin of error (MOE) is a measure of an estimate's variability. The larger the MOE in relation to the size of the estimate, the less reli-
able the estimate. This number, when added to and subtracted from the estimate, forms the 90 percent confidence interval. MOEs shown in
this table are based on standard errors calculated using replicate weights.
2
Includes the first two rounds of stimulus payments. Additional details available in the report appendix.
3
Refundable tax credits do not include stimulus payments.
4
SNAP: Supplemental Nutrition Assistance Program; SSI: Supplemental Security Income; TANF: Temporary Assistance for Needy
Families; WIC: Special Supplemental Nutrition Program for Women, Infants, and Children; FICA: Federal Insurance Contributions Act.
5
The data for 2019 reflect the implementation of revised Supplemental Poverty Measure methodology. More information is available in
the report appendix.
Note: Details may not sum to totals due to rounding.
Source: U.S. Census Bureau, Current Population Survey, 2020 and 2021 Annual Social and Economic Supplements (CPS ASEC).
U.S. Census Bureau 35
Appendix Table 8.
Comparison of 2019 Supplemental Poverty Estimates Using Production and Revised
Supplemental Poverty Measure Files
(Numbers in thousands. Margin of error in thousands or percentage points as appropriate. People as of March of the
following year. Information on confidentiality protection, sampling error, nonsampling error, and definitions is available
at <https://www2.census.gov/programs-surveys/cps/techdocs/cpsmar20.pdf>)
SPM 2019—Revised1 SPM 2019—Published Difference
Number Percent Number Percent
Characteristic
Margin of Margin of Margin of Margin of
Estimate error2 (±) Estimate error2 (±) Estimate error2 (±) Estimate error2 (±) Number Percent
All people. . . . . . . . . 38,300 876 11.8 0.3 38,163 895 11.7 0.3 138 Z
Sex
Male. . . . . . . . . . . . . . . . . . . . . . 17,666 485 11.1 0.3 17,655 489 11.1 0.3 11 Z
Female. . . . . . . . . . . . . . . . . . . . 20,634 498 12.4 0.3 20,508 508 12.4 0.3 *126 *0.1
Age
Under 18 years. . . . . . . . . . . . . 9,253 358 12.6 0.5 9,119 354 12.5 0.5 *134 *0.2
18 to 64 years. . . . . . . . . . . . . . 22,073 603 11.2 0.3 22,072 606 11.2 0.3 1 Z
65 years and older. . . . . . . . . 6,975 251 12.8 0.5 6,972 258 12.8 0.5 3 Z
Type of Unit
Married couple. . . . . . . . . . . . . 13,479 620 6.8 0.3 13,703 625 6.9 0.3 *–224 *–0.1
Cohabiting partners. . . . . . . . 3,147 330 12.1 1.2 3,167 324 12.1 1.1 –20 –0.1
Female reference person . . . 9,732 474 23.8 1.0 9,526 478 23.3 1.0 *207 *0.5
Male reference person. . . . . . 2,141 236 14.2 1.5 2,162 237 14.3 1.5 –21 –0.1
Unrelated individuals. . . . . . . 9,801 319 21.6 0.6 9,605 320 21.2 0.6 *196 *0.4
Race3 and Hispanic Origin
White. . . . . . . . . . . . . . . . . . . . . 26,029 645 10.5 0.3 26,089 669 10.5 0.3 –60 Z
White, not Hispanic . . . . . . 15,921 477 8.2 0.2 15,914 492 8.2 0.3 8 Z
Black . . . . . . . . . . . . . . . . . . . . . 8,144 409 18.9 1.0 7,907 408 18.3 0.9 *237 *0.5
Asian . . . . . . . . . . . . . . . . . . . . . 2,257 189 11.3 0.9 2,327 191 11.7 1.0 *–70 *–0.4
Hispanic (any race). . . . . . . . . 11,437 478 18.8 0.8 11,464 475 18.9 0.8 –27 Z
Nativity
Native-born . . . . . . . . . . . . . . . 30,460 733 10.9 0.3 30,238 761 10.8 0.3 *222 *0.1
Foreign-born. . . . . . . . . . . . . . 7,840 322 17.5 0.7 7,924 318 17.6 0.7 *–84 *–0.2
Naturalized citizen . . . . . . . 3,076 200 13.5 0.9 3,109 197 13.7 0.9 –33 –0.1
Not a citizen. . . . . . . . . . . . . 4,764 259 21.5 1.2 4,815 261 21.7 1.2 –51 –0.2
Educational Attainment
Total, aged 25 and
older. . . . . . . . . . . . . . . 24,312 547 10.9 0.2 24,319 558 10.9 0.2 –7 Z
No high school diploma . . . . 5,600 261 27.7 1.1 5,563 267 27.5 1.1 37 0.2
High school, no college. . . . . 8,588 305 13.9 0.5 8,543 319 13.9 0.5 45 0.1
Some college. . . . . . . . . . . . . . 5,449 231 9.5 0.4 5,472 238 9.5 0.4 –23 Z
Bachelor’s degree or
higher . . . . . . . . . . . . . . . . . . . 4,674 214 5.6 0.2 4,741 219 5.7 0.3 *–67 *–0.1
Tenure
Owner/mortgage . . . . . . . . . . 7,035 389 5.1 0.3 7,146 385 5.2 0.3 *–111 *–0.1
Owner/no mortgage/rent-
free. . . . . . . . . . . . . . . . . . . . . . 9,726 424 10.9 0.4 10,057 458 11.2 0.5 *–331 *–0.4
Renter . . . . . . . . . . . . . . . . . . . . 21,540 760 22.1 0.7 20,960 765 21.5 0.7 *580 *0.6
Residence4
Inside metropolitan
statistical areas. . . . . . . . . . . 33,359 871 11.8 0.3 33,426 881 11.8 0.3 –67 Z
Inside principal cities. . . . . 15,654 673 14.9 0.6 15,527 672 14.8 0.6 127 0.1
Outside principal cities . . . 17,705 631 10.0 0.3 17,898 640 10.1 0.3 *–194 *–0.1
Outside metropolitan
statistical areas. . . . . . . . . . . 4,941 458 11.6 0.7 4,737 442 11.2 0.7 *205 *0.5
Region
Northeast. . . . . . . . . . . . . . . . . 6,341 366 11.5 0.7 6,431 383 11.7 0.7 –90 –0.2
Midwest. . . . . . . . . . . . . . . . . . . 6,081 355 9.0 0.5 5,944 347 8.8 0.5 *137 *0.2
South. . . . . . . . . . . . . . . . . . . . . 15,662 644 12.6 0.5 15,466 628 12.4 0.5 *195 *0.2
West. . . . . . . . . . . . . . . . . . . . . . 10,216 405 13.1 0.5 10,321 402 13.2 0.5 –105 –0.1
Footnotes provided at end of table.
36 U.S. Census Bureau
Appendix Table 8.
Comparison of 2019 Supplemental Poverty Estimates Using Production and Revised
Supplemental Poverty Measure Files—Con.
(Numbers in thousands. Margin of error in thousands or percentage points as appropriate. People as of March of the
following year. Information on confidentiality protection, sampling error, nonsampling error, and definitions is available
at <https://www2.census.gov/programs-surveys/cps/techdocs/cpsmar20.pdf>)
SPM 2019—Revised1 SPM 2019—Published Difference
Number Percent Number Percent
Characteristic
Margin of Margin of Margin of Margin of
Estimate error2 (±) Estimate error2 (±) Estimate error2 (±) Estimate error2 (±) Number Percent
Health Insurance Coverage
With private insurance. . . . . . 11,933 500 5.4 0.2 12,202 491 5.5 0.2 *–269 *–0.1
With public, no private
insurance . . . . . . . . . . . . . . . . 19,993 534 25.8 0.6 19,600 556 25.3 0.6 *393 *0.5
Not insured. . . . . . . . . . . . . . . . 6,375 323 23.8 1.1 6,361 322 23.7 1.1 14 0.1
Work Experience
Total, 18 to 64 years.. . 22,073 603 11.2 0.3 22,072 606 11.2 0.3 1 Z
All workers. . . . . . . . . . . . . . . . 10,491 365 6.8 0.2 10,599 368 6.9 0.2 *–108 *–0.1
Worked full-time,
year-round. . . . . . . . . . . . . . . 4,371 225 3.9 0.2 4,487 226 4.0 0.2 *–116 *–0.1
Less than full-time,
year-round. . . . . . . . . . . . . . . 6,120 255 14.6 0.5 6,112 256 14.6 0.5 8 Z
Did not work at least
1 week . . . . . . . . . . . . . . . . . . . 11,582 388 27.0 0.8 11,473 390 26.8 0.8 *109 *0.3
Disability Status5
Total, 18 to 64 years.. . 22,073 603 11.2 0.3 22,072 606 11.2 0.3 1 Z
With a disability . . . . . . . . . . . 3,203 164 22.2 1.1 3,107 168 21.5 1.2 *96 *0.7
With no disability. . . . . . . . . . 18,804 554 10.3 0.3 18,899 556 10.4 0.3 –95 –0.1
* An asterisk preceding an estimate indicates change is statistically different from zero at the 90 percent confidence level.
Z Rounds to zero.
1
The data for 2019 reflect the implementation of revised Supplement Poverty Measure methodology. More information is available in the
report appendix.
2
A margin of error (MOE) is a measure of an estimate's variability. The larger the MOE in relation to the size of the estimate, the less reli-
able the estimate. This number, when added to and subtracted from the estimate, forms the 90 percent confidence interval. MOEs shown in
this table are based on standard errors calculated using replicate weights.
3
Federal surveys give respondents the option of reporting more than one race. Therefore, two basic ways of defining a race group are
possible. A group, such as Asian, may be defined as those who reported Asian and no other race (the race-alone or single-race concept) or
as those who reported Asian regardless of whether they also reported another race (the race-alone-or-in-combination concept). This table
shows data using the first approach (race alone). The use of the single-race population does not imply that it is the preferred method of pre-
senting or analyzing data. The Census Bureau uses a variety of approaches. Data for American Indians and Alaska Natives, Native Hawaiians
and Other Pacific Islanders, and those reporting two or more races are not shown separately.
4
Information on metropolitan statistical areas and principal cities is available at <www.census.gov/programs-surveys/metro-micro/about
/glossary.html>.
5
The sum of those with and without a disability does not equal the total because disability status is not defined for individuals in the
U.S. armed forces.
Note: Details may not sum to totals due to rounding.
Source: U.S. Census Bureau, Current Population Survey, 2020 Annual Social and Economic Supplement (CPS ASEC).
U.S. Census Bureau 37
File and source
- File
- census-p60-275-spm-2020.pdf
- Size
- 920,468 bytes
- SHA-256
- 43d7a9f4f67fd67d08114df4a503c92c672fdf0fa40a882f806a03e1e3333a0d
- Our copy
- census-p60-275-spm-2020.pdf
- Original
- www.census.gov