Reporters Notebook: The State Workforce Data Rabbit Hole

Telling a story with numbers seems like a straightforward endeavor.  That is, until numbers don’t match or add up.  I’ve been working with the state controller’s office to find out how many Idaho state employees have lost their jobs in the last five years.  (The story and the numbers are right here.)  But while I was in the process of gathering the numbers and asking questions, articles from The Times-News and The Spokesman Review’s Eye on Boise Blog were telling a similar story, but with a different set of numbers.

There are several complicating factors involved in telling the story of state government layoffs in Idaho.  First, who is a state employee?  Public school teachers aren’t considered state government workers in Idaho.  They’re contract employees.  Second, the data from the controller’s office is dependent upon each state agency reporting layoffs and retirements in a certain way, and those numbers can be flawed.  For example, StateImpact found the Idaho Department of Lands made coding errors from 2007-2009, categorizing the end of temporary firefighters’ employment as layoffs, instead of as seasonal jobs that came to an end.

Earlier this week, the Legislature’s budget director presented a slideshow to the Joint Finance Appropriations Committee highlighting the current number of state employees.  Budget Director Cathy Holland-Smith got her numbers from the controllers office’s annual ‘rainbow report’ (the same place StateImpact numbers came from).  And her presentation is where other news stories originated.  Holland-Smith told JFAC the number of active state employees is currently 24,198 and back in 2008 that number was 26,040.  That’s a drop of 1,842 state employees.

StateImpact reported the current number of state employees is 24,383.  That’s compared to 25,411 in 2008, a difference of 1,028.  So why the disparity?  Before 2009, the state controller’s office was using a different computer system to gather this data.  The data for the previous year would be rounded up just after the start of the new year.  After 2009 that changed and the data was gathered at the end of the year.  Holland-Smith got her 2008 data from that original report, rather than from the latest 2012 rainbow report, which I relied on for my reporting.

So which numbers are right?  Well, the controller’s office says they both are, or were, at the time the data was gathered. Every rainbow report issued since 2009 lists employment data taken from the same snapshot in time.  The bottom line? The controller’s office says the most accurate data is always from the most current report.

Either way, there are fewer state jobs today than there were in 2008.  Using the most current numbers, about three-quarters of the job losses between ’08 and ’11 resulted from layoffs.  The rest were due to retirement, or people leaving their jobs for other reasons.