About the Census & the Annual American Community Survey 


  • straight count of people who live in the U.S., the census provides basic demographics – sex, age, race, Hispanic origin, and homeowner status (aka ‘tenure’)
  • Mandated by the Constitution.
  • Takes place every 10 years. The next census will be published in 2020.
  • Determines the number of seats each state has in the U.S. House of Representatives.
  • Used to distribute billions in federal funds to state and local communities.
  • Now asks 10 “short form” questions (starting with Census 2010). Here’s a copy of the short form questionnaire:  Census 2010 Questionnaire

American Community Survey Estimates

Prior to the 2010 census, a ‘long form’ questionnaire was used to gather statistical data about the U.S. population.  This detailed information painted a picture of the demographic, social, and economic face of the nation – from states to counties to census tracts.   See sample here:  Census 2000 Long Form Questionnaire

Because data from the decennial census arrived in 10-year intervals, there was no way to harness and analyze population trends and other changes in the intervening years.

Congress asked the Census Bureau to bridge this gap with information and data about how the lives and circumstances of Americans change in those in-between years.

The American Community Survey estimates (ACS) is the result.

  • American Community Survey shows how people live.
  • Annual data for the ACS is acquired from a series of detailed questions asked of each respondent – not unlike the old ‘long form’ questionnaire of decennial census 2000 and earlier; here’s a sample ACS Questionnaire.
  • It is the source for detailed demographic and socio-economic data – citizenship, educational attainment, income, home ownership, health insurance coverage, travel time to work…, and data is available, down to small areas (census tracts- about 4,000).

Subjects included in the ACS survey


Important Note:

All ACS data are estimates, and should ALWAYS be identified/referred to as such.  To help interpret the reliability of the estimate, a margin of error (MOE) is included for every ACS estimate. The margin of error provides a measure of the range of uncertainty around each estimate; this range can be calculated with 90 percent confidence by taking the estimate +/- the MOE. If, for example, the ACS reports an estimate of 100 +/- 20, then there is a 90 percent chance that the value for the total population falls between 80 and 120.


The larger the MOE, the lower the accuracy of the estimate—and the less confidence one should have that the estimate is close to the true value.  If the margin of error is over 10%, use with caution.

A census tract is too small an area to get reliable ACS estimates; use the NTA (neighborhood tabulation area)

Each year, we get three types of datasets:

  • ACS 1-year estimates for populations of 65,000+ (2015)
  • ACS 3-year estimates for populations of 20,000+ (2011-2013)
  • ACS 5-year estimates for populations of any size (2011-2015)

The comparison table below (courtesy of the ACS site), is helpful in determining which estimates one should consider using when looking for data.

1-year estimates 3-year estimates 5-year estimates
12 months of collected data 36 months of collected data 60 months of collected data
Data for areas with populations of 65,000+ Data for areas with populations of 20,000+ Data for all areas
Smallest sample size Larger sample size than 1-year Largest sample size
Less reliable than 3-year or 5-year More reliable than 1-year; less reliable than 5-year Most reliable
Most current data Less current than 1-year estimates; more current than 5-year Least current
Best used when Best used when Best used when
Currency is more important than precision

Analyzing large populations

More precise than 1-year, more current than 5-year

Analyzing smaller populations

Examining smaller geographies because 1-year estimates are not available

Precision is more important than currency

Analyzing very small populations

Examining tracts and other smaller geographies because 1-year estimates are not available


When Making Comparisons with ACS Datasets:

  • DO compare similar period lengths, e.g. 1-year to 1-year.
  • DON’T compare estimates from different period lengths, e.g. 1-year to 3-year.
  • DO compare estimates from non-overlapping periods, e.g. compare a 2005-2007 ACS 3-year estimate to a 2008-2010 ACS 3-year estimate.
  • DON’T compare overlapping periods, for example, the 2005-2007 ACS 3-year estimates to the 2006-2008 ACS 3-year estimates. 

When comparing ACS estimates with decennial Census data, this Census Bureau link will prove invaluable:  ACS/Census Table Comparisons


Useful Tools for Finding Census and ACS Data for NYC CDs and Neighborhoods

NYC’s indispensable Department of City Planning produces the very familiar Community Data Portal.  Use the Population Data icon found in the menu bar for each CD Profile to access demographic data from the Census Bureau and the American Community Survey estimates.  Pay close attention to the descriptions (included below) of the various geographies at your disposal; your understanding of each will be most helpful in determining which geography best suits your editorial needs:

  • Census Tracts – small statistical subdivisions of counties used by the U.S. Census Bureau. In New York City, there are 2,168 census tracts, which typically have a population of about 2,500 – 4,000 each, and are typically a range of a city blocks.
  • Public Use Microdata Area(s) approximate New York City’s 59 community districts – about 100,000 persons.  A Census Bureau creation; there are 55 PUMAs in NYC. Here is a map of New York City PUMAs by CD (Public Use Microdata Areas as defined by the Census) and Community Districts. PUMAs are an approximation NYC Community Districts (CDs) borders.
  • Neighborhood Tabulation Area(s)are subsets of New York City’s PUMAs.  Created by the NYC Department of City Planning, these areas are aggregations of census tracts based on neighborhoods, and are a valuable tool for presenting published census data; the Census Bureau does not publish data for NTAs.  Find Census Tract-to-NTA-to-PUMA equivalencies here: Census Tract/ NTA/PUMA Equivalencies

Take Note:  This data is not the most current available.  And the data provided is selected, not complete

SocialExplorer (available via CUNY J-School database) has an easy-to- use interface, and contains datasets for all Decennial Censuses and American Community Survey estimates.  Have a look at Barbara Gray’s excellent how-to for finding community data with SocialExplorer here: Tipsheet.

Let’s have a go at the assignment you had for finding the foreign born from Iraq, Iran, Sudan, Somalia, Yemen and Syria.

The U.S. Census’ New American Factfinder is the repository for all things census; include the American Community Survey, the Decennial Census, the Annual Economic Surveys, the American Housing Survey, the Economic Census (published every five years).

The Census Reporter is a cool tool which provides a more pliable interface for reporters and easier access to American Community Survey census data to write stories. (Funded by the Knight News Challenge)

Sample Search:  

Use ACS data to generate story ideas


To provide context, always make comparisons – neighborhood to borough, neighborhood to city, city to nation.

Compare data year over year in an area to identify trends.

Use latest data to examine how they affect your beat or community.

Take a national story, and localize it.

If you hear about a trend or issue on your beat, use Census data to illustrate or disprove. 

There’s No Religion Data in the US Census!

  • The Association of Religion Archives and Data (ARDA)  ARDA Archives and Data has produced a 2010 study of US congregations and membership; data tracks to the state and county level.
  • The same data Religion 2010 (InfoGroup)is also available (PUMA level) on the J-School’s Social Explorer database.