Follow the Money: $40+ Million Gone Missing

[Note:   As we continue to await a full response from Richmond Public Schools Superintendent  Yvonne W. Brandon and RPS Chief Academic Officer Victoria Oakley concerning the recently released math SOL results that (incredibly)  had some RPS schools with only single digit pass rates, it should come as no surprise that basic math skills are apparently a challenge for the top RPS administrators. In anticipation of updated data due any day now from the Virginia Department of Education (VDOE), please take a moment to read John R. Butcher's trenchant analysis of last year's information. If you have any questions, please send them to us in the "comments" section and we will do our best to get you some answers.] 

The 2010-11 data on the Education Department Web report the total disbursements for each school division.

In terms of the total disbursements (corrected to remove payments to and from reserves, payments to upgrade or construct facilities, and debt service, the latter of which we are told Norfolk does not pay from the school budget) and the end-of-year ADM, we spent $2,957 per student more than the state average and $2,723 more than Norfolk. 

	$/ADM	v. State	v. Norfolk
Norfolk	$     11,130	$     234
Richmond	$     13,853	$  2,957	$    2,723
State	$     10,896

Responding to the down economy, these numbers are reduced from last year's, especially for Richmond:

Division	$/ADM	v. State	v. Norfolk
Norfolk	$ 11,928	$     782
Richmond	$ 14,672	$  3,525	$    2,743
All	$ 11,146

Notes: "ADM" is "average daily membership," which is educratese for the daily average enrollment.  This VDOE report uses the end of year number.  The disbursements in the table here are 2010-11 totals divided by the ADM.  The "State" number is for all school divisions; it omits the Governor's schools and regional programs. You might well ask why it took until April, 2012 to get the data for the 2011 fiscal year.  The answer is VDOE does not post some of the data it receives in October until about the following March.  Please be my guest and ask them why it takes so long.  (And for a larger delay for other data, see below.)

Multiplied by our ADM (21,432), the 2011 difference calculates to $63.4 million in disbursements above the state average (and $58.4 million above Norfolk).
 

Explanations Please!

How shall we explain the very high cost of the Richmond system?  It turns out that, even if we correct for all major differences between Richmond and the State, about $50 million of excess Richmond school expense is just that: Excess.  The data (and a string of audits) suggest that Richmond is wasting that very large amount of money.

Follow the Money

Let's start by looking at the categories in the State report.  Here are the Richmond and Norfolk expenditures by category (with debt service, contingency, and facilities removed), expressed as percentages of the State average disbursement.

Note: The total disbursements are the day school expenditures (administration, instruction, attendance & health, pupil transportation, and O&M) + food service + summer school + adult education + preK + other educational + facilities + debt service + contingency reserve, if any.  See the footnotes to the State's Table 13 for the details about these categories.  The numbers above omit facilities, debt service, and contingency reserve.

Richmond's costs are well above average for every category except administration and transportation; our costs are extraordinarily high for summer school, preK, and for other instructional programs.

If we recast the data in terms of dollars, we see attendance, food, summer school, and adult ed. to be relatively minor players, with the excess costs of instruction, O&M, preK, and other ed. predominating.

Most of that "preK" category is Head Start.  In the context of the current debate about funding public preschool education, it would be interesting to find out what, if anything, Richmond is getting for the extra 9.7 million dollars (plus the other $4.9 million for "other" education).  For sure, a major study suggests that Head Start doesn't do any good.

Salary Differences

One source of these excess costs might be instructional salaries.  Ours, however, are lower than the state average.

	All Instructional Positions	District Wide Instructional Positions
Norfolk	$       49,852	$       45,739
Richmond	$       49,947	$       50,610
State	$       52,780	$       46,260

Note: The State defines (fn 3) "all instructional positions" to include "classroom teachers, guidance counselors, librarians, principals, and assistant principals."

Versus the State, multiplication of the $2,833 difference by our 2,223 instructional positions shows a saving of $6.3 million from Richmond's lower salaries. 

Note the interesting anomaly in the salary data: The "District-Wide" salary is for "division-wide technology instructors and instructors in programs such as summer school and adult education."  That is, teachers.  One would expect the district-wide salary to be near the average salary in the division for classroom teachers.  For Norfolk and the State the division-wide salaries are, in fact, lower (by $6,521) than the average instructional salary.  But in Richmond, the division-wide salary is $663 higher than the average for all instructional positions.  So we're paying a premium for these division-wide positions.  How many of Richmond's expensive bureaucrats do you suppose are masquerading as division-wide instructors?  Indeed, the district-wide excess over the state average, $4,350, suggests that Richmond was concealing ca. $0.78 million worth of administrative salary as "district-wide instructors."

Number of Teachers

As to the state average, another source of excess cost is number of teachers.  Here are the State's numbers per 1,000 kids (again, 2011 data from preliminary Tables 17b and 19; the final 2011 data are not ups as of 6/12/12(!)):

	District-Wide Instructors	All Instructional Positions
Norfolk	7.39	108.86
Richmond	7.80	115.69
State	2.47	102.53

Compared to the state average, we have an extra 13.2 instructional personnel per thousand, i.e., 305 for the school division.  At our average salary of $49,947, that calculates to $14.1 million in extra cost per year. 

Notice that we have 3.2 times the state average number of district-wide instructors.  And, as we saw above, they are very expensive.  What do you think: Great teachers, or concealed bureaucrats?

Corrected for Salary Difference and Number of Teachers

Applying these corrections to the cost differences, we obtain:

Category	XS v. State
Gross Excess	63.4
Salary Diff.	6.3
No.Teachers	-14.1
Net Excess	55.6

That leaves $55.6 million in excess spending unaccounted for.

Special Education, Poverty, and Geography

We often hear that special education and poverty are major contributors to Richmond's higher costs.  It is difficult to come by numbers to quantify that so I turn to SchoolDataDirect, the now defunct data service from Standard & Poors that was supported by the Gates Foundation.

S&P calculated adjustments to spending on core operating activities to account for poverty, special education, and geographic factors.  On the 2006 data, here are their numbers

	Core Spending, $/kid	Core Spending Adjusted	Factor, %	v. State
Norfolk	$       8,400	$        5,000	60%	0.7%
Richmond	$     11,181	$        5,927	53%	-5.8%
State	$       8,598	$        5,059	59%

The difference in the adjustments is 5.8%, or 9.9% of the State correction.  In short, on core educational spending, SchoolDataDirect calculates that Richmond is 9.9% more expensive to run than the State average (and 11% more expensive than Norfolk).

What we are doing here is broad-brush approximations, so let's apply the 9.9% to the total expenditures, after correction for the differences in salary and number of teachers:

Category	XS v. State
Gross XS	63.4
Salary Diff	6.3
No. Techers	-14.1
Net XS	55.6
Cost Adjustment	-5.5
Corrected XS	50.1

After these adjustments we are left with an excess in our school budget of some 50 million compared to the state average (recall that this does not account for disbursements for facilities, debt, or contingency).  (If we were to apply the 9.9% to the gross excess, it would reduce the corrected excess to 49.3 million).  Until the RPS can explain where all that money is going and what we are getting for it, there is no reason at all to increase the school budget.  To the contrary, these data make a good case to cut the budget by at least $50 million.

O&M

Richmond often claims that its infrastructure is larger than Norfolk's, reflecting Richmond's "small school" policy.  Indeed, there are 106.5 principals and assistant principals in Richmond but only 100.3 in Norfolk, although Norfolk's ADM is almost 50% larger (31,111 v. 21,432).  Richmond has never demonstrated any educational benefit from this relatively larger number of schools, however.  And they've been right slow to do anything about the excess 4% of the capacity in elementary schools and 19% in middle and high schools. 

The empty seats in the Richmond system suggest that much of Richmond's extra infrastructure is wasted space.  Thus, much of Richmond's extra cost for infrastructure probably is wasted money.

In any event, Richmond's excess expenditure for O&M is $7.4 million.  Even if all that is justified, Richmond is spending an excess $42.7 million.

The Bottom Line:  $40+ Million Gone Missing

Most of the $42.7 (or $41.9) million extra spending relative to the state average must be in those categories where Richmond vastly outspends the state average: Instruction and "Other" educational spending or preK (gross excess $51.9 million).  The 09-10  budget (pdf, ~1 MB download) shows about $6.8 million in grant income for the Head Start program, which would fund two thirds of the excess in preK activities.  There is no information to show that we are getting any value for that money or for the remaining $8 million excess expense of "other" and preK instruction that apparently comes out of local and state taxes.  In any event, the largest pools of unexplained expenditures are in the instructional budgets and those should be the first places to look for savings. 

Without a zero based budget, it probably is impossible to find the waste in the instructional categories.  Doubtless that is the reason that the City Auditor's forays into the RPS budget have targeted the low-hanging fruit in the administrative operation: In June, 2007, the City Auditor released an audit that identified potential savings of $16.7 to $19.8 million.  About 20% of that ($3.4 to $5.3 million) was in non-teacher instructional staffing; most of the rest was non-instructional categories (notably custodial outsourcing and vehicle replacements).  In April, 2008, the City Auditor released an audit that found a further $6.7 million of waste in the RPS' purchasing and accounts payable operation.  A December 2008 audit concluded that RPS fleet maintenance contract lacked provisions that would allow RPS to make necessary cost control and other operational decisions.  

A February, 2009 audit contains a four-page list of deficiencies in the Schools' IT program, including an absence of monitoring procedures that would assure that RPS was receiving proper value for its IT budget.  An August, 2009 audit of RPS's grants administration found that the lack of proper monitoring of grants poses a "major risk."  If the auditor would look at the incompetent (or worse) expenditures for ADA compliance (see the discussions here and here), he would find still more waste.  Even so, the (very rough) estimate here suggests there is lots more waste to be found and that the instruction budget is the place to find it.