My Two Census

Formerly the non-partisan watchdog of the 2010 US Census, and currently an opinion blog that covers all things political, media, foreign policy, globalization, and culture…but sometimes returning to its census/demographics roots.

Posts Tagged ‘Crew Leaders’

Daily Sound Off: The real problems with payroll

Friday, May 28th, 2010

Here’s today’s Daily Sound Off:

I work for the payroll department in my LCO.  I wanted to explain some things about how Census payroll works and why people are getting paid late.  I would appreciate if my name were left out of this, but feel free to publish some or all of the information contained below.

As you may know, in order to get paid for a day’s work a Census employee must submit a daily payroll form that we lovingly refer to as a “308.”  The 308 contains several redundancies to help catch potential errors.  For instance, the employee must mark both the date worked and the day of the week worked, and if these do not match the 308 will not be processed until the office can determine what date the employee actually worked.  The employee also must enter the number of hours worked and the times worked, and if these do not match the employee will be paid for the lesser of the two numbers.  Finally any expenses incurred must be explained and any over $5 must be accompanied by a receipt; in order to save taxpayer dollars we regularly reject claims for ridiculous things that the employee does not need to complete their assignment.

The reasons that we’re having so much delayed payroll come down to the problems with processing these time sheets.  First of all, as I mentioned before, if there are any errors with a paysheet, that sheet may be placed into a problem file to be dealt with later.  Ideally we deal with all problem 308s in their appropriate pay period, but the first three weeks of NRFU were not ideal.  You’ve heard of all the paperwork new employees have to fill out?  All of that has to be processed by the admin department *before* an employee can be paid.  Admin departments basically had to begin processing one to two thousand hiring packets plus five to ten thousand pay sheets starting at the end of the first day of training and be finished by the following Monday.  For many LCOs, that just didn’t happen.  That’s why we all put in overtime that week – to try to get as many people paid as possible.

Now, from the perspective of someone whose job it is to process paysheets, the thing about problem 308s is that some are very easy to deal with and some are very difficult, but almost none of them would exist if the employees themselves took the time to fill these things out right.  Everyone who works for the census was tested on the ability to read and count and everyone who works for the census was hired basically to enter information on forms, and filling out pay sheets does not require any skills beyond these.  And yet we continuously have problems with people who apparently cannot count to 40 – who either claim overtime with under 40 hours a week worked, or claim no overtime with more than 40 hours a week worked.  We continue having problems with people who apparently cannot glance at a calendar long enough to verify both the date and the day of the week.  So while we try to get these errors fixed, a large portion of the employees who are getting paid late are being delayed because they made mistakes on their paperwork that we cannot easily deal with.

Of course the other problem we’re facing is that we can’t process payroll that we don’t have.  I’ve heard numerous stories of FOSes and CLs who don’t submit 308s on time.  I understand from the Crew Leaders’ position that they have a lot to do, but most of our CLs get their 308s in on time.  The maybe 5% who don’t account for 90% of the phone calls we get from enumerators who have missed several days’ pay from their checks.

This is a personnel problem.  We simply don’t have a good way to motivate large numbers of temporary employees to do their jobs promptly and correctly.  Every job has its share of lazy or incompetent employees.  The Census does work to terminate these, but if we have to give each CL who brings payroll in late (or never) at least two warnings, that’s at least three weeks of delayed payroll before we can replace the person, which is why we’re getting stories from across the country of whole crews who haven’t been paid for two or three weeks of working.  Rumor around the office has it that the terminations for unsatisfactory performance are going to start coming fast and furious starting next week, although we’ve already got a decent pile going now.

Now, the admin department gets well over a hundred calls a week inquiring about missing hours or days.  In the vast, overwhelming majority of cases – including every single call I have personally handled – these hours or days are already processed and on their way to the employee on the next pay period.  I understand that it is difficult for many people, especially those whose only job is the Census, to have to wait three weeks instead of two to be paid for a particular day’s work.  Some people may be counting on being paid on time.  I think that the situation would have been helped immensely if we had issued a blanket disclaimer at training or even during the hiring process that it is normal for it to take up to four weeks to be paid for any particular day worked.  Somehow, people formed an expectation that a gigantic government bureaucracy staffed entirely by people with virtually no experience would be fast and efficient at handling paperwork, which makes me wonder if none of these employees who are calling us up or going to the media because their pay is a week late have ever tried to mail a letter or get a driver’s license.  Anyhow, we try to stay cheerful but a certain fatalism develops when all we can do is tell people, essentially, that their check is in the mail.

I can say that fortunately our department is now caught up with payroll on a weekly basis, and it is only when CLs or FOSes bring 308s in late that we process them late.  However, payroll is already on a delayed basis by design – so if I work on a Monday, that 308 gets processed by the LCO and “closed” the following Monday, which means that a direct deposit will be issued the week after that, usually on a Wednesday – a delay of up to 17 days.  So people who missed hours on their last paycheck were actually missing hours for the week of May 9-15 – which was basically the second week of actual work, and third week of employment, and at that point we had many but not all of our glitches ironed out.  By that point we had issued directives to FOSes and CLs about how and when to fill out and bring in 308s and started getting positive responses, which should be reflected in even fewer errors in next week’s checks.

However, the heart of this issue is actually in how the Census approaches the hiring process.  While the recruiting process stretches over two years, the hiring process is basically crammed into a week.  Queens LCOs had to hire 1600 – 2200 employees over the week of April 19th, for a training session that started April 26th.  This has obvious problems.  First of all, we were asking people – many of whom had taken the test months ago, in the fall or even summer – to drop everything and come in for training with a week’s (or in some cases, a day’s) notice.  This is pointless and disrespectful and also resulted in the loss of many promising candidates.  Basically, we weeded out everyone who had a job, or responsibilities, or the ability to plan, or the self-respect to demand to be treated courteously by an employer; then we hired whoever was left.  Certainly we found some people who were competent and hard-working and just down on their luck or hit by the economy, but the overall caliber of employees is lower than what it would have been if we had given people adequate notice or contacted them in a timely fashion after they took their test.

The second problem is, as I have said, the logistical difficulty of processing 2000 new hires at once.  If we had hired people on some kind of rolling basis we could have gotten their paperwork filed and their payroll started up before they had to start working.  If we had started hiring and taking care of administrative matters in, say, March or even April 1st, as most test-takers were promised, then we could have gotten people trained, processed, and into payroll before NRFU even began.  This would have eased the burden on admin, but also on NRFU and the people who had to get training sites for thousands of people all during one week.  This would also have reduced the number of people who were verbally hired but never contacted again, or who attended training but were never assigned a CL, or who were assigned a CL but never any work.

Also, there simply has to be a less resource-intensive way to handle payroll than having each employee hand a piece of paper to their CL each day, to be handed to the FOS each day, to be brought into the office each day, to then be audited by one clerk and then entered into the payroll system by another clerk and then sent to a different agency entirely for final processing.  We did payroll exactly the same way in the 2000 Census, and guess what?  We’ve had ten years and the internet since then.  We have secure banking, we have ebay, amazon, paypal (all of which, I realize, we also had in 2000).  Why can’t we have a server that the employee can log on to to enter their information; that the CL can log on to to approve the hours worked and digitally sign; that can automate the auditing process and eliminate the need for a separate data entry process?  I believe I was promised a paperless society when this whole internet thing started, so what gives?

In short, we in payroll are struggling to get everyone’s pay processed correctly and on time, but the system for doing so is incredibly inefficient, incapable of surviving the level of human error presented by barely-trained temporary Census employees, and compressed into a set of arbitrary and irrational time-frames that make actual prioritization of tasks or long-term planning impossible.  So some of us are doing the best we can, some people aren’t doing well at all, and are being fired ASAP, but ultimately I think we have to blame the planners.  There’s really nothing any of us on the ground can do to remedy the systemic problems that come from an unnecessarily paper-heavy and error-prone operation in which everything is rushed and the right hand never seems to know what the left hand is doing.

EXCLUSIVE: Former NYT Reporter Exposes Census’ Mishandling of Operations

Wednesday, April 29th, 2009

By Laura Mansnerus

Well, the census job in Philadelphia is over. They hired way too many of us. We finished a couple months ahead of schedule. They trained people for a week so that they could work for two weeks. We all miss the paychecks we thought we’d have. The Census Bureau goofed. Does Secretary of Commerce Gary Locke know this? Probably not.

My census job was a total accident. This winter, I was on hiatus from my career in the dying newspaper industry, having left my job to take a fellowship that, of course, came to an end. When I regained consciousness, the whole publishing industry was staggering toward a cliff. Uh-oh.

In January, I was prowling on Craigslist — and the Census Bureau was recruiting for the address canvassing phase of the 2010 Census: Work 20 to 40 hours a week for 10 to 12 weeks.

And before long I was a “crew leader,” hired at $19.25 an hour to supervise canvassers, known as listers, who would be verifying every address in my district of Philadelphia. All over the nation, listers would be updating maps and address lists so the bureau would know where to send questionnaires for the 2010 Census. Fine.

I loved the prospect of spending springtime on the streets while putting off a hideous job search. Moreover, I tend to believe that the public sector does important things, even though I once worked for the E.P.A., one of the worst government agencies known to mankind. The census is a worthy undertaking. The Constitution told us to do it! So I walked into this job with a pretty sunny outlook.

The crew leaders started in March. We were told we’d be working into June. The official deadline for the address canvassing operation, I learned later, is early July.

In my chunk of Philadelphia, seven of us crew leaders were trained the first week, and the next week we trained assistant crew leaders. The third week, each of us would train about 15 listers. Then we would parcel out “assignment areas,” or AA’s, smaller districts with 300 or so addresses each, to the listers.

My supervisor, Ian Hemphill, said we would be all working all-out, full-time and furiously, because the Philadelphia region was behind schedule. It took me a week and a half to start wondering about this. The crew leaders and assistants had already knocked out a few assignment areas. Even fumbling with the balky software on our little hand-held computers, you could finish an AA in a day. My district had 90 AA’s left to assign. I’d have 15 listers.

“Ian, am I missing something?” I asked him when he called a meeting of the crew leaders. “I have 90 AA’s left, and they’re sending me 15 listers. That’s 6 per lister.”

“Oh, ho, I think your math is a little off there,” Ian said.

“Wellllll, I have 90 AA’s, divided by 15, and that’s 6.”

“Let’s get our terms straight here. Six per what? Six per day? Per week?”

Another crew leader, a quiet young guy headed for graduate school, entered the conversation: “SIX … for … the … rest … of … the … entire … operation.”

We could finish the whole thing in a week. Given some slow listers and half-functioning computers, it would take two weeks. But Ian was not persuaded. We all went home.

Wait, I said to myself. The Census Bureau was about to screw a bunch of barely employed people. They had been told they would work until June. They wouldn’t. Worse, this is the agency charged with sophisticated demographic analyses for government and business throughout the nation — no, the world — and with 10 years’ planning time the people running it didn’t know how long it takes x listers to verify y addresses? I had a supervisor who couldn’t divide 90 by 15?

I went to Google News. All around the country, newspapers were running friendly features as address canvassing started in their communities. The Census Bureau, as many of these reporters had copied from its press releases, hired 140,000 people to verify 145 million addresses. Ooops. So they were doing this everywhere. They hired one person for each 1,000 addresses. In Philadelphia they had hired one per 2,000 or so addresses. How long could that take?

The next week, while our listers were chewing through AA’s, Ian trained a second wave of listers who would serve (if needed) as replacements. Meanwhile, we were encouraged to fire people wherever possible. I did not fire anyone. I got two more listers. I got 19 in all, though one was fired because he had failed the security check. (When the local Census authorities learned this, he was already on the job — your federal security clearance dollars at work! — so I had to confiscate his computer and badge immediately while we were out on the street. But that’s another story…)

They people who were hired were top scorers on the same test I had taken. Some had not worked for months and were hugely relieved to have jobs, any jobs. They took their mission very, very seriously. We got lost in discussions about how to treat houses that appeared to be abandoned or buildings that might be divided into illegal apartments, even though our supervisors did not want us to spend much time on concerns like this.

The listers were in their second week on the job when the AA’s dried up. My big crew leader map was plastered with check marks. The listers had put in their first full day on April 3, and most were on their last AA’s when the crew leaders were summoned to headquarters on April 15. The assistant manager for field operations, a mild sort named Wayne Wolfgang, who seemed to be trying to do the right thing, announced, “There’s rumors out there that we’re running out of work.” Not true, he said. “We have a lot of opportunities to move people around. We still have work for everybody.”

Since the crew leaders had been instructed not to open our mouths at this meeting, we didn’t. The office manager distributed a memo from the regional director, Fernando Armstrong, stressing the importance of meeting deadlines. Attached was a warning about driving safely, and anticipating hazards like black ice and moose. We were supposed to pass these out to our listers. And we did! You’re out of work, but if you had any, you should watch for moose on Washington Avenue!

Ian wanted us crew leaders to tell people, “We will find you work.” I ventured, “I’m telling them, ‘we might find you more work.’” Ian got mad. “No! We WILL find you work!”

Nothing materialized, though. Listers who needed to reopen their unemployment claims asked if the Census Bureau would provide letters to certify that they were out of work. No, The Census Bureau wouldn’t do such things.

Many people had passed up other temp jobs or even quit jobs to take jobs with the Census Bureau. “What am I going to do?” said one middle-aged woman whose other job was passing out samples at a liquor store a few hours a week. “I don’t know what I’m going to do,” she said, six or seven times. She gave me phone numbers of a couple market research firms that convened focus groups.

Others just wandered away. A few called or visited the office to deliver a piece of their minds. This was to be avoided at all costs. Keep them away from us, the last text message from Ian said. It also said, “Congratulate Listers [sic] on a quick accomplishment of their mission. Stress the positive. Stress the value of a period of federal employment with the possibility of recommendations.”

Today was my last day. I turned in the equipment I’d gathered up from listers, and on the way out of the office I saw Wayne Wolfgang. I asked him a question: “Why did the Census Bureau hire 20 people, including me, to do 97 AA’s?”

“Well, that was the number he was given by headquarters,” he said.

There was no mistake. We finished early. So what?

Wolfgang, apparently looking for congratulations until the end, said, “Did it ever occur to you that we’re efficient?”

“No,” I said.

Laura Mansnerus was an editor and staff reporter at The New York Times for 22 years. She was a 2007 Soros media justice fellow. She is also a no-longer-practicing lawyer.

Note: This article does not represent the opinions of MyTwoCensus.com, Stephen Robert Morse or Evan Goldin. The views expressed are those of the author. That said, MyTwoCensus welcomes written, video, photographic, and multimedia contributions from any individual with a 2010 Census-related story to tell.