American Community Survey data suppression

Our users frequently wonder why certain American Community Survey (ACS) estimates are not available. This is, more often than not, caused by data suppression. 

Data suppression is a restriction applied to certain estimates to prevent disclosing sensitive information about individuals and minimize the number of estimates with unacceptably low reliability. There are multiple data suppression factors ranging from population thresholds to topic restrictions and we’ll cover all of them below.

Population Thresholds

Population thresholds are applied to ACS 1-year and Supplemental estimates only. The estimates will be published if an area has a population at the threshold or exceeds it. Areas with a population of 65,000 or more are eligible for 1-year, 5-year, and Supplemental estimates, whereas areas with a population of 20,000 or more are eligible for 5-year and Supplemental estimates. Areas with a population of 20,000 or fewer are eligible only for 5-year estimates.

As with every rule, there are certain exceptions: 

• 1-year estimates are published for all Metropolitan Statistical Areas (MSAs) with populations of at least 50,000. 
• 1-year estimates are published for all areas in which total population is within 5% of the 65,000, and the estimates were published for previous year. This is particularly useful for areas in which population decline temporarily due to unusual events, such as natural disasters.

Data quality

The ACS tables consist of a series of estimates. It’s only natural that those estimates are subject to sampling variability, which can be summarized by its standard error. Data quality filtering rules are devised to moderate table filtering based on variability and are applied to 1-year and Supplemental estimates. 

If more than half of the estimates in the table are statistically equal to zero at a 90% confidence level, the table will not be published.

Base and collapsed tables

As a part of the ACS, two types of tables are released: base and collapsed tables. Base tables are the most detailed estimate. For collapsed tables, categories and estimates are collapsed and combined. Since base tables contain more details split into more tables, they often do not meet the requirements for publishing, in which case collapsed tables come in handy.

Disclosure Review Board rules

The Disclosure Review Board sets additional rules in order to ensure confidentiality and protect respondent privacy, and all ACS tables are submitted to the Disclosure Review Board for approval each year. These rules pertain to a minimum number of cases required to publish a cell in a table, table cell restrictions, and table topic restrictions.

Table suppression for specific geographies

All of the ACS data goes through a rigorous review, during which analysts may find microdata errors that occurred either during collection or processing. If the management determines that correcting errors is not possible without affecting the release schedule, the affected tables are suppressed for specific geographic areas and not published.

Head over to the U.S. Census Bureau website to find out more about data suppression.