The analyzing of this week’s letter from the Census Bureau about the cost-effective way it increases its response rate to the Census forms that arrive next week continues.
What we’ve learned so far is the estimated $45 million spent on the letters you received this week is more than offset by the estimated $500 million the Census Bureau says it would otherwise have spent getting them. I calculated yesterday that it would cost about $75 for Census workers to snare the forms.
What was still left unresolved, however, is how the Census Bureau calculated that $75 figure. I suggested today that if a letter motivated 7 million households to return the form, it might well cost less than what it costs to retrieve the forms from people who don’t respond to three phone calls and three personal visits.
So I’ve been asking the Census Bureau to break it down for me. A few minutes ago, I got its response:
Thank you for your email. We here at the Census Bureau certainly understand your concern, and I want to assure you we care a great deal about being good stewards of the taxpayer’s money.
The short answer to your question is that even on the eve of census forms arriving next week, as many as 45% of Americans are unaware that this month is when the Census starts. That’s probably due to the fact we only do a Census once every ten years. Based on historical response rates, we expect roughly two thirds of households will mail back their form. The rest we will have to send an enumerator to collect the data required by the Constitution. You can imagine that follow-up is an expensive proposition.
In fact, every one percent increase in the number of households who mail back the form saves the taxpayers about $85 million in expensive door-to-door follow up. That’s why we advertise and promote, to increase the mail back response rate and help save on expensive labor to follow up.
We have extensive research that shows additional mailings alerting households to the arrival of the census form increase response rates by about 6 to 12 percentage points. The savings from that increase more than pay for these mailings. It costs about $85 million to print and mail the advance letter and reminder postcard. The potential increase in response rates demonstrated by our research could result in a savings of more than $500 million.
After the 2000 Census we returned to the Treasury some $305 million in savings. Then Secretary of Commerce Don Evans testified in 2001 to the U.S. Senate that those savings came about from our advertising, promotion and PR efforts encouraging households to mail back their forms, increasing response rates over the prior census for the first time in three decades.
The total we spend on all promotion and advertising is about one dollar perperson in the U.S. It costs just 42 cents to mail back the form. But it costs $57 for follow up with non-responding households, many of which we must visit several times to reach someone at home. We wish we did not have to advertise, and that 100% of households mailed their form back, but that is just not the case. Not everyone is as active and engaged as yourself.
We appreciate your civic minded spirit, and your concern for our current fiscal situation. I hope this information helps explain the steps we are taking to reduce the burden on taxpayers of meeting our Constitutional mandate to count every person in the country.
And that answer led me to another question that I’ve sent:
Thank you for your response. Unfortunately it didn’t answer the question I asked. What I’m trying to determine is how the Census Bureau calculates that it would take $75 worth of work to get the forms from the — in this case — 6 percent of people who wouldn’t have sent it back, were it not for the letter? (You quote a $57 figure below which contradicts — a little bit — some of the information I received via some Census press releases earlier this month. But in any event, I’d be interested in how that $57 figure is reached)
In the comments from the previous post on the subject, I indicate my reason for asking the question:
As I said above, the EASIEST people to get to return the survey, are people like me who were excited to think that’s what was in the envelope.
The NEXT most-easy is the 7 million households that will send it in because they were sent a letter.
My theory is that they came up with the $500 million figure by calculating how much time/overhead it takes to get the LEAST-likely respondents to respond. That could take three phone calls and three personal visits and that COULD cost $75 per household.
But that figure may not — and logic says it DOES not — apply to the targets of this letter.
In other words: The more work you have to do to get people to send back their forms, the more expensive it is. As people have pointed out, sending out a letter a week ahead of time to get a response rate increase is an easy — and acceptable — method. But in determining whether it’s the most cost-effective method can only be concluded by determining how much it would cost to get the forms from those same people, not how much it costs to get them from people who don’t respond to the “a letter is coming” letter.
Is it dwelling on the trivial? Maybe. But it’s also an exercise in questioning the answers we get when ask questions to be sure they’re correct and add up. That’s never a bad thing.