Data Sources in Census Studio
Census Studio provides access to five major Census Bureau data programs. Jump to any section:
| Data Source |
What It Covers |
Update Frequency |
Smallest Geography |
| ACS |
Demographics, income, housing, education, employment |
Annual (1-year and 5-year) |
Block Group (5-year) |
| Decennial |
Complete population count, basic demographics |
Every 10 years |
Block |
| PEP |
Population and component estimates between censuses |
Annual |
County |
| Migration Flows |
County-to-county and metro-to-metro migration |
Annual (from ACS) |
County / Metro |
| LEHD LODES |
Jobs by workplace, residence, and commuting patterns |
Annual (2-3 year lag) |
Census Block |
ACS
American Community Survey
What is the American Community Survey?
The American Community Survey (ACS) is an ongoing survey conducted by the U.S. Census Bureau. Unlike the decennial census which counts everyone, the ACS samples about 3.5 million households annually to estimate demographic, social, economic, and housing characteristics.
1-Year vs 5-Year Estimates
| Feature |
1-Year Estimates |
5-Year Estimates |
| Time Period |
Single year |
60 months of data |
| Currency |
Most current |
Less current (represents 5-year average) |
| Reliability |
Larger margins of error |
Smaller margins of error |
| Geographic Coverage |
Areas with 65,000+ population |
All areas including small geographies |
| Best For |
Large areas, current data |
Small areas, detailed analysis |
Recommendation: Use 5-Year estimates (acs5) for most analyses, especially at tract or block group level.
Margins of Error
Every ACS estimate comes with a margin of error (MOE) representing the uncertainty in the estimate. The MOE defines a confidence interval around the estimate.
For example, if median income = $50,000 with MOE = ±$5,000, you can be 90% confident the true value is between $45,000 and $55,000.
Important: Small geographies and small population subgroups often have large margins of error. Use the MOE Reliability Filter tool to identify unreliable estimates.
Geography Hierarchy
| Level |
Description |
Typical Count |
| Nation |
United States total |
1 |
| State |
50 states + DC + territories |
52 |
| County |
Counties and equivalents |
~3,200 |
| Tract |
Statistical areas of 1,200–8,000 people |
~85,000 |
| Block Group |
Subdivisions of tracts, 600–3,000 people |
~240,000 |
| ZCTA |
ZIP Code Tabulation Areas |
~33,000 |
Common ACS Tables by Topic
Income & Poverty
B19013 Median Household Income
The most commonly used income variable. Single value per geography.
Key variable: B19013_001 — Median household income in past 12 months
B19001 Household Income Distribution
Counts of households by income bracket. Use to calculate percentages in income ranges.
Key variables: B19001_001 (Total), B19001_002 (Less than $10K), ... B19001_017 ($200K+)
B17001 Poverty Status
Population by poverty status. Use to calculate poverty rate.
Key variables: B17001_001 (Total), B17001_002 (Below poverty level)
Poverty Rate: B17001_002 / B17001_001 × 100
B19301 Per Capita Income
Average income per person.
Key variable: B19301_001 — Per capita income
Education
B15003 Educational Attainment
Population 25+ by highest level of education completed.
Key variables:
B15003_001 — Total population 25+
B15003_017 — High school diploma
B15003_022 — Bachelor's degree
B15003_023 — Master's degree
B15003_025 — Doctorate degree
B14001 School Enrollment
Population 3+ by school enrollment status.
Key variables: B14001_001 (Total), B14001_002 (Enrolled in school)
Housing
B25003 Housing Tenure
Occupied housing units by owner/renter status.
Key variables:
B25003_001 — Total occupied units
B25003_002 — Owner-occupied
B25003_003 — Renter-occupied
B25077 Median Home Value
Median value of owner-occupied housing units.
Key variable: B25077_001 — Median home value
B25064 Median Gross Rent
Median rent for renter-occupied units.
Key variable: B25064_001 — Median gross rent
Population & Age
B01003 Total Population
Basic population count.
Key variable: B01003_001 — Total population
B01001 Sex by Age
Population by sex and detailed age groups. Use for population pyramids.
Structure: Male ages (B01001_003–025), Female ages (B01001_027–049)
B01002 Median Age
Median age of population.
Key variable: B01002_001 — Median age
Race & Ethnicity
B02001 Race
Population by race (alone).
Key variables:
B02001_001 — Total
B02001_002 — White alone
B02001_003 — Black/African American alone
B02001_005 — Asian alone
B03003 Hispanic or Latino Origin
Population by Hispanic/Latino origin.
Key variables: B03003_001 (Total), B03003_003 (Hispanic or Latino)
Employment
B23025 Employment Status
Population 16+ by employment status.
Key variables:
B23025_001 — Total population 16+
B23025_003 — In labor force (civilian)
B23025_004 — Employed
B23025_005 — Unemployed
Unemployment Rate: B23025_005 / B23025_003 × 100
Transportation
B08301 Means of Transportation to Work
Workers 16+ by commute mode.
Key variables:
B08301_001 — Total workers
B08301_002 — Drove alone
B08301_010 — Public transit
B08301_019 — Walked
B08301_021 — Worked from home
B08303 Travel Time to Work
Workers by commute time in minutes.
Key variables: Ranges from B08303_002 (Less than 5 min) to B08303_013 (90+ min)
Health Insurance
B27010 Health Insurance Coverage
Population by health insurance status and type.
Key variables:
B27010_001 — Total population
B27010_017 — No health insurance coverage
Decennial
Decennial Census
What is the Decennial Census?
The decennial census is a complete count of every person living in the United States, conducted every 10 years. Because it counts everyone rather than sampling, it has no margins of error and provides data at the finest geographic levels, including census blocks.
Available Census Years
| Year |
Dataset |
Table Prefix |
Notes |
| 2020 |
PL 94-171 Redistricting Data |
P1, P2, etc. |
Population, race, Hispanic origin, housing occupancy |
| 2020 |
Demographic and Housing Characteristics (DHC) |
P1–P18, H1–H10 |
Age, sex, race, households, group quarters |
| 2010 |
Summary File 1 (SF1) |
P001, P003, etc. |
Full count data; different variable naming than 2020 |
| 2000 |
Summary File 1 (SF1) |
P001, etc. |
Full count data |
Variable naming changed between censuses. The 2020 Census uses different variable codes than 2010 and 2000. For example, total population is P1_001N in the 2020 PL file but P001001 in the 2010 SF1. Always verify variable codes for the specific census year you're working with.
Common 2020 Decennial Variables
P1 Total Population (PL 94-171)
P1_001N — Total population
P2 Hispanic or Latino by Race (PL 94-171)
P2_001N — Total
P2_002N — Hispanic or Latino
P2_005N — Not Hispanic: White alone
P2_006N — Not Hispanic: Black alone
H1 Housing Occupancy Status (PL 94-171)
H1_001N — Total housing units
H1_002N — Occupied
H1_003N — Vacant
When to use Decennial vs ACS: Use the decennial census when you need complete population counts at block-level geography or when you want exact counts without margins of error. Use the ACS for detailed socioeconomic characteristics (income, education, employment, etc.) that the decennial census does not collect.
PEP
Population Estimates Program
What are Population Estimates?
The Population Estimates Program (PEP) produces annual estimates of the population between decennial censuses. These estimates use the most recent census count as a starting point and update it using administrative records for births, deaths, and migration.
What PEP Provides
| Component |
Description |
| Total Population |
Annual population estimate for states and counties |
| Components of Change |
Births, deaths, net domestic migration, net international migration |
| Age & Sex |
Population by single year of age and sex |
| Race & Hispanic Origin |
Population by race/ethnicity groups |
Geographic Availability
PEP estimates are available at the nation, state, and county level. They are not available for tracts, block groups, or other sub-county geographies.
When to use PEP: Use Population Estimates when you need the most current population figures between decennial censuses, or when you want to analyze population change components (births, deaths, migration) over time. PEP estimates are considered more accurate than ACS population estimates for total counts at the state and county level.
Vintage years: PEP data is organized by "vintage" — the year the estimates were produced. A vintage 2023 estimate set, for example, includes estimates from the 2020 Census base through July 1, 2023.
Migration
Migration Flows
What are Migration Flows?
Migration flow data, derived from the American Community Survey, tracks the movement of people between geographies over a one-year period. It shows where people moved from and where they moved to, along with characteristics of movers.
Geographic Levels
| Flow Type |
Description |
Example |
| County-to-County |
Migration between individual counties |
Travis County, TX → Williamson County, TX |
| Metro-to-Metro |
Migration between metropolitan statistical areas |
Austin-Round Rock MSA → Dallas-Fort Worth MSA |
What the Data Contains
For each origin-destination pair, migration flow data includes the estimated number of movers, margin of error, and the previous residence one year ago. The data can reveal patterns of in-migration (who's moving in), out-migration (who's leaving), and net migration (the balance between the two).
Margins of error can be large. Migration flow estimates for individual county pairs are based on small ACS sample sizes and often have very large margins of error. Focus on the largest flows or aggregate to broader patterns for more reliable analysis.
When to use Migration Flows: Use migration data to understand population movement patterns — which areas are gaining or losing residents, where new residents are coming from, and where departing residents are going. This is valuable for regional planning, economic development, and housing demand analysis.
LEHD
LEHD LODES (Workforce & Commuting Data)
What is LEHD LODES?
The Longitudinal Employer-Household Dynamics (LEHD) program produces the LODES (LEHD Origin-Destination Employment Statistics) dataset, which provides detailed job counts at the census block level. LODES combines state unemployment insurance records with Census data to create a comprehensive picture of where people work and where workers live.
LODES Data Types
| Type |
Name |
Description |
wac |
Workplace Area Characteristics |
Job counts by workplace location. How many jobs are located in each block? |
rac |
Residence Area Characteristics |
Job counts by worker residence. How many workers live in each block? |
od |
Origin-Destination |
Commuting flows between home and work blocks. Where do workers commute from and to? |
Job Type Segments
LODES data can be filtered by job type:
| Segment |
Description |
S000 |
All jobs |
SA01 |
Jobs with earnings $1,250/month or less |
SA02 |
Jobs with earnings $1,251–$3,333/month |
SA03 |
Jobs with earnings greater than $3,333/month |
SE01 |
Jobs in Goods Producing industries |
SE02 |
Jobs in Trade, Transportation, and Utilities |
SE03 |
Jobs in All Other Services |
Key Variables (WAC and RAC)
Total and Demographic Variables
C000 — Total number of jobs
CA01 — Workers age 29 or younger
CA02 — Workers age 30 to 54
CA03 — Workers age 55 or older
CE01 — Earnings $1,250/month or less
CE02 — Earnings $1,251–$3,333/month
CE03 — Earnings more than $3,333/month
Industry Variables (NAICS Sectors)
CNS01 — Agriculture, Forestry, Fishing, Hunting (NAICS 11)
CNS02 — Mining, Quarrying, Oil/Gas (NAICS 21)
CNS05 — Construction (NAICS 23)
CNS07 — Retail Trade (NAICS 44-45)
CNS09 — Information (NAICS 51)
CNS12 — Finance and Insurance (NAICS 52)
CNS15 — Professional, Scientific, Technical Services (NAICS 54)
CNS16 — Management of Companies (NAICS 55)
CNS17 — Administrative and Waste Services (NAICS 56)
CNS18 — Educational Services (NAICS 61)
CNS19 — Health Care and Social Assistance (NAICS 62)
CNS20 — Arts, Entertainment, Recreation (NAICS 71)
No API key required. LEHD/LODES data is downloaded directly from the Census Bureau's servers by the lehdr R package. You do not need a Census API key for these tools.
When to use LEHD: LODES is the go-to dataset for workforce analysis, economic development, and commuting studies. Use WAC data to see where jobs are concentrated, RAC data to see where workers live, and OD data to map commuting flows. Because it's available at the block level, you can aggregate it to any geography you need — tracts, ZIP codes, custom study areas, or municipal boundaries.
Data lag: LODES data typically has a 2–3 year lag. As of 2024, the most recent LODES data available is for 2021. Check
lehd.ces.census.gov for the latest available year.
Tips for Working with Census Data
Always check margins of error (ACS)
Use the MOE Reliability Filter tool to identify estimates with CV > 40%. These should be used with caution or suppressed from analysis.
Use the right denominator
When calculating percentages, make sure your denominator matches the universe for your numerator. For example, unemployment rate uses civilian labor force (B23025_003), not total population.
Don't compare overlapping 5-year estimates
The 2019–2023 ACS 5-year and 2020–2024 ACS 5-year share 4 years of data. For valid comparisons, use non-overlapping periods (e.g., 2014–2018 vs 2019–2023).
Understand inflation-adjusted dollars
Income and value variables are adjusted to the final year of the estimate period. The 2019–2023 ACS reports values in 2023 dollars.
Choose the right data source
Need block-level total population? Use the Decennial Census. Need current county population estimates? Use PEP. Need detailed socioeconomic data at the tract level? Use the ACS. Need job counts and commuting patterns? Use LEHD LODES. Each source has its strengths — pick the one that fits your analysis.