Lies, Damn Lies and Statistics in the Virginia Department of Education – Average Teacher Salaries Edition

by James C. Sherlock

I was in the early stages of researching a column on school salaries in Virginia when I came upon yet another bad report.

In 2021 Special Session I, the General Assembly directed the Superintendent of Public Instruction to provide a report on the status of staff salaries, by local school division, to the Governor and the Chairmen of the Senate Finance and Appropriations and House Appropriations Committees.

The appropriations committees wanted to know how much teachers and others were getting paid so they could raise the state contribution. It would seem to be a report that VDOE would like to get correct.

As with many other reports I have documented, the January salary report on its face cannot possibly be correct. VDOE and thus the Governor and General Assembly have no idea how much teachers and other instructional staff are paid in Virginia.

This report was a parting gift from the Northam administration.

The question itself — average salaries — may prove not to provide information useful for legislation and appropriations however accurately it is answered.

Presently the House and Senate are at odds over whether school salaries get a 4% or a 5% bump from the state. Those are across-the-board increases, not targeted at specific types of teacher certification that are in critical shortage. Why? Perhaps because, as I documented earlier, VDOE’s teacher shortage data are also useless.

So, it appears that those ubiquitous rankings of teacher salaries among states are equally flawed.

More importantly, any attempt to compare school salaries in various jurisdictions in Virginia is frustrated. So too is any effort to compare salaries to cost of living in each jurisdiction.

That is as important as the salaries themselves.

Finally, if the elected officials of the 132 school divisions use these data, they will not know whether and how much to raise local contributions, especially in any attempt to match salaries in adjacent jurisdictions.

I’d say it can’t go on like this, but it can — and has. And it really matters.

But there are hopeful signs.

The report was sent to the General Assembly on January 12 of this year by Rosa Atkins, Acting Superintendent of Public Instruction, in the last days of the Northam administration. It is titled “2020-2021 Teachers Salary Survey Results” dated Jan 4, 2022.

The source of the problem. The basic problem appears to be that the Department of Education staff prepared the report based on Annual School Report (ASR) expenditure data submitted by each school division and regional program.

From the report:

Department of Education staff prepared the report based on Annual School Report (ASR) expenditure data submitted by each school division and regional program. Statewide and school division salary averages were computed for fiscal years 2021 and 2022. The report shows salary information from fiscal years 2020, 2021, and 2022 for comparative purposes. Year-to-year percentage change in salary averages is shown for statewide and division data.” (Emphasis added).

If someone had actually looked at the results of those calculations they would have seen that they were fatally flawed. Maybe that process should have worked if the input data were accurate, but it did not because they were not.

Failed the test of reasonableness.  Some examples of report data in which Year one is FY 20 to 21 and Year two is FY21 to 22.

Teacher salaries:

  • Essex County Public Schools showed a decrease of salaries by 15.93% in year two.
  • Greensville County showed an increase of salaries 8.82% in year one and another increase of 23.28% in year two.
  • Isle of Wight County showed a decrease of 19.14% in year two.
  • Middlesex County showed an increase of 43.84% in year two.
  • Prince Edward County is shown with the data blank after FY 2020.
  • Buena Vista City Public Schools is shown with a decrease of 12.96% the first year and another decrease of 15.99% in year two.
  • Norfolk schools showed a decrease of 15.77% in year two.
  • Governor’s Schools display some enormous swings in average salaries, some in excess of 20% year over year.
  • Piedmont Regional Education Program showed a 155.3% increase in average salaries in year two.
  • Three of the Regional Alternative Educational Programs showed increases of over 100% in average salaries in year two.

Salaries of other instructional personnel:

  • Brunswick County is shown decreasing the salaries of its Assistant Principals 60% in year one.
  • Middlesex County is shown increasing the salaries of its APs by 43.8% in year two.

What to do?

So, what are we to think of the rest of the data? Of what use are overall averages computed from those data by VDOE and reported to the General Assembly? The answer is “none.”

I don’t know what plan B should be to ascertain average salaries, but using these ASR budget expenditure data to calculate them won’t work unless the divisions submit better baseline data.

And even correct answers won’t provide much information because they leave un-illuminated the differences that are due to the effects of time in service and graduate degrees on the compensation ladders.

Perhaps the question needs to be reframed or some other data used as a baseline. Or perhaps the question itself — average salaries — is not one the answers to which will be useful in the appropriations process or in any other form of decision making.

One suggestion, ask each district for their pay scales for teachers, principals, AP’s, etc.  Pick a common point or two on each scale, say a teacher with an certificate but no graduate degree with four years of service or whatever and compare them across the divisions.  It will get much closer to the information sought than the current system.

A final point on what to do. There is no indication in the report that the numbers of personnel in any classification (teachers, principals, etc.) in each division is accounted for in the state average salary. Averaging Franklin City with Fairfax County without that factor yields useless information if you wish to know what the average teacher in Virginia is paid.

Be that as it may, I have three observations:

  • VDOE staffers should never have accepted those numbers on face value, much less averaged them and prepared a report for the Governor and General Assembly.
  • The Secretary of Public Instruction should have never signed it.
  • The General Assembly should have sent it back for correction.

The Youngkin Administration is the unhappy inheritor of this long-term cancer on public schools policy.

Good news.

Yesterday I received, unsolicited, a personal note from the highest levels of VDOE. In it my correspondent recognized the problem presented in my columns and promised VDOE management is taking on the task of fixing it.

I received a similar note from the leadership of Richmond Public Schools that acknowledged their own problems with data reporting. They promised to not only fix it going forward but also to go back and correct the staffing vacancy report that I recently called out.

It is a very hopeful sign that both VDOE and RPS acknowledge these issues, because, as we see above with average salary reports that were sent to the appropriations committees, they matter a lot.

But there are 131 other divisions, some of which take their reporting seriously and some clearly do not. Nothing new in that.

It will be a tough slog.

But the acknowledgements represent a breakthrough, and the citizens of Virginia thank VDOE and RPS for them.

Updated May 22 at 7:04 AM.


Share this article



ADVERTISEMENT

(comments below)



ADVERTISEMENT

(comments below)


Comments

24 responses to “Lies, Damn Lies and Statistics in the Virginia Department of Education – Average Teacher Salaries Edition”

  1. Timothy Watson Avatar
    Timothy Watson

    Did they take x expenditure amounts for y positions and calculate the averages that way? And the bizarre numbers you highlighted are because a position sat empty for 6 months with zero expenditures for the position, but the position was still used in the average?

    1. That is exactly what I was thinking.

    2. James C. Sherlock Avatar
      James C. Sherlock

      I can come up with three or four scenarios for how they got it wrong. But it does not matter. It is wrong. By a lot.

    3. Dick Hall-Sizemore Avatar
      Dick Hall-Sizemore

      The position was used in the average for the second year because that is the “budgeted” amount, not the actual expenditure. The budgeted amount accounts for salary increases, replacing experienced, retired teachers with lower paid new teachers, increases for experience, and filling all positions.

  2. Good start on digging into the “average” pay numbers, Jim. Mr. Watson nearby raises an interesting issue — does the average for each school district incorporate unfilled positions which, by virtue of being unfilled, are paying zero dollars? That may explain the wide swings in some of the smaller school districts as the the number of unpaid positions fluctuates year to year. It also might explain why Virginia numbers are so low compared to other states.

    Another question I have is what kind of positions are we talking about? Full time? Part time? Substitutes? Are we comparing apples with apples?

    One final thought. This reflects a longer-term trend and can’t explain the crazy fluctuations for individual localities, but are schools seeing a turnover of retiring older teachers with seniority at the top of the pay scale, who are replaced by younger, junior employees who are paid less?

    Perhaps VDOE’s methodology took all of these variables into account. But… perhaps not. Citizens need to dig deeper and make sure we’re getting honest information.

    1. James C. Sherlock Avatar
      James C. Sherlock

      It did not take any variables into account, or the averages would not have been so skewed. They took data that was wrong on its face and averaged it. As I wrote, I think they asked a question of the data that those data were never submitted to answer.

      That does not forgive the inaccurate inputs on the front end, but suggests the authors of the report were more interested in getting out of town than providing a professional product.

  3. Bubba1855 Avatar
    Bubba1855

    ‘average’ teachers pay is a statistical/mathematical joke…too many variables. I’m a math BS/MBA…various posts have pointed out the numerous gaps in the numbers…not to mention the input data…(Prince Edward County’s input was ‘zero’…fyi I graduated from a Prince Edward County college). Hey…if the GA wants to grant a one time increase to teachers I won’t argue with it…just don’t base it on some stupid/flowed statistical model.

    1. James C. Sherlock Avatar
      James C. Sherlock

      Exactly.

  4. f/k/a_tmtfairfax Avatar
    f/k/a_tmtfairfax

    Just the kind of mathematics & statistical analysis to be accepted wholeheartedly and without question by the MSM. In decades past, the media would jump all over this mess and hammer those who created it.

  5. Stephen Haner Avatar
    Stephen Haner

    I guess none of the people who ran the stats went to TJ in Fairfax….

    In my reporter days I tended to use in stories the starting salary and the top-of-the-scale for somebody with a master’s. In other words the range. That is what somebody considering the profession will likely look at. As noted before, I started watching this “teacher salaries above national average” kabuki dance when Chuck Robb ran for Governor.

    1. James C. Sherlock Avatar
      James C. Sherlock

      You’ve been around state government a long time.

      You know that data are tortured for answers that they cannot usefully provide to make points they cannot supported. This is one of those examples.

      But it is also a perfect example of data compiled for a “requirement” and then no one even glances at it, just the”averages”.

  6. Bubba1855 Avatar
    Bubba1855

    oops.. I meant to say ‘I DON’T DISAGREE WITH THAT’…

  7. Dick Hall-Sizemore Avatar
    Dick Hall-Sizemore

    I disagree that this is a worthless report. There are some anomalies at the district level that can probably be explained. And, yes, there is probably some bad data. But, most of the wide swings took place in smaller school divisions where the retirement of a few senior teachers and replacing them with new teachers who are paid much less can make a big difference in the system average expenditure. Those disparities likely had little effect on the overall state data.

    This data reflects the difficulties of using data from numerous jurisdictions that differ widely. A senior teacher retiring and being replaced by a new teacher or a teacher position going unfilled for six months can make a much bigger percentage difference in Wise County than it would in Fairfax County. In some jurisdictions, the differences may occur because a new school opened. There are a lot of variables at the local level that could explain fluctuations from year to year. For comparisons between jurisdictions, it would probably be best to use the one that Steve Haner suggested: starting teacher salary and highest salary for teacher with experience and advanced degree.

    At the state level, the best number to use is the one in the next to last column, FY 2022 Budgeted Average Teacher Salary. That is what each division expected to pay in teacher salaries, assuming all positions were filled and using the salaries in each approved or anticipated contract.
    Using those numbers, the “average” teacher in Virginia had a salary of
    $62,101 in school year 2021-2022.

    By the way, “classroom teachers” includes librarians and
    counselors.

    1. James C. Sherlock Avatar
      James C. Sherlock

      Dick, I agree with Steve.

      Another point, there is no indication in the report that the numbers of personnel in any classification (teachers, principals, etc.) in each division is accounted for in the state average salary. Averaging Franklin City with Fairfax County without that factor yields useless information if you wish to know what the average teacher in Virginia is paid.

      1. Dick Hall-Sizemore Avatar
        Dick Hall-Sizemore

        You are right. Or one could take the data, delete the outliers, say the two top and two bottom divisions, or perhaps Northern Virginia altogether, and then recalculate the average of the rest of the state.

        The data in the report can be used in many ways.

    2. f/k/a_tmtfairfax Avatar
      f/k/a_tmtfairfax

      Then, for smaller divisions wouldn’t it make sense to do some weighting, such as salaries per years of teaching experience? Wouldn’t that address the issues that Dick raises?

    3. DJRippert Avatar
      DJRippert

      Sounds like VDOE believes that the report is flawed:

      “Yesterday I received, unsolicited, a personal note from the highest levels of VDOE. In it my correspondent recognized the problem presented in my columns and promised VDOE management is taking on the task of fixing it.”

      1. James C. Sherlock Avatar
        James C. Sherlock

        They do. Along with other flawed reports such as personnel vacancies that I have been pointing out. They are taking on the whole process of reporting and assessing. First step is admitting the problem, and both VDOE and RPS have done that.

        1. James Wyatt Whitehead Avatar
          James Wyatt Whitehead

          That is a nice feather in the cap Captain! Double check their fix to the problem. VDOE believes in test retakes.

          1. James C. Sherlock Avatar
            James C. Sherlock

            I trust the new management at VDOE until proven otherwise. They have years of mismanagement to clean up. I am not referencing policy here, which offers another set of problems. What we are discussing here is just the blocking and tackling of governance.

            I am going to focus my attention on the Department of Health next. Similar issues of both policy and management.

    4. James C. Sherlock Avatar
      James C. Sherlock

      One suggestion, as Steve pointed out, ask each district for their pay scales for teachers, principals, AP’s, etc. Pick a common point or two on each scale, say a teacher with an certificate but no graduate degree with four years of service or whatever and compare them across the divisions. It will get much closer to the information sought than the current system.

      1. Dick Hall-Sizemore Avatar
        Dick Hall-Sizemore

        Your approach might be a better way of comparing salaries across districts. But, that is not what the General Assembly asked for. The language in the Appropriation Act specifically included average salaries for each division. DOE gave the GA what it asked for.

        Keep in mind one of the primary purposes, if not the most important purpose, of this data: comparing Virginia teacher salaries to those in the rest of the country. Because the only measure common with all states is the average, that is what is used. Of course, that has many shortcomings, among them the differences in cost of living. Thc cost of living in New York or California is certainly higher than it is in most of Virginia.

        By the way, the DOE report does have a chart that addresses one measure suggested by some commenters: starting salaries by school division.

      2. Bubba1855 Avatar
        Bubba1855

        i tried your ‘pay scale’ approach. easy to find FCPS’s scales…but for teachers there are 6 scales based upon the length of the contract…from 194 to 260 days. each has the pay for service time and advance degrees, etc. I went to Prince Edward County schools and could not find their pay scales. It might be there somewhere but I couldn’t find it.
        The deeper we dig the murkier it gets.

        1. James C. Sherlock Avatar
          James C. Sherlock

          I agree, but VDOE could create a standard format for pay scales, require it and replace the current salary reporting.

Leave a Reply