Bureau of the Census, U.S.

Department of Commerce

Small Area Income and

Poverty Estimates: 2015
December 14, 2016

Small Area Income and Poverty Estimates (SAIPE): 2015

Introduction
This document presents summary highlights of the 2015
data released by the Small Area Income and Poverty
Estimates (SAIPE) program of the U.S. Census Bureau in
December 2016. Each year, the SAIPE program provides
timely, reliable estimates of income and poverty for the
administration of Federal programs and the allocation of
Federal funds to local jurisdictions and school districts.
SAIPE is the only source of data on single-year median
household income and poverty statistics for the 3,141
counties and 13,236 school districts in the United States. 1
Some state and local programs also use SAIPE income
and poverty estimates to distribute funds and manage
programs.
The Census Bureau and other Federal agencies created
the SAIPE program to provide annual income and poverty
statistics for states, counties, and school districts in the
United States. 2 The SAIPE program produces yearly
poverty estimates for the total population (all ages) and by
selected characteristics for counties and states. These
estimates include the number of children under age 5 in
poverty (for states only), the number of related-children
aged 5 to 17 in families in poverty, the number of children
under age 18 in poverty, and median household income.
The Department of Education uses SAIPE data to aid in
determining annual Title I allocations of Federal funds to
states and school districts. At the school district level,
estimates are generated for the total population, the
number of children aged 5 to 17, and the number of
related-children aged 5 to 17 in families in poverty.
Due to the comprehensive geographic coverage and oneyear focus, SAIPE data can be used to analyze
geographic variation in poverty and income, as well as
changes over time. The purpose of this document is to
highlight several key aspects from such analysis. 3

Highlights

Median household income at the county level ranged

from $22,894 to $125,900 with a median county-level
value of $46,800.

Based on poverty rate estimates for the 3,141 counties

for all ages, 18 percent of counties (569) had a
statistically significant increase in poverty between
2007, the year before the most recent recession, and
2015. Only two percent of counties (63) had a
statistically significant decrease in poverty during that
period.

Small Area Income and Poverty Estimates (SAIPE) are

model-based enhancements of the American Community
Survey (ACS) estimates created by integrating additional
information from administrative records, intercensal
population estimates, and decennial census data. SAIPE
methodology employs statistical modeling techniques to
combine this supplemental information with survey data to
produce estimates that are more reliable. SAIPE are
broadly consistent with the direct ACS estimates, but with
help from other data sources, SAIPE estimates are more
precise than the ACS 1-year and 5-year survey estimates
for most counties and school districts. ACS 1-year
estimates are not available for most of these smaller
geographic areas (approximately only 800 counties with a
population of 65,000 or more are included in the ACS 1year estimates). A 2015 ACS map of unpublished counties
is available at:
https://www.census.gov/did/www/saipe/data/statecounty/m
aps/2015.html.
Additional detailed information on the various input data
sources used in producing SAIPE is available at:
http://www.census.gov/did/www/saipe/data/model/index.ht
ml.
SAIPE estimates are subject to several types of
uncertainty. Details on SAIPE methodology are available
at:
http://www.census.gov/did/www/saipe/methods/index.html.

There were 3,142 total counties in the United States. Kalawao County, HI
was omitted due to small sample size. There were also 13,245 school districts
in the United States; however, nine were excluded due to lack of school-age
children.
2
For more information on the creation of the SAIPE program visit:
https://www.census.gov/did/www/saipe/about/origins.html
3
All data shown are estimates containing uncertainty. Unless specifically
noted in the text, apparent differences among the estimates may not be
statistically significant. All direct comparisons cited in the text have been
statistically tested at the 10 percent significance level. See text box on page 4
for additional information on the sources of uncertainty.

County-level Estimates

Figure 1a. Map insert (Boston MSA to DC MSA): 2015

Median Household Income

The 2015 SAIPE provides estimates for 3,141 counties in
the United States. At the county level, median household
income ranged from $22,894 to $125,900 with a median
value of $46,800. 4
Figure 1 highlights the range of median household income
throughout the United States. Seventy-seven counties had
a median income within the highest range ($81,129 to
$125,900). Forty-two of these high-income counties were
located in the Northeast region, Maryland, and Virginia.
Figure 1a depicts the metropolitan statistical areas (MSA)
which includes Boston, New York, Philadelphia, Baltimore
and Washington, D.C. 5 There were 33 high-income
counties located within this corridor. Eighty percent of
counties in the lowest income range ($22,894 to $38,139)
were located in the South.

Figure 1. Median Household Income of the Total Population by County: 2015

The median of county-level values ($46,800) is not the same measure as

the median household income in the United States. The legends in Figures 1
and 1a show the median household income for the nation ($55,775).
5
A reference map on metro/micro area status is available at:
http://www.census.gov/did/www/saipe/data/statecounty/maps/2015.html.

From 2014 to 2015, more counties experienced an

increase in median household income than a decrease.
Figure 2 shows the percent change in median household
income between 2014 and 2015. All changes were
adjusted for inflation using the national Consumer Price
Index (CPI-U) and are reported as calendar year 2015
dollars.
Seventy-four percent of counties (2,314) had an increase
in their median household income, only 14.0 percent (440
counties) had a statistically significant increase. During the
same period, 26.3 percent (827 counties) had a decrease
in median household income, only 1.6 percent (51
counties) had a statistically significant decrease. In 22
states, more than fifteen percent of counties have a
statistically significant increase in median household
income between 2014 and 2015.
However, when comparing 2015 with 2007, the year
before the most recent recession, more counties had a
decrease in median household income than an increase. 6
Figure 3 shows the percent change in median household
income between 2007 and 2015.

Household income includes income of the householder

and all other people 15 years and older in the household,
whether or not they are related to the householder.
Median is the point that divides the household income
distributions into halves, one-half with income above the
median and the other with income below the median. The
median is based on the income distribution of all
households, including those with no income.
Related-children aged 5 to 17 in families denotes
children who are related to householder by birth,
marriage, or adoption. Foster children are not included in
families.
School-age population refers to children aged 5 to 17
who live within the geographic boundaries of a school
district and who are in an appropriate grade range. It is
not a measure of school district enrollment.

Figure 2. Percent Change in Median Household Income of the Total Population by County: 2014 to 2015

6
The National Bureau of Economic Research (NBER) is the official source for
recession timing. The NBER pinpoints December 2007 and June 2009 as the
beginning and end of the most recent recession.

Figure 3. Percent Change in Median Household Income of the Total Population by County: 2007 to 2015

Between 2007 and 2015, 58.5 percent (1,836 counties)

displayed a percent change in median household income
ranging from -29.0 to 0.0 percent. In the same period, 41.6
percent of counties had a percent change ranging from
0.1 to 74.7 percent. Only 47 counties displayed a percent
change in median household income in the top range of
28.5 percent or more, with 39 of these counties located in
the Great Plains states of Nebraska, North Dakota, and
Texas.
Of the 3,141 counties in the United States, 25.5 percent
(802 counties) had a statistically significant change over
the eight-year period. Of these, 57.9 percent (464
counties) had decreases in median household income.
Clusters of counties with statistically significant decreases
in median household income exist throughout all regions.
Forty-two percent (338 counties) had a statistically
significant increase in median household income between
2007 and 2015. North Dakota, South Dakota, and
Nebraska contained the highest proportions of counties
with median household income gains: 84.9 percent (45
counties), 50.0 percent (33 counties), and 50.5 percent
(47 counties), respectively.

Poverty
The SAIPE data also include poverty estimates for all
counties in the United States. In 2015, county poverty
rates for all ages ranged from 3.4 percent to 47.4 percent
across counties. Figure 4 shows how poverty rates varied

among counties throughout the United States. Counties

with
higher
poverty
rates
were
concentrated
predominately in the South. Eighteen percent of counties
within the South had poverty rates in the top ranges of the
map legend (from 24.7 up to 47.4 percent), while each
other region (West, Midwest, and Northeast) had no more
than 6.0 percent of their counties with poverty rates in the
top ranges. Most counties with lower poverty rates were
concentrated in the Northeast and Midwest where over
70.0 percent of counties in each region were in the bottom
ranges of the map legend (from 3.4 to 14.6 percent).
Figure 5 shows the poverty rate by county for relatedchildren aged 5 to 17 in families. In 2015, 54.7 percent
(1,717 counties) had poverty rates for related-children
aged 5 to 17 above the groups U.S. poverty rate of 19.5
percent. Among these counties, 60.9 percent (1,045
counties) were statistically different from the national
poverty rate.
In Alabama, Georgia, Louisiana, Mississippi, and New
Mexico, 70 percent or more of their counties poverty rates
were statistically greater than the national average.
Seventy percent or more of counties in eight states had
poverty rates for school-age children statistically lower
than
the
national
rate:
Connecticut,
Hawaii,
Massachusetts, Minnesota, New Hampshire, North
Dakota, Rhode Island, and Wyoming.
Thirty-four percent (1,063 counties) fell into the top two

Figure 4. Poverty Rates of the Total Population by County: 2015

Figure 5. Poverty Rates of the School-Age Population by County: 2015

ranges between 24.7 and 60.0 percent. Of the 1,063

counties, 73.8 percent (785 counties) were located in the
South. Six states in the South had 65 percent or more of
their counties within the top two ranges: Alabama,
Arkansas, Georgia, Louisiana, Mississippi, and South
Carolina.

poverty rates greater than 20 percent, but only 10.0

percent (324 counties) were statistically greater than 20
percent. In the Midwest, 7.5 percent of counties had
poverty rates of 20 percent or more; in the Northeast, 2.8
percent; in the South, 41.4 percent; and in the West, 18.1
percent.

Change in County Poverty Rates

How is poverty measured?

From 2014 to 2015, more counties had a decrease in their

poverty rate than an increase. Figure 6 shows the change
in county-level poverty rates for the total population
between 2014 and 2015. Sixty-six percent (2,061
counties) experienced a decrease in their poverty rate,
only 7.8 percent (244 counties) had a statistically
significant decrease. Thirty-four percent (1,080 counties)
had an increase in their poverty rate, only 1.5 percent (46
counties) had a statistically significant increase.

Poverty status is determined by comparing total annual

family before-tax income to a table of federal poverty
thresholds that vary by family size, number of related
children, and age of householder. If a familys income is
less than the dollar value of the appropriate threshold, then
that family and every individual in it are considered to be in
poverty. For people not living in families, poverty status is
determined by comparing the individuals total income to
their threshold.

However, more counties have higher poverty rates in 2015

than in 2007, the year before the most recent recession.
Figure 7 shows the change in county-level poverty rates
for the total population between 2007 and 2015. Seventythree percent (2,293 counties) displayed changes in their
poverty rates ranging between 0.1 and 15.9 percentage
points. Twenty-seven percent (848 counties) displayed a
change in their poverty rate ranging from -17.2 to 0.0
percentage points. There were only 109 counties in the
top legend range between 5.5 and 15.9 percentage points.
Of these, 82.6 percent (90 counties) were in the South,
with 30 counties in Georgia and 18 in Florida.

For more general information on poverty, please see:

http://www.census.gov/topics/income-poverty/poverty.html.

Of the 3,141 counties in the United States, 20.1 percent

(632 counties) had statistically significant changes in their
poverty rates between 2007 and 2015. Of these counties,
90.0 percent (569 counties) showed a statistically
significant increase in poverty rate during the eight-year
period. Approximately 10.0 percent (63 counties) had a
statistically significant decrease in poverty rate during this
period.

Poverty by Region and Metro Status

Figure 8 compares the share of the total population and
people in poverty by region and metro status. Among
those living in poverty, 41.1 percent live in the South; in
the West, 23.8 percent; in the Midwest, 19.7 percent; and
in the Northeast, 15.5 percent. This is similar to the share
of the total U.S. population that lives in each region.
Eighty-three percent of the people in poverty live in
metropolitan areas, 9.8 percent live in micropolitan areas,
and 6.9 percent live in non-metro/micro areas.
Figure 9 depicts county-level poverty data for all ages by
region and the largest 25 metropolitan areas. The lighter
shaded counties have poverty rates less than 20 percent,
while the darker shaded counties have poverty rates of 20
percent or more. Twenty-four percent (755 counties) had

The table of federal poverty thresholds is updated annually

by the U.S. Census Bureau to allow for changes in the cost
of living using the Consumer Price Index (CPI-U). The
thresholds do not vary geographically.
SAIPEs primary input is the estimates of poverty from the
American Community Survey (ACS), a monthly survey with
people responding throughout the year. Since income is
reported for the previous 12 months, the appropriate
poverty threshold for each family is determined by
multiplying the base-year poverty threshold (1982) by the
average of the monthly CPI values for the 12 months
preceding the survey.
For more information, see How the Census Bureau
Measures Poverty at:
http://www.census.gov/topics/incomepoverty/poverty/guidance/poverty-measures.html.

What are the sources of statistical uncertainty?

All data shown are estimates containing uncertainty.
Sources of uncertainty include model error, sampling error,
and non-sampling error. Confidence intervals for all state
and county estimates are available at:
http://www.census.gov/did/www/saipe/data/statecounty/ind
ex.html.
Guidance on the uncertainty contained in the school district
estimates is available at:
http://www.census.gov/did/www/saipe/methods/schooldistri
ctuncertainty.html. Unless specifically noted in the text,
apparent differences among the estimates may not be
statistically significant. All direct comparisons cited in the
text have been statistically tested at the 10 percent
significance level.

Figure 6. Change in the Poverty Rates of the Total Population by County: 2014 to 2015

Figure 7. Change in the Poverty Rates of the Total Population by County: 2007 to 2015

Figure 8. Share of the Total Population and People in Poverty by Region and Metro Status: 2015

Figure 9. Counties with Poverty Rates Above or Below 20 Percent of the Total Population by Region: 2015

School District Level Estimates

Boundary Updates
To estimate the number of children living in poverty within
a school district, the SAIPE program must obtain the most
recent school district boundary updates. The SAIPE
program provides funding for the Census Bureaus
Geography Division to collect, update, and release
biennial school district boundaries through its School
District Review Program (SDRP).
In the latest SDRP update, there are 13,245 U.S. public
school districts compared with 13,486 in the previous
SDRP update. This net decrease of 241 school districts
reflects the deletion of 329 previously defined school
districts and the creation of 88 school districts. Since
2007, there has been a net decrease of 509 school
districts. Changes in the number of school districts are
typically the result of school districts shifting, splitting, or
consolidating boundaries, which are often driven by state
or local policy changes.

Poverty
The 2015 SAIPE data utilize the newly updated school
district boundaries effective as of January 1, 2016. This
accounts for all school districts in the Title I universe. 7
Since nine school districts did not have any school-age
children, these districts were excluded from the analysis
(13,245 school districts in universe, but 13,236 school
districts in the analysis).
Figure 10 shows the distribution of the number of school
districts, the number of school-age children, and the
number of school-age children in families in poverty by
school district resident population size. School-age
children, including school-age children in families in
poverty, often were concentrated in school districts with a
population of 20,000 or more. In 2015, an estimated 25.8
percent of school districts had a total population size of
20,000 or more. These school districts contained an
estimated 81.6 percent of all school-age children in the
nation and an estimated 82.6 percent of school-age
children in poverty.
Figure 11 shows the distribution of school-age children
(aged 5 to 17) living in families in poverty by school
district. This map provides an overview of the variation in
poverty throughout the United States by school district. 8
The lighter colors show school districts with lower poverty
rates (0.0 to 19.4 percent) and the darker colors show
school districts with higher poverty rates (19.5 to 100.0
7 The Title I universe is the set of U.S. school districts for which Title I of the
Elementary and Secondary Education Act pertains.
8 For a large-scale view, refer to the following link:
http://www.census.gov/did/www/saipe/data/statecounty/maps/2015.html.

percent). School districts with high and low poverty rates

are scattered throughout the nation, with some clustering
within regions.

Why are the Small Area Income and Poverty

Estimates Important?
The SAIPE data are designed primarily for use in the
U.S. Department of Educations annual Title I allocations
of Federal funds to states and school districts. Most
school districts in the United States, about 93 percent,
have a total population less than 65,000 and so do not
have ACS 1-year estimates available. The SAIPE
program was designed specifically to provide estimates
for school districts in the United States on a yearly
basis.
For additional detailed information on the use of SAIPE
estimates, please visit the FAQ webpage at:
http://www.census.gov/did/www/saipe/about/faq.html.
The SAIPE main page is located at:
http://www.census.gov/did/www/saipe/index.html.
Additional information is available by data release year
from 2005 to 2015. For example, annual highlights
reports, datasets, maps, figures, and ranking tables can
be downloaded from the SAIPE webpage at:
http://www.census.gov/did/www/saipe/data/index.html
or
http://www.census.gov/did/www/saipe/data/highlights/in
dex.html.
The online SAIPE Interactive Data Tool provides
detailed customized data tables by selected year(s) from
1989-2015, geography (state, county, and school
districts), poverty characteristics (all ages, under age 18,
aged 5-17 in families, under age 5) and median
household income. Data at the school district level are
available by total population, number of school-age
children (aged 5-17), and the number of school-age
children (aged 5-17) in families in poverty. Maps
showing school district boundaries are also available.
These custom tables can be downloaded to a PDF or
CSV file. The interactive data tool can be accessed
online at:
http://www.census.gov/did/www/saipe/data/index.html.
For video tutorials on SAIPE methodology, see:
http://www.census.gov/did/www/saipe/methods/index.ht
ml.

Figure 10. Distribution of School Districts, School-Age Children and School-Age Children in Families in Poverty
by School District Population: 2015

Figure 11. Percentage of School-Age Children in Families in Poverty by School District: 2015

10

Acknowledgements
The Small Area Estimates Branch of the Census Bureau
prepared this document.

Contact
For questions related to the contents of this document,
including the SAIPE program's estimates and
methodology, contact the Small Area Estimates Branch at:
(301) 763-3193 or [email protected].
For questions related to income and poverty definitions,
the American Community Survey, or other Census Bureau
surveys, contact the Census Bureau call center at 1-800923-8282 (toll free) or visit ask.census.gov for further
information.
A related program to SAIPE is the Small Area Health
Insurance Estimates (SAHIE) program, which produces
estimates of health insurance coverage for all counties
and states. Information about the SAHIE program is
available at
https://www.census.gov/did/www/sahie/index.html.

Suggested Citation
U.S. Census Bureau, Small Area Income and Poverty
Estimates (SAIPE): 2015 Highlights, U.S. Government
Publishing Office Washington, D.C.

Income and Poverty Data Sources Available

from the Census Bureau
SAIPE is one of several sources of income and poverty
data available from the Census Bureau. Other sources
include the Annual Social and Economic Supplement to
the Current Population Survey (CPS ASEC), the
American Community Survey (ACS), the Survey of
Income and Program Participation (SIPP), and the
Census 2000 Long-Form. Each of these sources differs
from the others in various ways, such as the length and
detail of its questionnaire, the number of households
included (sample size), and the methodology used to
collect and process the data. With its detailed
questionnaire, the CPS ASEC is the source of both the
official national estimates of poverty rates and of widely
used estimates of the distribution of household income
and individual earnings. The CPS ASEC provides a
consistent historical time series at the national level
beginning in 1959 and can also be used to look at statelevel trends and differences (through multi-year
averages) beginning in 1980.
Since 2006, the ACS has released annual subnational
estimates of income and poverty for places, counties,
and metropolitan statistical areas with a population of at
least 65,000 as well as for all states and the nation. The
sample size of the ACS is about 3.5 million addresses
per year, making this survey exceptionally useful for
subnational analyses. Three-year ACS estimates were
made available for 2008 through 2013 for areas and
subpopulations as small as 20,000. Five-year ACS
estimates became available for census tracts/block
groups and for small subgroups of the population
starting in 2010. More information on the ACS is located
at: http://www.census.gov/acs/www/.
The SIPP is useful for understanding the dynamics of
income and poverty (changes in income and poverty
rates for the same households over three or four years)
and for examining the nature and frequency of poverty
spells. The SIPP also permits researchers to look at
monthly or quarterly changes in income and poverty.
Decennial Census Long-Form estimates offer the best
measure of change between 1990 and 2000 for
subnational areas and for subpopulations. Since the
ACS replaced the Long-Form, the 2010 Census does
not provide income and poverty estimates. Since 2010,
ACS 5-year estimates provide data at the census tract
level that are comparable to earlier Decennial Census
estimates.

11