The Census in Context
The Census Bureau, part of the Commerce Department, is under tremendous political pressure from the White House, Congress, and the general public. This pressure can come in the form of oversight by Congress, rules imposed by the White House, or lobbying from both citizens and organizations alike. There is a specific purpose behind every question on the census form, and many are the result of advocacy from stakeholders. Birthdate/age is asked to determine the potential voting age population. Questions are also added to shine light on new demographic trends or because of a law. For example, a question about grandparents raising their grandchildren was added in 2000 as this trend was identified as one important to track; computer and internet use was added in 2013 as a requirement of the Broadband Data Improvement Act.
Pressure from various groups, in addition to financial concerns, led to a major change in the Census in 2005 with the introduction of the American Community Survey (ACS). Prior to the ACS, researchers, government officials, and others had to wait 10 years for new decennial census data; now they can get data for a single year, or 5-year rolling averages. The ACS surveys 250,000 households a month, continually. While this yields a much smaller sample than the decennial Census long-form (which was given to 1/6th of the population), ACS demographic data is available much more quickly, and in most cases is just as accurate (with the limitation that since it is a sample, it may not be as accurate or precise for small populations/geographies).
Article 1, Section 2 of the U.S. Constitution is about the make-up of the House of Representatives and apportionment, but clause 3 directs an enumeration of the population, aka a census, to be taken every ten years.
The Census vs the American Community Survey
There are a number of differences between the information which was available from the decennial census long form and the ACS data we have now. The main advantage of the ACS is that it is administered continually, meaning that the data can be released yearly rather than every 10 years
- The ACS data comes from a sample of the population. Since the sample is chosen systematically, it can be used to represent data for a population fairly accurately.
- Why doesn't the Census ask everyone all the questions of interest? There are many reasons (cost and speed of data collection being the two most important) that it's more efficient to have only a sample of the population fill out the longer questionnaires (until 2005 in the census long-form, and since then in the ACS).