Most Read: Politics

Behind the Numbers
Posted at 01:26 PM ET, 03/09/2010

Census "heads up" prompts questions

As "Census Day" approaches, the U.S. Census Bureau has come under fire for some of its edgier attempts to encourage people to respond to its mailings, and now, some have begun to take umbrage with standard technique - the advance letter.

On National Review's The Corner, John J. Miller questions the mailing itself while Bill S. on RedState writes, "I'm having a difficult time deciding if this letter is: 1. Supposed to be helpful or informative in some way. 2. A joke. 3. Some sort of Obama stimulus plan for the postal workers. 4. My imagination."

But to those in the survey research world, including the Census Bureau's new director, Robert M. Groves, the letters are standard operating procedure. The bulk of the research on the topic finds that advance letters explaining the purpose and benefits of survey research improve response rates.

For an effort like the Census, which attempts to contact more than a hundred million households, the educated bet is that the cost of postage and printing for the preliminary letters saves subsequent future outlays in sending Census workers to follow-up with those who don't return the questionnaires.

The Bureau's announcement of the letters puts a dollar amount on response rates, "Census Bureau research shows that reaching out to respondents with an advance letter and reminder postcard if necessary can boost census mail-back rates and save money. For every 1 percent increase in households that respond by mail, taxpayers save about $85 million in operational costs associated with census takers going door to door to follow up with households that did not mail back the form."

The Bureau says that research on such advance letters used during the 2000 head count finds they increase participation by about 6 percentage points. Groves writes in defense of the letters, "The research is clear that the advance letter can save money for all of us."

Groves himself literally wrote the book on this. In the oft-used text "Survey Nonresponse," edited by Groves and several other prominent survey researchers, he writes of advance letters: "[i]t has become commonplace in face-to-face surveys to mail a letter to the sample household that alerts the unit to the upcoming call. From a theoretical perspective, this design feature can be used to manipulate a variety of influences known to affect survey participation."

By Jennifer Agiesta  |  01:26 PM ET, 03/09/2010

Categories:  Polls, Polls, Polls

Read what others are saying

    © 2011 The Washington Post Company