We have (in theory, anyway) a representative government by which the people elect officials whose job it is to run the government in the way they prefer. In order to maintain some consistency, the government hires professionals who are supposed to implement and assist the representatives’ policies objectively.

So, you hire a Ph.D. economist to analyze the government’s data and to objectively explain it to the elected officials (who aren’t necessarily expected to be deeply trained in economics), and that’s what you get. Right?

Sometimes I wonder.

State Director of Revenue Analysis Paul Dion (who recently received a retroactive pay raise of $5,000, bringing him to $125,610 per year) mounted a defense of Governor Lincoln Chafee, tonight, as a member of the legislative commission to study the elimination of the state sales tax. In doing so, he made one of the several surprising statements that livened up the hearing for the handful of people in the room:

We started [the governor’s term] four years ago at eleven-something percent [unemployment]. Now we’re down to nine. OK, nine-one. I don’t know why we focus on one blip, but the reality is that consistently the [federal Bureau of Labor Statistics] has underestimated the number of jobs in the state, year after year after year. 6,500, they were off by, on June 30th. When we look through, the revisions are coming; they do it every single year. We’ll get the revised data, and things will look far better.

When the governor took office — and I’m not defending this particular governor, because I’ve worked for different governors, for the record — but when the governor took office, the projections were coming back to us at 13-odd percent [unemployment], and we never broke 12, and that’s because the BLS data is faulty. …

When the governor took office, that unemployment rate was 11.9… now it’s down to 9, so to say that he’s not done anything or tried to move things forward, I just don’t think that’s a fair characterization of the governor’s [inaudible].

It’s difficult not to read this as spin from a government employee defending his current boss and, by extension, the whole hierarchy of career government officials.

It’s true that there’s periodically been some difference between the BLS’s monthly count of RI-based jobs and the numbers that the state Department of Labor and Training sometimes releases, based on data as soon as it’s available. (See here for a graph of the difference.)

But Dion is surely aware, though his comments are at best unclear and at worst misleading about it, that the jobs data that the DLT was correcting is from a different source than the employment and unemployment rates. RI-based jobs numbers come from surveys of employers, checked against tax records. Employment and unemployment are from a survey of households.

Yes, for estimating purposes, the jobs data is one of the benchmarks that the BLS uses to adjust its employment data. And yes, it is subject to significant revisions. But when the BLS released revised data for the years 2008 through 2012, some of the revisions of the unemployment rate were positive and some were negative. The revisions for 2012 were the most dramatic, at an average of 0.33 percentage points. That’s the difference between 9.0 and 9.3, not 9.1 and 13.0.

That gets to the most disappointing part of Dion’s harangue. He makes it sound as if BLS data flaws and the governor’s activities are to credit for lowering the state’s unemployment rate. That is simply false.

The real reason that the unemployment rate never topped 12% is that Rhode Islanders gave up looking for work and dropped out of the labor force. Period. If they had not done so since the recession began, unemployment would have peaked at 13.9% at the end of 2011 and would now be a little over 13%. If our labor force had grown (as our last-in-the-nation peer, Nevada, has experienced), unemployment would actually be worse.

I’ll be updating this chart tomorrow, with December data. The blue line is the reported unemployment rate; the red line is what the unemployment rate would have been if the labor force had remained at its January 2007 level; and I’ve thrown in the the pre-revision numbers for 2012.  (Note that, through 2012, the blue and red lines reflect the revised data.)

Dion didn’t specify when the 13% predictions were being made, so to be fair, I’ll acknowledge that holding the labor force at its position when Governor Chafee took office changes the peak, but it’s still not a picture of growth.

Either way, the lesson isn’t pretty. Rhode Island is improving its employment statistics by driving people out of the labor force. That’s not a healthy trend, and one of the state’s top payrolled economists shouldn’t be trying to sell the story to legislators and the public.