Wednesday, June 9, 2010

A Tale of Two Districts

The Columbus Dispatch published an article in its June 2, 2010 edition, reporting on the signing of new collective bargaining agreements between the Boards of Education and the teachers' unions of Dublin City Schools and Gahanna Jefferson City Schools. Those agreements are quite different, and I thought some investigation was merited to see if we could understand why such differences exist.

Update: on June 11, 2010, the Dispatch reported that the union representing the support staff of Dublin City School (the "Dublin Support Association") radified substantially the same deal as the teachers' union.

The new deal between the Gahanna Jefferson School Board and the Gahanna Jefferson Education Association (the formal name of the teachers' union) is – according to the Dispatch – a one year extension of the current agreement with no increase in base pay. In their contract for the prior three years, the G-J teachers received 3% base pay increases each year.

However, as Dispatch reporter Charlie Boss correctly points out, the G-J teachers will continue to receive their "step increase," which is an additional raise given for years of service – in those years for which a step increase is specified. In the case of the G-J teachers, step increases for a teacher with a Masters degree are paid in years 0-15, then 17, 19, 21 and 27. The size of the step starts at 5.25% for brand new teacher with a Bachelor's degree, and diminishes to less than 3% (of a larger salary) for teachers with higher degrees and more seniority. The G-J Treasurer indicates in his assumptions for their last published Five Year Forecast that overall, step increases would average 1.66%. That suggests that half or more of their teachers have many years of experience, and are in the sparse part of the step grid.

So by agreeing to freeze their base pay at 2010 levels, the G-J teachers will average a 1.66% increase for the 2010-2011 school year, with their most junior teachers receiving around 5% and the most senior teachers generally receiving no increase at all. Because this is a one year extension, the G-J School Board and the GJEA will be negotiating another contract next year. 

One year agreements are usually signed when one or both of the parties feel the deal shortchanged them, and want the opportunity to get back to the bargaining table sooner, hoping economic and political conditions will favor them more next time.

Contrast this to the contract just negotiated between the Dublin School Board and the Dublin Educator's Association. In their case, a new three-year agreement was negotiated, with base pay increases of 1% this year, 1.25% next year, and 1.5% in the final year. Their prior agreement was a two-year contract, with a 3% increase in the second year. As is the case with the G-J salary schedule, the step increase schedule for the Dublin teachers is non-linear, with the newest teachers receiving 5.2% increases and more senior teachers getting smaller percentage increases.

So why is there a one year deal with no base pay increase for the Gahanna-Jefferson teachers, and a three year deal with growing base pay increases for the Dublin teachers?

I have spent a good deal of time trying to correlate a number of school district statistics to teacher compensation, and have come to believe that the single greatest influence over teacher compensation from district to district is simply the wealth of the community. In other words, the higher the income of the people of the school district, the more they pay their teachers. Let's look at some comparative numbers:

The average classroom teacher salary in Gahanna-Jefferson was $67,493 in 2009, according to the CUPP Report published by the Ohio Department of Education (ODOE), while in Dublin, it was $66,885, a difference of less than 1% (for Hilliard, the average classroom teacher salary was $64,703). The statewide average classroom teacher salary was $55,581, and it ranged from $30,810 in Bettsville (Seneca County) to $78,624 at Orange City Schools (Cuyahoga County).

In those same school districts, the average community income is: Dublin = $90,715; G-J = $75,060; Hilliard = $66,602; Orange = $240,582 (!!); Bettsville = $38,445.

One might think that teacher experience plays a part in compensation, and that it should be relatively consistent across districts. Here is a chart that shows the fraction of teachers with 10+ years of experience plotted against the average classroom teacher salary, for each district. The slope of the red line is an indicator of how well these two statistics are correlated, and the near-horizontal nature of this line suggests that there isn't much correlation between these two statistics at all.

(note: the yellow triangles in these charts indicate values for Hilliard City Schools)

(click to enlarge)

How about compensation vs some measure of workload? One of the most used measures of workload is the student/teacher ratio. The notion is that the more kids there are in the classroom, the greater the burden on the teacher, and therefore, the more the teacher should be paid. As this chart shows, there does seem to be some positive correlation between these two statistics, in that the higher the average number of kids per teacher, the more the teachers get paid.

(click to enlarge)

Then there is the most controversial of all measures – student performance vs teacher compensation. Again, the CUPP data shows that there is a positive correlation between classroom teacher salaries and student performance, as measured by the Performance Index on the State Report Card.

(click to enlarge)

But we have to be very careful not to draw cause-and-effect conclusions from these correlations. In other words, it would be inappropriate to say that test scores would go up simply by paying teachers more. Or conversely, that when the Performance Index is low, it is because the teachers aren't paid enough. Note that this correlation shows that some of the highest paid teachers are in the lowest performing districts. There are a huge number of other factors that need to be examined. 

In all the pairs of statistics I studied, two produced strong correlations. The first is the Performance Index vs Students in Poverty, and the correlation is strongly negative. In other words, the more kids that are living in poverty in a school district, the lower the Performance Index. In fact, no district with more than 5% its kids living in poverty has a Performance Index of 100 or greater. By the way, while most of the high-poverty districts are the big urban systems, a number of them are in the rural Appalachian areas of the state.

(click to enlarge)

The other strong correlation is the average income of the people of a community versus the average income of the community's teachers. In this case, the correlation is definitely positive, meaning that as community income increases, the compensation of the teachers increases as well. In the case of Dublin, the average teacher salary of $66,805 is 74% of the average community income of $90,715, while for G-J, the average teacher salary of $67,494 is 90% of the average community income of $75,060 (in Hilliard, the teachers average $64,703, which is 97% of the average community income of $66,602).

(click to enlarge)

So might it be that the Dublin teachers got the apparent better deal of these two negotiations simply because of the greater income of Dublin residents? I think so. 

If we compare Hilliard Schools and our neighboring school district, Jonathan Alder, we note that JA is rated as "Excellent with Distinction," the same as us, with a Performance Index of 100.9 compared to our 101.5. However, their average classroom teacher pay is $49,646 vs our $64,703, making our average pay 30% higher than JA's. I believe it is significant that the average community income for JA is $55,277, which means the average income in Hilliard is also 20% higher.

So here we have two neighboring districts with virtually identical performance, yet the teachers have significantly different pay. I believe the reason for the difference is simply the relative wealth of the communities, and how that has – over many years of negotiations and levies – allowed the Hilliard teachers to bargain for better deals than have the Jonathan Alder teachers.

How are we to make sense of all this? Smart people all across our country have tried to discover an algorithm that can be used to nail down what it should cost to run our schools, which is substantially the same as asking how we should determine how many teachers, administrators, and staff a school district needs to employ, and how much those folks should be compensated. Governor Strickland claims to have the answer in his so-called Evidence Based Model, but it has never been fully funded, and therefore never tested, so we don't really know if it works or not.

Meanwhile local School Boards and Unions continue to struggle year after year with how to answer these fundamental questions.

I certainly don't know the answer to these questions either, and therefore suggest to you that whatever other rationale one might like to apply, the most truthful answer is that teachers and staff members get paid what they can negotiate. Their union leadership will work hard to drive a good bargain on behalf of their members, and the school board will do the same on behalf of the people of the community. 

As with any negotiation, the outcome will depend on the relative strength, empathy, resources, skill and resolve of the folks on each side of the bargaining table. The union negotiators are good at what they do, and know how to bring the power of their membership to bear on the process.

How are the people of the Hilliard School community going to participate in the upcoming union negotiations?

Speaking for myself, as only one School Board Member, I'd very much like to hear what you think should be the goals of this negotiation. The leadership of the teachers' union is asking the very same question of its members right now.


  1. If I remember correctly, the current contract granted the teachers around 3% for the final year of the contract, plus of course, the steps for those eligible. Their contribution for health benefits went up from 7 to 10% (correct me if those are wrong)
    Your average salary figure for 2009
    appears to be updated on the HCSD website to $68,058, or $3355 more. Given 1182 teachers, that would figure to a total increase in the budget of very close to $4M dollars, offset by the extra 3% in health benefits minus any increase in the total premiums (if applicable), for the last school year. That $4M works out to about 5%, so a lot of teachers must have been eligible for the step increase.
    My point in all of those figures is whether we as taxpayers can afford another $4M plus increase in the school budget for salaries, and how the Board and the union intend for it to be paid. This thought process was not used in the last negotiations - the contract was approved AFTER the voters had just turned down a levy request and the Board and union knew that programs were going to have to be cut, yet still agreed to a pretty healthy raise. I don't know where the negotiations are at this exact point, but I have not seen any request for input from the taxpayers, other than, of course, yours in this post. Once again, I find that irritating.
    I won't begrudge teachers a raise, and I don't find 1-2% at all out of line but what I hate is the blatant "veil" when it comes to the step raises - it is always glossed over, and in fact quite often never mentioned at all. What I also hate is the results of the poll which you wrote about last year where many thought that teachers deserved raises but those same people thought that the average salary was much lower than what it is. Both the average salary, and the total cost of any proposed increase, including step raises, should be made available on the schools web site, and any responsible reporters articles should include that as well (meaning that when questioned, the Board and unions MUST be up front with those facts.
    A lot of us (well, me anyway lol) have bitched about the lack of transparency in the missives from the Board and all I would ask is that they pay attention to the ones who are going to foot the bill. I don't expect the unions to follow suit - although I wish the same teachers who claim it is all about the kids would exert some influence on the union leadership to not come across as so arrogant as to ignore us, and to not put a totally false spin on the facts. I recall teachers claiming they were taking a pay cut when they had to pay that initial 5% of their health benefit premiums and some simple math shot that conclusion down. I would hope that we are not subject divisive misinformation this time around. I followed that article in the Dispatch rather closely the last few days, and the general take on the forum was that the people are still hurting economically, and people are becoming more aware of the increases in teacher compensation; it would behoove the union to consider that. If the teachers get what the public perceives to be too good of a deal, the next levy request is going to have a tough hill to climb, and once again, it will be the kids who lose.
    To be specific, what I would like to see is a one year contract with no raise, other than the steps which I don't think can be sidestepped. I don't think I'll get my wish but you asked. i do know that I would be dead-set against a contract which results in a 5% increase over each of the next 3 years.

  2. Hillirdite:

    Thanks as always for your comments. Some corrections to your assumptions:

    1. The base pay was increased 3% in all three years of the current contract. Therefore, their 2010 base pay is 9.27% higher than it was in 2007 (the last year of the prior contract) not counting step increases.

    For the 65% of the HEA members who got step increase all three years as well, their 2010 pay was 19.0% higher than the 2007 pay.

    2. The contribution to health insurance premiums increased from zero prior to 2008, to 6.0% (of the premium, not of salary) in 2008, 8.0% in 2009, and 10.0% in 2010.

    3. Using your example of a one year contract with no base pay increase, but with the current 4.15% step increase, I estimate that the new compensation level, including 30% for benefits, would increase by about $2.4 million, approximately equal to the amount of money raised by a permanent property tax increase of 1 mills.

    4. Your estimate of another 3% increasing the annual salary cost by $4m is right on. Adding the benefits takes it to $6.4m, which is equivalent to a permanent property tax increase of 3 mills.

    5. You are correct that the salary figure I used in this analysis is not current. Because the CUPP report is based on 2009 numbers, I used 2009 data for all statistics so the correlations would not be distorted. For 2010, the average base salary of the 1,002 fulltime HEA members is $70,613.

    Thanks again for your comments.

  3. Paul, thanks for clarifying my post. I am not always all that clear in my rants, lol, although I am trying to do less ranting and more rationalization these days.
    1.I realize it was 3% in each of the 3 years and should have made it clear that it was not just the final year.
    2. I also realize it was a percentage of actual premiums; since you mention it though, I AM curious as to what the exact total dollar cost is. As a businessman, I am not all that impressed by percentages - I am only concerned with actual dollars. You know,what it is going to cost, the actual amount of the check that will be written. I have never seen a dollar cost associated with either our share or the teachers share. Who gets family, who gets single, etc etc.
    3. I was not really sure how many of the teachers would be eligible for steps, and I guess I am still not; I thought it was a higher percentage. That goes to the point about transparency - I am sure the Board knows but I don't know where to find the data to discern that figure. I appreciate your estimate as it comes from someone who would know.
    4. Thanks for converting the math into actual millage, as that is what we all need to know - again, what is the actual cost, in dollars, for budget items.
    5. I had no intention of "calling you out" on the salary figure; I was merely using the numbers to point out the increase from one year to the next and the figure on the web site was fresh in my mind from some fact checking while perusing the Dispatch article regarding Dublin.
    Regarding that article, I was amazed (or maybe not) how even the defenders of the Dublin schools, had their facts wrong. Statistics such as per pupil spending were used to show how great the schools were, and, same as the poll in Hilliard which I referred to in my previous post, showed me that way too many people are just plain ignorant of the true picture.

    On a personal level, my business has enjoyed a pretty decent upsurge in business, but you know what? We have some debt to pay off as a result of the last 3 years of a very crappy economy as well as some infrastructure improvements which have been delayed - we are not quite ready to hand out 3 year guaranteed raises. As I said, all I would like to see is ONE year of patience and lets operate in the real world instead of some half-baked projections or some sense of entitlement that "all" white collar workers are getting 5% per year. It isn't happening
    and why the trend seems to be to exempt teachers is beyond me. Except that once again, the unions are powerful, and once again, I STILL question why highly educated professionals need, or rely on, a union to make their case.

  4. Hillirdite:

    #2 - I don't know the exact numbers either, but the last thing negotiated (i.e. the last point of contention) in the contract was a cap of $50.19/mo for single coverage, and $135.52/mo for family coverage.

    #3 - The 65% number comes from the actual list of HEA members which I asked for and received from the Personnel Dept. The raw numbers are that in FY10, 710 of 1,089 HEA members got a step raise (65.22%).

    #5 - Didn't think you were. You made some very astute observations using scanty info. I just wanted to clarify why the numbers I used are different than yours.

    Thanks again.

  5. One other point I forgot to make in regard to the teacher's salaries I reported in this article: the CUPP report labels these numbers as "Classroom Teachers Average Salary." There are a number of folks with other job titles who are also members of the HEA, and who, as a group, have a higher average salary than classroom teachers.

  6. Interesting graphics. It seems to me that Hilliard is one of the more economically diverse places around. Assuming that the wealthy tend to vote in higher numbers than the lower middle class and poor, that may skew Hilliard towards being overly generous with teachers. 97% is awfully high.

  7. I don't think that Hilliard voters have been overly generous with teachers - we don't really get a say in that. We vote No on a levy, the Board/Union agrees to a particular contract, and then announces cuts in programs/services because the money is not there for those services. The levy appears on the ballot again, and this time the voters say yes. The Board/Union assumes the voters "accepted" the contract by voting Yes on the levy, when in actuality, the voters are once again apathetic to the contracts but fully in tune with the cuts to the programs/services. THIS is the cycle that must be broken - voters must be aware of the big picture that is the HCSD budget, and must let the Board know that if they want our support at the polls, they can't focus on threatening us with cuts to programs while ignoring the 800 pound gorilla which is salaries.
    Again, suck it up for a year, HEA. Accept your steps, for those who get them; accept nothing or a token for those who don't; come back to the table in a year and let's see how things are going for those of us who have to foot the bill. Save the money "for the kids".

  8. Is it possible that the average teacher salary is in any way impacted by the average age of that pool of teachers? I believe Cleveland used its stimulus money to pay more experienced teachers to retire so that younger (cheaper) teachers could be hired. It could be irrelevant, but I can't help but wonder if that number is skewed somehow. (Special thanks to "The Flaw of Averages," an excellent read.)

  9. Certainly each school district has its individual profile of teacher experience, and that will influence the average teacher salary statistic for that district. But as I try to show in the chart correlating the fraction of teachers with 10+ years of experience vs salary, there doesn't seem to be much connection between the two.

    In other words, if experience were a significant influence over compensation, one would expect the districts with the greater population of more experienced teachers to be paying their teachers more. But this scattergram does not show that to be the case.

    A good friend of mine retired a few years ago as a teacher in Meigs County, one of the poorest areas of the State. She had more than 30 years of experience, and a Masters degree. Nonetheless, her final salary was in the low $40s, about the same as a Hilliard teacher with a Bachelor's degree and two years of experience.

    Is a Hilliard teacher that much better than a Meigs County teacher? If experience is so important, note that 69% of Meigs teachers have 10 or more years, compared with 65% in Hilliard.

    Or might it be that 1 mill of property tax raises $67/student in Meigs county, while 1 mill raises $161 in Hilliard. Or that the average income of Meigs county residents is $35K vs $67K in Hilliard?

    You're correct that doing analysis strictly on averages can lead one to flawed conclusions. The sets {0,0,100,100} and {0,25,50,75,100} have the same average (50), but are clearly quite different in nature.

    Nonetheless, this kind of analysis can help give directional guidance for further study. My takeaway from this analysis is that community wealth seems to have the greatest influence on teacher salary.

    In other words, I believe the primary reason teachers get paid more in wealthy districts is that wealthy districts have more to spend on their teachers, and are less likely to balk at union demands for higher pay.

    Which leads me back to the hypothesis that teachers get paid what their local unions can negotiate on their behalf.

    My motivation is not to pay teachers less, but to get our community to wake up to this dynamic, and play their role in the negotiation process.

    Many reasons are given as to why school districts keep coming back to their voters for more levies. The simple answer is because school districts spend more every year and most of that increased spending is to fund step increases and base pay increases.

    Yes, the State of Ohio is struggling to meet its school funding obligations, but State funding is unlikely to ever increase as aggressively as has our teacher compensation.

    This is a local question: How much do we want to pay our teachers? The answer to that question will determine the size and frequency of future levies.

    It's that simple.

  10. I've added a note in the main article stating that, according to the Dispatch the union representing Dublin's support staff has agreed to substantially the same terms as the teachers' union....

  11. Hillirdite:

    The Treasurer has made this information available relative to health insurance costs:

    Family plan
    Total Cost - $1,221.07 per month
    employee cost 122.10 per month
    employer cost 1,098.97 per month

    Single plan
    Total Cost - $452.25 per month
    employee cost 45.22 per month
    employer cost 407.03 per month

  12. It certainly seems like a salary level at 97% of average is pretty generous. I understand that just looking at averages glosses over some interesting data points, but that number seems to stick out a bit.

    I'd also be interested in learning whether any of these numbers would change if we were looking at fully-loaded compensation values (ie, salary plus benefits). I know that the benefits I've seen from my employer have decreased substantially in value over the last few years, and it doesn't seem fair to completely ignore the time off that teachers have for breaks, etc., that most of the rest of us would love to have. Although I suspect that data would be much harder to come by, I think it would be interesting to look at a total compensation number that includes the value of all benefits and paid time off.

    I certainly wouldn't expect to underpay teachers, because the work they do is incredibly important to the future of our community, but let's also not forget that lots of teachers get into that field because it's got some unique "lifestyle" benefits for families when you look at hours and working days. I'm not sure you'd find a lot of 60K+ jobs in Hilliard where you're able to be home when your kids are off from school over breaks and summer vacation, right?

    When we're doing apples-to-apples comparisons, let's make sure we're looking at all the apples.