The following is a guest column by Carl Graham, CEO, Montana Policy Institute:

Bozeman - MEA-MFT's Eric Feaver asserts in a recent open letter to me that the state employee compensation data at MPI's website is "...flawed and distorts the reality of state employee salaries and benefits." I think his frustration is misplaced.

While we simply presented data provided by the state and cannot independently verify whether it's "flawed," it's also true that the data would be much more useful if it contained greater detail, omitted certain things, and provided more context. We'd like that, too.

MPI waited over two years, won a lawsuit, and paid the state over $1,000 to obtain the pay data used in What you see is what we got, with one caveat. The 189 Mbytes of data - equal to nearly 120,000 printed pages - provided by the state broke down pay items by earning codes, and our original intent was to present them that way. But for reasons too technical to describe in 700 words, we simply were not comfortable with the validity of that breakdown. So instead we went with the state-provided year-to-date total figure that included most benefits and other payroll costs.

Mea Culpa on us for not explaining that better. But after months of delays and unexpectedly high fees from the state, we simply felt we had done the best that we could. That being said, the best MPI can do from the outside is certainly not the best that can be done.

This point is illustrated by Mr. Feaver's recitation of's shortcomings, with which I mostly agree. Let's take a look.

First, he points out that year-to-year pay increases due to job changes or as a result of going from a partial year's work to a full year's work are not highlighted and explained as such. Yep. The data we obtained does not include hire date or job changes, only current job and whether anything was paid that year. Breaking that down further would require the state to create yet another custom report at our expense. We had to draw the line somewhere on how much we were willing to pay the state to create electronic reports from their SABHRS personnel system, and that level of detail didn't make the budget. Why a system as comprehensive as SABHRS has to be reprogrammed at great expense to create custom reports with existing data is a different question altogether, and one we'd leave for the Department of Administration to address.

Next, Mr. Feaver points out that we did not disaggregate employee reimbursements for business expenses. I agree completely that expense reimbursements are not pay and would have omitted them if it were possible without compromising other data. But since we were unable to pick and choose pay codes we were forced to aggregate it all. MPI was clear in our court case that we were not interested in reimbursements, but since the state mingled them with the full data set we could not confidently omit them.

Mr. Feaver also notes that we did not disaggregate employee benefits and severance payments. Same answer as above. We wanted to disaggregate and display individual types of pay (other than reimbursements) all along but could not confidently do it with the data set that we received. Where we erred was in not making that tradeoff clearer in the website's overview and methodology notes.

Finally, Mr. Feaver asserts that we have an obligation to verify the data (with whom he does not say since it came from and resides at a single source: The Department of Administration), and to suspend until we make the changes he requests. Well, that's not going to happen.

What can and should happen, though, is working together as Mr. Feaver suggests to "...compel the state to create its own state employee pay site..." Getting the state to post spending online is something MPI has worked for unsuccessfully since 2008. State employee compensation would be a relatively simple place to start.

MPI and MEA-MFT working together on something like this would not only be a Disney moment with birds twittering and flowers blooming; it would also be a service to state employees and taxpayers. If the state publishes a timely, credible, and comprehensive website: one that includes information currently at while addressing Mr. Feaver's concerns, the pay portion of MPI's site will go dark the very next day. We can pull the plug on it together.