Census Survey Bill Would Leave Policymakers in the Dark

House Republicans want to eliminate American Community Survey or drop its mandatory requirement.
by Mike Maciag | May 28, 2012 AT 12:00 PM

If there’s a single set of statistics most indispensable to public officials as they make decisions shaping policies, it’s probably the Census Bureau’s American Community Survey.

The annual survey is the gold standard for current demographic data and a key factor in funding allocations for state and local governments – providing a snapshot on everything from vacant housing units to immigrant populations. Without it, officials would rely on less precise estimates to guide policymaking.

But that’s exactly what would happen if a Republican-led effort to eliminate the American Community Survey (ACS) is successful.

Earlier this month, the House passed a measure along mostly party lines to completely cut off its funding. Some GOP lawmakers favor keeping the survey, but want to drop the mandatory requirement to complete it.

Much of the debate centers on privacy concerns. Conservative lawmakers argue the survey’s compulsory requirement is unconstitutional and intrusive, with those willfully neglecting to answer questions technically subject to fines up to $5,000.

The Census Bureau is not an enforcement agency, though. Director Robert Groves testified at a House Oversight and Government Reform Committee hearing that he was unaware of a single prosecution for non-compliance, emphasizing that interviewers’ ability to educate respondents on the survey’s importance is far more effective than threatening to levy fines.

Non-compliance doesn’t appear to be a problem. The 2010 ACS reported a response rate of 97.5 percent, far exceeding typical rates for surveys. Only about 1 percent of households refused to complete surveys in recent years, according to Census figures.

Still, Rep. Darrell Issa (R – Calif.) and other lawmakers would prefer that the ACS be voluntary.

Random surveys with large sample sizes and high response rates are considered reliable. But converting the ACS to a voluntary survey could result in bias among response rates across different demographic groups, diminishing its accuracy.

Comments made by Daniel Webster, a Florida Republican who sponsored the legislation, suggest the freshman legislator isn’t familiar with this concept. At least he wasn’t when the New York Times interviewed him last week:

“We’re spending $70 per person to fill this out. That’s just not cost effective,” he continued, “especially since in the end this is not a scientific survey. It’s a random survey.”

Ironically, Republican lawmakers used an online constituent survey to justify the proposed legislation. Voters on House Majority Leader Eric Cantor’s YouCut website chose to eliminate the ACS over a number of other programs also deemed wasteful by Cantor.

Visitors to Cantor’s website hardly represent a cross-section of Americans. But this didn’t stop Webster from declaring the apparent widespread support for the initiative. “The American taxpayers agree,” Webster said on the House floor. “Eliminating the ACS was overwhelming the winner when the citizens were polled what federal spending they would cut.”

The bill appears to have little support in the Senate. The White House has also pledged to veto the law should it make it that far.

In 2010, a decision by Canadian lawmakers to lift the compulsory requirement for the country’s National Household Survey was met with sharp criticism in policy and academic circles. A third of Canadian households received the now-voluntary 2011 survey, and the government has assumed a response rate of 50 percent.

Two University of British Columbia economics professors wrote a policy paper denouncing the move, showing voluntary surveys tended to misrepresent income figures at the top and bottom of the scale. They also noted other surveys had used mandatory census figures as benchmarks to weight results and better reflect populations, a method not possible without a true census.

The extent to which the change might sway survey responses, though, likely will be unknown until Canada releases the results next year.

The Census Bureau estimates the ACS is sent to 2.5 percent of homes each year, requiring an average of 38 minutes per household to review instructions and answer questions. At this rate, the typical American would respond to the survey about twice in their lifetime.

Webster also called the ACS “intrusive,” citing questions on emotional condition and commute times. "It’s the definition of a breach of personal privacy,” he said.

At first glance, some questions do seem unwarranted. But many fulfill multiple statutory requirements in other legislation.

One question regarding a person’s physical or mental condition helps assess policies aimed at assisting the disabled, such as transit agencies providing adequate facilities under the Americans with Disabilities Act. Another survey item requests the number of times a person has married for measuring changes in households that guide child care, social services and related programs.

Census workers are sworn to protect confidentiality, facing prison sentences up to five years for disclosing any personal information as outlined in Title 13 of the U.S. Code. No employees are known to have violated the provisions, according to the agency.

Much ACS data is employed in administration of programs benefitting marginalized groups. The Voting Rights Act, for instance, specifically mandates the Census Bureau use the ACS to identify areas with high concentrations of non-English speakers. Historically, Americans with most to gain from the survey typically respond at lower rates.

Policymakers would often end up utilizing less reliable or outdated figures without the survey, essentially governing blindly.

And public agencies aren’t the sole beneficiaries of the data. Relators and other businesses manage operations with the figures. Research organizations regularly factor ACS data into their studies. Journalists also use the data to detect trends and find stories.

Even Rep. Webster links to Census information on his own website.

House Republicans cited the price tag – an estimated $2.4 billion over the next 10 years – as another reason to cut the ACS. But its importance in decision-making far exceeds the cost, with ACS data helping to direct $416 billion in federal assistance in fiscal year 2008, according to a Brookings Institution study.

The survey’s worth doesn’t hinge on cost savings, though; it's true value centers on being able to make policy and budgeting decisions based on a reliable, factual analysis.