Incompetence by the numbers
A worker finds out the hard way that helping with the US Census isn’t what it used to be
By David Czamanske 05/21/2009
In April 1999 I worked for the US Census Bureau as an address verifier, ensuring that residential addresses were correct for census forms to be mailed on April 1, 2000, for the decennial census. The work was fulltime for approximately nine weeks, until early June.
I enjoyed it. An active retired person who likes walking and has an educational background that includes urban geography, I liked venturing into new neighborhoods. Plus, I felt I was making a small contribution to the census required by the Constitution for determining the number of members each state is entitled to in the House of Representatives.
In early 2009, I took and passed the exam that qualified me to work again in this capacity, and in mid-March I was contacted to begin training as an enumerator. Along with 12 others, I attended a basic one-week training class in early April. Most of the first morning was taken up with filling out required forms — there must have been a dozen — each requiring us to provide our name, date and signature. I wondered why the forms had not been consolidated and why we were generating so much paper in this era of computerized data collection.
In the week-long training, we learned how to use the Census Bureau’s new Hand-Held Computer (called our HHC), one of them issued to each of us to help verify the addresses of residents in single-family and multifamily dwellings.
The HHC, which included a Global Positioning System enabling us to “Map Spot” each residence, was accompanied by a needlessly elaborate handbook that contained multiple errors (the list of errors, in small type, accompanying the handbook went on for four pages). I couldn’t help but ask myself: Why was the basic handbook for 140,000 enumerators hired nationally not proofread carefully before 140,000 copies were printed?
The paid training lasted longer than I felt necessary, but did include a day of field work. During the training our crew leader told us that the Census Bureau had set June 5 as the target date for completing all address verification, with June 12 as the absolute deadline. We all looked forward to nine, perhaps 10 weeks of employment.
But toward the end of the second week, our crew leader told us that the regional census office was running out of assignment areas and that our work was just about over. By April 20, all the work was done in our district.
What had gone so grossly wrong in the estimate of time required for the work? Could it have been an error in basic arithmetic? According to national news reports, 140,000 enumerators were hired to verify the addresses of the nation’s 145 million households.
That’s an average of just over 1,000 addresses per enumerator. We were told that enumerators had processed an average of 19.2 addresses per hour in test runs, or almost 800 addresses in a 40-hour week.
In fact, I was often able to accurately verify 30 or more addresses in an hour. In two weeks of full-time work, I estimate that I processed more than 2,000 addresses.
Whatever the reason for the incorrect time estimate, my co-workers and I were extremely disappointed that the Census Bureau had more or less promised nine weeks of employment and then left most of us hanging out to dry after two weeks.
Moreover, we were paid for one week of training and then worked a little more than two weeks, so 30 percent of our pay was invested in training. Additional workers were trained in the week after our training, and they worked for just one week — so almost 50 percent of their pay was invested in training!
Despite my disappointment, I hope to work next year as an interviewer — as I did in 2000 — for the 2010 Census. After all, we do need to know how many members of Congress the state of California is entitled. I just have to hope that’s one calculation they get right. n
David Czamanske can be reached at firstname.lastname@example.org.