Later, Sooner

Ruins at Mount Alban outside of Oaxaca, Mexico ID’d by Ryan Mahoney

Over $66,000!  Did you know that’s how much teachers are paid in Oklahoma?  Well I didn’t either until I saw it in the 1889 Institute’s report “Teacher Pay:  Facts to Consider”, which I read about in a blog post on the Oklahoma Council of Public Affair’s website.  I debunked the number in my post Miserables Love Company, but it is likely a number that will get repeated and taken as gospel by many, including policy makers, in this state.  That’s how alternative facts get legs.  Someone with seeming credibility, in this case the authors have at least four college degrees (including a PhD in economics) between them and are employed by a think tank which presumably does real research, puts out a number, which then gets restated by another think tank’s publication which gives it another layer of credibility, and then gets picked up and restated by others again and again.

Earlier this week I corresponded directly with one of the number’s fabricators, Byron Schlomach, who is also listed as the 1889 Institute’s Director, to see if he would share the “Author’s Calculations” that mysteriously grew their starting point from $44,921 to $66,034.  He actually replied, not with bluster and threats as did the Limited Thinker in Chief at the OCPA (see You’re Not in Kansas Anymore), but with actual information which I appreciated.  Here is his explanation:

“Table 236.50 of the 2016 Digest of Education Statistics from NCES provides total salaries and total benefits by state for 2013-14. Looking at the total benefit number, it was clear that it did not include the expenditure by the state on health insurance. Adding that in ($339,243 in thousands) and dividing the resultant total of benefits by total salaries, the percentage comes out to 47.7%. I then multiplied the reported average salary, $45,751, by 1.477. I don’t think your number comes out that much different from mine if you add in the federal contribution to retirement, but any such number is rough. The intention, as stated in the paper, was to estimate the total cost of an average teacher to taxpayers, not to estimate take-home pay. If the rest of the analysis had depended on that number I’d have tried to be more precise. Your number may very well be a better one, but I doubt that, absent our number, yours would have been any less “eye-popping.””

Notice the underlined sentence above, suggesting their work was to estimate “cost…to taxpayers”, not “take-home pay”.  True, there are different ways to look at it, but if the $66,034 is meant to communicate about “cost”, not “pay”, then why does his “report” say it like this:

When benefits are included, pay increases considerably, by about 47 percent, or $21,113 for a total average cost per teacher of $66,034.  This takes average hourly teacher pay to $39.45.

I’m sorry, Byron, I don’t think your calculation is intended to inform policy makers about what an average teacher costs taxpayers, I think you want to put out a number for “average hourly teacher pay” that will cause many Oklahomans to think, “Wow, teachers are paid a lot more than I am, so what’s all the fuss about?”

In addition to what I point out in my calculations (see Miserables Love Company), here’s why he’s so far off.  He says “Looking at the total benefit number, it was clear that it did not include the expenditure by the state on health insurance.”  Actually it is clear that the total benefit number does include “the expenditure by the state on health insurance”.  I followed his link to Table 236.50 (tabn236.50.nces) in the Digest and the most recent amounts I found were for FY 2013, not FY 2014 as he states.  That really doesn’t matter because the proportions don’t change dramatically from year to year.  The table says Oklahoma spent $2,073 million on instructional salaries and $637 million on instructional benefits yielding a benefits/salary ratio of 30.7%.  My average teacher calculation, based on FY 2016 because that’s what the “report” said, yields a ratio of $14,426/$59,347 = 32.1%, pretty close to the 30.7%.

The NCES table data must come from Oklahoma’s OCAS reporting.  The OCAS system uses object codes in the 100s for salaries, stipends and other forms of direct compensation to school personnel; it uses object codes in the 200s for employee benefits, such as employer paid health insurance which is code 213 for teachers.  The FY 2014 OCAS statewide totals (ExpenditureReportFromOcasStatewide2014) for instructional code 100s is $1,929 million and $632 million for code 200s which yields a benefits/salary ratio of 32.7%, again very close to the other ratios I calculate above.  It is very clear that the NCES data for benefits absolutely includes every dollar the state, through school districts, expends for teachers’ health insurance.  So where does Byron get his $339,243,000 that he adds to the NCES benefits total?

I don’t know except to conclude he doesn’t understand how school employee insurance is funded.  Either by law or custom the legislature in most years (they didn’t for FY 2011 which led to then Superintendent Barresi trying to short-fund teachers’ health insurance) funds school employees’ Flexible Benefit Allowance with a specific line item appropriation, one for certified personnel and one for support personnel.  The earliest line item amounts I could easily find, shown in the State Department of Education budget, on its website are for FY 2015; again these numbers change year to year but not dramatically, though the FBA has steadily taken a larger share of the overall  budget.  The line items that year are $263,634,696 for certified employees which are mostly teachers, and $143,648,937 for support personnel which includes many teacher assistants whose salaries are also included as “instructional”.  So I can see Byron finding a number somewhere saying that $339 million out of $400+ million expended on health insurance was “instructional”, maybe using the OCPA’s data tool; anyhow that’s plausible.

But, but and but, the health insurance payments are already in the NCES and OCAS “Benefits” numbers.  The legislature appropriates it to the State Department of Education, which in turn pays it over to the school district employers, who in turn pay the premium amounts (coded 213 and 223) directly to the Employees Group Insurance Department of the Office of Management and Enterprise Services of the State of Oklahoma, and that is all that is paid, there is no other “expenditure by the state” for school employees health insurance.  So our Byron is a double counter.  His “it was clear” adjustment takes the benefits/salary ratio from the just over 30% that is supported by the data to his silly 47.7% which is not, but certainly helps support the 1889 Institute’s narrative that teachers are paid and cost a lot.

Byron’s bio says he’s an Aggie from Texas A&M, but his employment by the 1889 Institute now makes him a Sooner, and we fourth generation (my grandmother remembered seeing settlers headed to the 1889 land rush when she was a girl in Oklahoma living in a sod house) Okies know what Sooners were—land rushers who played fast and loose with the law, just like Byron is playing fast and loose with teacher pay data.  A PhD economist should know better than to double count, but he didn’t and my next post will show how he double counts again in his Report “Saving Money:  School District Consolidation vs. Breaking Up Big Districts”.  So…later, Sooner.

As always lunch on me for the first to ID the photo location.