How should I interpret the Margins of Error?

Each Census ACS estimate comes with a margin of error (see What do Margins of Error indicate?) which is the range over which there is a 90% probability that the actual value lies. So if the estimate is 90% ± 2%, you have a 90% probability that the actual value is between 88% and 92%. If the margin is ± .02%, then there is a 90% probability that the actual estimate is between 89.98 and 90.02 — which is very tight.

When there are only a few households contributing data, the margins might be large, even up to 100%.

Large margins of error indicate that the estimate itself is pretty soft.

In addition, for smaller communities or areas, there may be no valid estimate. These values will show up as “Not available (NA)” in tables and on maps.