Dat A Tool? Thinker Exposes a Frank Distortion

Artist Cha’ Tullis hilltop sculpture of Indians on horseback Hominy, OK; be sure to visit his gallery in town.

This Thinker just finished reading “A Higher Loyalty:  Truth, Lies and Leadership” so a part where James Comey shares some regrets from his college days came to mind as I decided on a sophomoric title for this post, namely that after writing previously about the “Texas Envy” rampant among the fellows at the Oklahoma Council of Public Affairs, I needed to more carefully examine the OCPA’s tool, “Data Tool” that is.  Go to their website homepage, select “Projects” at the top, then “Data Tools”, and lastly “Oklahoma Education Spending and Revenue”.  It used to be prominently displayed, but now is more hidden, so you can miss it and I’ll next explain why you should.  Also, if you bear with me to the very end, we’ll find what I think is the source of Frank Keating’s $300 million distortion when he wrote “In fact, through fiscal year 2016, Oklahoma public schools received more than $9.2 billion in revenue, a near record high.”  ( Takes One To Know One  and  Wayne’s World )

The “tool” purports to provide data about “Education spending, student enrollment, and spending per student from 2005-2006 to 2016-2017”.  I’ll use the “Note” following to organize my analysis.

OCPA says:  Data for this application are compiled from the Oklahoma Cost Accounting System (OCAS) and provided by the Oklahoma State Department of Education (OSDE). 

My comment:  I am intimately familiar with the OCAS used by school districts for financial reporting to the State Department of Education.  It is the best source for financial data, though it can be easily misunderstood as it often is by the OCPA.  Here is a table that intersperses expenditure and revenue statewide totals from the OCAS site and the OCPA’s “data tool”.

You will note that for no year, neither revenue nor expenditures, does the OCPA “data tool” report the correct number from the OCAS data.  “Compiled” is the key word here.  They couldn’t just provide a more user friendly access to OCAS data, rather they had to “compile” it to fit their narrative about how Oklahoma schools have lots of money to spend.

Some of the differences we’ll try to understand later, but for now look at the “ocpa revenue” column compared with the “ocas revenue” three columns to the right; the numbers differ greatly.  Part of the explanation is that the OCPA ignores the obvious, that “nonrevenue” is not part of total revenue, and adds it in anyhow.  I have written about this many times (see  Double, Double, Toil and Trouble  ) so won’t elaborate except to point out that borrowed money, i.e. a district selling bonds to finance construction of a new school, is not income, rather it is an advance of income, local property taxes, to be levied later to retire the debt.  There is no reason to include the “nonrevenue” totals as part of total revenue except ignorance or deception—if the shoe fits…

OCPA says:  Visitors to the OSDE website will notice a sharp drop in expenditures after the 2015-2016 school year. This is due largely to HB 1202, which codifies NCES current spending as the official definition of per-pupil spending. 

My comment:  I noticed this change in the “State Reports” and commented on it in my post  The Bice Is Not Right earlier this year.   Here’s the SDE page we’re talking about.

Here is the 2016-2017 Expenditure Report  I assumed it was an error in uploading the data and would eventually be corrected.  Simply stated no Function 4000 or 5000 amounts are reported for FY 2017, but they are for all the prior listed years.  I see nothing in the legislation that requires this omission.  It only requires that “per pupil spending” be reported using “current expenditures” which means, of course, eliminate double counting by exclusion of Function 5000, and include no “capital” expenditures being Function 4000.  But defining the per pupil spending calculation that way doesn’t require not reporting those excluded transactions.  Anyhow my 2017 “ocas expend” amount is way low for that reason—OCAS simply does not report those functions for 2017 on the state reports.   However, check this out.

SDE does report the data both ways for each individual district, namely “District Expenditures Report per HB 1202” which omits functions 4000 and 5000, and “District Expenditure Report” which includes them.  I expect that eventually SDE will add a report to its State totals that will again include those functions. 

OCPA says:   However, in order to maintain transparency and year-to-year consistency, this application will continue to report nearly all spending. 

My Comment:  I assume the OCPA must have called SDE and obtained the missing data so their “tool” remains consistent. 

OCPA says:   (One notable exception: hundreds of millions of dollars in annual state contributions to the Oklahoma Teacher Retirement System. This money bypasses the school districts’ payroll and is sent directly to TRS.) 

My Comment:  There are two payments to OTRS that bypass school district payrolls.  One is budgeted for by the State Board of Education, $27.7 million in FY 2018, which pays a percentage of all teachers’ required 7% contribution, according to a years of service scale that was part of a compensation package many years ago.  It requires effort to continue, ought to be eliminated with some appropriate offset, but is truly part of current payroll costs.  So that statewide amount does escape notice in most calculations.  If the OCPA fellows are referring to this, they are correct, sort of.

The other payment is from the State itself, a dedicated 5% of four major revenue sources being sales, use, corporate income and individual income taxes.  In FY 2017 that amounted to $278.9 million—a very large sum.  You can see in this excerpt from the FY2017 OTRS Actuarial Report where it fits in; note the $27.7 million “state credit” is referenced as part of “Member contributions”.

In the lead up to the 2017 legislative session I wrote a couple of posts (see What’s Up Doc? or Should Teachers Eat A Carrot?  ) explaining how all this fits together so I’m not going to repeat it here.  Bottom line is that for FY 2017 out of the essentially $1 billion ($998 million) in “Contributions”, only $450 million of that was needed to support the retirement commitment to teachers and other school employees currently employed; their contributions paid for $293 million of that with the balance of $157 million coming from the employers and/or grant program contributions.  That means that ALL of the State’s direct contribution, the $278.9 million, and $271.1 million of employers contributions went to pay for the State’s unfunded future obligation to current retirees, like me.

So if we want to be accurate here in determining what it costs to fund Oklahoma’s public schools, then we should include the $27.7 million from the SBE budget paid in to OTRS; but we should not include the “hundreds of millions” the OCPA refers to AND we should subtract about $270 million from payroll costs, BECAUSE if the State didn’t spend another dime educating its children going forward, it would still bear the legal obligation of its past retirement promises costing the $550 million per year for an estimated 17 additional years. 

OCPA says:   All spending and revenue graphs and tables on this application are adjusted for inflation using the CPI Inflation Calculator of the U.S. Department of Labor’s Bureau of Labor Statistics. 

My Comment:  In other words the numbers in their “tool” are “compiled”, i.e. distorted, by being multiplied by an inflation factor.  They don’t disclose the base year.  They don’t disclose how often they engage in this silly exercise.  They don’t examine whether the CPI is even a relevant measure of inflation for school district expenses.  This silliness makes their tool of no use except to intentionally distort what public education receives in revenue and expends each year—and their intent, as shown by their consistent double counting and other misrepresentations, is to inflate the actual numbers.

Look again at the table I compiled above.  The last two columns, “cpi rev” and “cpi exp” are the result of dividing “ocpa revenue” by “ocas rev+nonrev” and “ocpa expend” by “ocas expend” respectively.  The results should be the CPI inflation factor used to inflate each OCPA data entry from whatever mysterious base year they are using.  If their (and my transcription) math is correct then the two columns should be identical from 2006 through 2016; 2017 is distorted as discussed above.  Interestingly only five of the eleven years match and my hunch is the OCPA fellows simply botched the annual adjustments they make to this data.  Also, unless prices actually declined year over year, the numbers should rise from 2016 through 2006.   They do except between 2009 and 2008, the Great Recession, so that deflation makes sense.

One of the few external realities that made managing school district finances from 2006 to 2016 easy was the absence of inflation like that which rocked our economy in the 1970s.  When we went about the business of projecting expenditures and revenues general inflation measures were simply not relevant.  Student population growth mattered; legislative appropriations mattered; employee compensation changes, namely none, mattered; the closing of a major manufacturing facility mattered; but not once did I think to see what the CPI index was up to—it just didn’t matter.  Using it to distort the real numbers does nothing helpful to assist policy makers.

Here’s what it does do.  It provides some plausible justification for Frank Keating to write that school district revenues in 2016 were “more than $9.2 billion” when in fact, even using the OCPA’s double over counting, they were $8.812 billion (my table, ocas rev+nonrev).  But, you say, the ocpa revenue number for that year is only $8.964 billion.  Here’s what I think Frank did.  He was writing in February, 2018 so when he asked for the 2016 revenue amount, which is the OCAS number $8,812,222,077, the clever fellows at the OCPA couldn’t resist inflating it by the CPI.  Here’s the CPI Inflation Calculator for January 2016 to January 2018.

Multiply that real OCAS number (includes double counting) by 1.04622 and, drum roll please, we get $9,219,522,981—or, “more than $9.2 billion”.   I think Frank’s distortion is solved, and it was simply an indefensible distraction from what was a very serious statewide discussion about the future of our public education system.

What is really sad here is that the OCPA has invested some resources in having the OCAS data in an online program that could provide some useful and interesting features for policy makers.  But it is of no practical use because the numbers are distorted for no reason other than to exaggerate the investment Oklahomans make in public education.

Fantasy Epilogue:  An OCPA fellow goes into his year-end 2018 performance review with Frank as chair of the personnel committee armed with the latest CPI Inflation multiplier being 1.03 (3% inflation).  The fellow, whose salary is $100,000, states that his salary should be increased to $103,000.  Frank responds, “Well that’s perfect because I see here that, according to our “OCPA Salary Data Tool”, that adjustment has already been made!  Thank you for your service.”

As always lunch is on me for the first to ID the photo location, a blown up version of the last one.  It’s just too good to go unclaimed.