Sunday, April 1, 2012

Funding or Spending?

The Sunday, March 25, 2012 edition of The Columbus Dispatch included a story about the seemingly never ending debate about how to fund Ohio's Schools.

Two primary factors drive the funding needs in school districts - growth in student population and growth in the cost of compensation and benefits. Most school districts in Ohio, including ours, have not had much growth in student population for the past several years, so it has mainly been the compensation and benefits costs which have driven our spending growth rate.

To put it bluntly, this dialog about school funding is really about how much more money the people and businesses of Ohio should pay in taxes so that more money can be spent on the teachers, staff and administrators employed by our school districts.

That's not an ultra conservative, extreme-right, gonzo Limbaugh-loving radical viewpoint. It's just the raw truth.

Most of the professionals employed in our school districts are classroom teachers - and that's as it should be. The amount that is spent on compensation and benefits for those teachers is driven by two primary factors: a) the ratio of kids per teacher; and, b) the average compensation per teacher.

There is no 'right' number for either of these two factors.

It is generally believed that the lower the student/teacher ratio, the better it is for the kids. It seems intuitive that - all other things being equal - the fewer kids a teacher has to work with, the better. But is there one "right number" that everyone agrees we should shoot for?

We know it can't be 1/1. And we know it can't be 100/1. The so-called Evidence Based Model put into place by the Strickland administration set the Regular Education Teacher ratio to 19/1 for grades K-3 (line 6A) and 25/1 for grades 4-12 (lines 6B-6D). It calls for Hilliard Schools to have 674 teachers.

According to the State Board of Education's Cupp Report, the overall student/teacher in Ohio's public schools is 17.2/1, and it ranges from 10.8/1 in Fayette Local Schools (Fulton County), to 25.3/1 in Lorain City Schools (Lorain County). The chart below shows what percentage of districts have various average Pupil/Teacher ratios. Most districts have 18 to 20 kids per regular education teacher. Notice that of the districts ranked Excellent with Distinction, the percentages at higher student ratios were greater.

click to enlarge
Our ratio here in Hilliard City Schools is 19/1, with a total classroom teacher team of 687 - about the same number calculated by the Evidence Based Model.

I won't try to sell what I'm saying as a high-quality analysis, but it suggests that increasing our student/teacher ratio by one or two kids won't necessarily hurt performance. This could reduce the number of teachers required by about 40, which would reduce our cost for salaries and benefits by about $3.5 million/year. I think it merits discussion.

As for teacher compensation, the Strickland EBM also set guidelines. It said that statewide the compensation plus benefits cost for a teacher should average $57,812 (line 5C). Of course, it was never funded to that level, and it was never clear where the money would come from to do so.

The data strongly suggests that teacher compensation is tied to the wealth of a community. Statewide, the average pay for a classroom teacher (without including benefits) is $56,850, but it ranges from $79,656 in Orange City Schools (Cuyahoga County) to $32,020 in Southern Local Schools (Meigs County). 

This chart shows the relationship between the average community income for all 608 school districts in Ohio versus the average teacher salary. The yellow point is Hilliard City Schools, with an average community income of $66,258 vs the average teacher compensation of $69,369:

Click to enlarge
The red line is an indicator of the relationship between these two variables, and it suggests that the correlation is strong - as community income goes up, so does the compensation of the teachers. Teacher compensation is set via the adversarial collective bargaining process, with the teachers' union on one side of the table, and the school board on the other (actually, both parties are represented by professional negotiators).

So what this chart really tells us is that the higher the incomes of the people of a community, the more likely they are to support the local taxation necessary to pay higher salaries to their teachers. That's not an unexpected or unreasonable outcome, but we rarely if ever talk about teacher compensation in these terms.

The real challenge is relating the rate of change of local community compensation to the rate of change of teacher compensation. In the booming economy we enjoyed for a couple of decades, it came to seem normal that everyone's incomes in the private sector were going up (and house prices were marching upward, along with the balances of our 401k retirement accounts). Folks working in the private sector were getting nice raises, bonuses were pretty common, 401k matching was generous, and some even got a nice boost from employee stock option plans.

During that same period, the teachers were able to introduce the 'step and lane' salary grid approach into their collective bargaining agreements which, along with decent base pay increases, rapidly accelerated their compensation to catch up with the growth of the private sector.

But what happens when the economy slides in a deep recession, where unemployment climbs to very high levels, and those who are lucky enough to keep their jobs must figure how to survive with little or no raises, increasing workloads, and sometimes even salary cuts?  Does the teacher compensation philosophy used during the boom years work in a prolonged recession, when both income and net worth has taken a deep hit in the private sector?

In their last contracts, our teachers agreed to freeze their base pay through 2013, delay two step increases, and eliminate one altogether. The support staff union agreed to similar terms in their contract. Additionally, they have stepped up their share of the health insurance premium to 15%. Remember this contribution had been zero until 2008, and rose to 10% in the 2008-2010 contract.

This was key to allowing our last levy to be 5.9 mills in spite of substantial state funding cuts, and will allow us to not ask for more funding before 2014 - barring any more surprises from the state government. We can debate how much the teachers' union was motivated by the threat of SB5 vs their empathy with the economic pain being felt in the private sector, but regardless, the result was a deal which is more appropriate to our community's current situation.

Governor Kasich and the General Assembly may figure out how to allocate more state budget dollars to K-12 public education. Maybe the economy will pick up, and state tax revenues will start growing again. Maybe, but I suspect that the rate of state funding growth will be very gradual for a while.

Maybe cost cutting in other government programs will free up money for public education. Maybe, but I doubt that it will be significant. The politicians talk big about cost cutting, but in practice, it's very hard to eliminate or reduce programs.

If the Governor and the General Assembly do find more money for K-12 education, I wouldn't count on any of it coming our way. New dollars will first go to urban and rural districts, like Meigs County were the average compensation for teachers is less than the starting salary for ours (a good friend of mine retired a few years ago from Meigs County schools, having taught for 30+ years, and her final pay was less than a 3rd year teacher makes in our district).

As I've said many times - this is a local matter. We shouldn't expect any help from the statehouse, although we shouldn't stop making our voice heard. We have to figure out how to set realistic expectations for both the teachers and the community members.

That will take education, dialog and a healthy dose of respect, trust and empathy on both sides of the bargaining table.


  1. I see the phrase "Excellent with Distinction" in this article.

    Since that has now been replaced by an A, B, C, etc. grade system (of which we merit a B), maybe you can update the chart to reflect that.

    Seems we have been sold a bag of goods for our "Excellent with Distinction" rating which turns out to not be as good as it appeared.

    Maybe it's now time to be honest with the taxpayers?

  2. Bag of goods M? How so? That was based on the past evaluation that the state had in place. A new model is now place. How do you know which is good, bad or indifferent? Maybe some research would help instead of making a bag of goods?

  3. Paul, I'm curious how that student teacher ratio is calculated and where the classes are in Hilliard that have 19 kids or less. My son's elementary class has 26 students. I'd hate to see his class size rise any more than that.

  4. Here is the most recent Enrollment Report. We can pretty easily see the kids/classroom ratio for the elementary levels. Ignoring kindergarten, the size of the grades 1-5 classrooms range from 19 to 30, and average 23.

    It's when you get to the high school level that the section sizes can get pretty small, pulling down the overall average. It's a consequence of the large number of courses we offer (around 300).

    I know my kids were in classes in high school with only 4-5 kids, such as their advanced foreign language classes. Selfishly, I got benefit from those kinds of classes because they helped our kids get college credit, which could have the effect of lowering total college tuition costs. But is it fair to ask the whole community to share the cost burden for the benefit of a few?

    The question isn't that simple though. A Latin 5 class can only be taught by a teacher certified to teach Latin. So if the schedule works out that the one Latin teacher in a high school has a period available to teach Latin 5, the incremental cost to the district is zero.

    By the way, we don't offer Latin in our district.

    High school scheduling is a complex beast. Required course have lots of students enrolled, and the section sizes tend to be in the 25 student range. Elective courses have fewer students in general, but it can range from a handful, to say 100 in a band or choir class. Teachers often have licensure in narrow subject areas, and can't be cross-assigned to other subjects to try to even out the workload.

    Unwinding some of the variety from our high school course catalog would not be simple, or without pain. But if the planning is done with a long horizon - the pain can be minimized (e.g. we'll stop offering X no later than 2020, depending on things like when the incumbent teachers of that subject intend to retire).

    The Innovative Learning Center may play into this strategy as well. Perhaps some classes are no longer offered in the high school buildings, but a section or two is offered at the ILC, and made available to students across the district.

  5. @Anonymous

    Maybe you've not been paying attention, but we've been hearing about "maintaining our 'Excellent with Distinction' rating" in every levy fight, and every discussion about HCSD financing, since it was first achieved.

    Since we only now merit a 'B', it would indicate that the way the previous rating was applied and/or achieved was not by true academic standards, which the new rating system appears to put more emphasis on.

    My point was simply that the district isn't as good as people were being told it was.

    The flip side of course is that our teachers and support staff now have something to strive for.

    For some strange reason, I suspect it will involve more money from the taxpayers...

  6. We need to recognize that the state report cards are a political instrument. Some want to use them as a hammer to drive reform into school districts, which can be done by negotiating 'standards' in exchange for funding or curriculum control.

    In other words, the folks downtown might have put this new rubric on the table, knowing it would cause an apparent downgrade of many districts, so as to use it as a bargaining chip to get school districts to acquiesce to new curriculum requirements (e.g. "if you add the study of Ancient Egyptian Engineering to the curriculum, we'll ease up on some the the report card standards, restoring you to an "A" district").

    The local districts are inclined to give into such pressure because they need the "A" rankings to get levies passed.

    This is just a made-up example, but the point is that most political discussions are designed to appear one way to the public, when there's other stuff going on behind closed doors.

  7. Paul, but the elephant not yet chosen to be revealed by this district is EXACTLY WHAT IS OUR REMEDIATION RATE. Last discussion some time ago was over 20% and so I would suggest it may have not gotten any better. At any rate, WHY NOT TELL US, the taxpayer, parent, student etc. Politics or not
    if anyone thinks 20% plus remediation is Excellent with distinction then one needs a drug screen or deliberatly

  8. Rick: The Ohio Board of Regents has published remediation data for several years. The most recent reports are available here on their website.

    I would read the 2009 report to say that of the 485 graduates of Hilliard high schools in 2009 who entered an Ohio public college, about 40% had to take either remedial Math or English, and about 9% had to take both.

    Although the report tells us that about 38% of these 485 students enter community college, it doesn't give us the breakdown of remediation of kids entering Ohio universities vs those who start out at a community college. One would assume that remediation is higher for the community college kids just because entrance requirements tend to be higher at the universities.

    The state report cards have 26 indicators for performance, ten of which are associated with the Ohio Graduation Test. To get the checkmark for each of these ten indicators, a school district must have at least 75% of its students score at least "proficient" on the OGT subject area.

    According to our 2011 Report Card, we have well more than 95% who score at least "proficient."

    But I think most people see the 95%, and somehow take that to mean our kids are all getting an "A" grade on the OGT. What it really means is that 95% got a "D" or better on the test. That's a different thing altogether.

    We need to understand that the State Report is a political document. Not to diminish all the great work done by our students, teachers and staff, but the structure of our current report cards can lead casual readers to feel overly optimistic about the level of performance of the students.

    The new report card system is supposed to be less so, and in simulations with the current test data, the new system would rank us at a "B" performance level. In fact, no Franklin County public school district would get an "A" under the new system.

    And that means there is political pressure to change the rating methodology.

    We'll see how that works out in the Ohio Dept of Education and the General Assembly.

  9. I would like to throw in my two cents here, as the engineer spouse of a teacher in another Ohio district (just to be clear of my own biases):

    First off, the fit on your curve of teacher income vs. community income is really extremely poor. There is a point to be made with a similar analysis, but a linear curve fit is likely just as accurate as the parabolic that you chose, though it wouldn't support your conclusions as much. I'd recommend looking at fitting average net compensation per FTE as one change that might reduce noise in the plot. Why? One local district here reduced part-time employment significantly, which increased average salaries (1 teacher making $50k instead of 2 making $25k), but actually reduced district total cost through lower benefits costs.

    But second, I truly believe that comparisons to community income are just a poor way to go. Ohio until recently required a masters degree for teachers to renew their licenses. As a result, you have pretty much 100% of all teachers have a bachelors degree and in most districts 60-80% of teachers have a masters degree. By comparison, fewer than 25% of all Ohio residents over 25 have a bachelors' degree. Some districts that is higher, and others it is lower. Not surprisingly, household income changes across the board with education level. So comparing a workforce with 100% having college degrees to those with 25% or less having degrees isn't exactly an even comparison.

    That doesn't negate your point that higher resident incomes tend to translate to higher teacher salaries, but it builds to this point:

    The economy has been brutal to some. However, those with college degrees have largely been spared. According to the Bureau of Labor Statistics, unemployment has remained low and real wages have actually grown for this group throughout the recession and recovery. Real wages for the population as a whole have not. If you want to tie teacher wages to average incomes of the population as a whole, you will effectively move to a point where becoming a teacher has a high entry cost (college education) with extremely low returns relative to other professions which could be chosen. The returns already aren't great compared to alternatives, but making them worse certainly won't do anything to attract qualified, talented employees. Should you want to tie wages to growth in salaries of college-educated professionals, you could avoid this. However, growth in teacher wages in Ohio since 1986 have already trailed growth in salaries for college educated professionals quite significantly.

  10. Thanks for your comments.

    The curve fit I used was nothing more than the standard "exponential trendline" available in Microsoft Excel, which in this case is a 3rd degree polynomial, rather than a parabolic function.

    I'm not sure how you can support saying a linear fit would be more accurate. Visually this "exponential trendline" looks pretty good. But here's a link to the same chart with a linear trendline added. The slope is still pretty positive, indicating a positive correlation.

    Let me put the question back to you - if the wide variances in pay scales between Ohio's public school districts can't be explained by their correlation to community income (or wealth if you prefer), then what do you propose is the cause of these pay scale differences?

    I see compensation in general as being a matter of negotiation. That is, there are no absolutes which relate to degrees, experience, or anything other than a meeting of the minds between the employer and the employee. We can't claim that lawyers are more valuable to society than diesel mechanics, and therefore should be paid more, even though a law degree is very expensive and takes at least seven years of study. And I'm quite sure than the best diesel mechanics are paid more than the worst lawyers.

    In the case of unionized employees, the bargaining process is between the employer and the union as a whole, rather than with individuals. Again, I have no problem with that, as long as both parties play their role ethically.

    Employers in general want to pay less, and employees in general want to be paid more. But employers know that they have to pay enough to attract candidates who will be an asset to the organization. If an employer sets the salary for a particular role to $X, and no satisfactory candidates apply, then perhaps the salary needs to be raised.

    But I don't think we can say that this is the case at the moment, at least not in Hilliard. We had a job fair a few weeks ago, and had approximately 700 applicants show up. Sounds to me like there is no problem attracting lots and lots of teachers to our district at our current pay levels.

  11. Paul - Same anon. poster as you replied to here....

    You make some very good points.

    When it comes to the curve fit, I do believe a better fit, visually, at least, would be a piecewise linear fit with a break around $55-60k on the x-axis. That would still show an increasing trend, but would likely result in a higher r-squared value and wouldn't exaggerate the higher values as much - a third degree polynomial implies extraordinarily high growth if extrapolated, growth that isn't supported by the data.

    When I look at pay variance among districts, there is often little apparent explanation at first glance. Some of the districts that would provide more challenging work environments that you would expect to have to provide higher compensation don't. Some do. Wealthier districts that could pay more to attract top talent (and be pickier in their hiring) sometimes do, sometimes don't. The one factor I can think of off-hand that would be interesting to consider, in addition to average household income, is the # of children in the district per household. This number varies wildly. I'm not familiar with Hilliard, but as an interesting comparison, consider these two examples. My wife works in a district that has seen rapid growth and has a very large number of children per household. Their average household income is fairly high (on the upper end of middle class). The district has below average spending per pupil and the teachers are paid good wages, but well below surrounding districts. We live in one of those surrounding districts. Over 80% of the households in our district do not have children in the school system, and average household income is considerably lower. We spend over 50% more per pupil than the district my wife works in, and teachers generally make 20% more for the same qualifications and experience. We do get more programs for the extra cost, though. Interestingly, the state provides her district over 5x as much per pupil in state funding as they do our home district. Despite all these differences, our effective millage in our district is about half of what it is in the district for which she works.

    Unfortunately, the state supreme court has ruled that impact fees, the one way in which they could moderate the impact of residential growth, are illegal.

    I completely agree about the balance in compensation - and I believe we should let districts decide what they want to support, without artificial limits from the state. If they want to pay more for better talent, that should be their prerogative. I would like to see a base level that is reasonable for all districts, too, so that we don't end up shortchanging poor areas by simply letting them push wages to the point where they get people willing to take the jobs, but people who aren't competent.

    One final point w/ respect to Hilliard's job fair. I took a look, and it did appear that you might have had a fair number of job openings, though it wasn't possible to determine how many from their webpage. Naturally, we should look at the number of applicants per opening as a bit better indicator of the level of supply. Some of those positions may have had high supply and some low (for the record, I am opposed to the treatment of all teacher roles as identical when it comes to compensation - I think unions have it wrong there). In addition to considering that, we should also consider that no candidate in their right mind would ever apply to a single position, and there is no guarantee that Hilliard could even hire the people they wanted even after their fair. A local district here pays a little more than Hilliard for similar degrees/experience, and I know that they've actually had a hard time finding candidates willing to take their jobs, as the people they've been offering have been getting other offers.

    Sorry if this is a little disconnected - I'm operating on very little sleep.

    1. I remember working with one of our math experts at CompuServe to try to build an extrapolation function for a large set of data we had (modem usage by time of day). Fortunately, we had some pretty cool software to help with this stuff. Turns out that the equation that produced the best fit to the observed data would generate a looping curve if you extrapolated very far past the upper bound of the observed data. We ultimately simplified it down to a two-term, two-variable equation, and worked well enough to tell us how to scale up our global network without overspending - potentially by tens of millions of dollars.

      Interesting thought re kids per household. The district you're talking about sounds like Olentangy. If so, their situation is that they experienced their growth boom in the past 10 years, so most of their teachers are pretty young, and therefore on the low end of the pay scale. Here in Hilliard, our growth boom started about 20 years earlier, so the majority of our teachers are toward the upper end of the pay range, which is the an early retirement incentive program made economic sense for us.

      Olentangy will see its cost/student rise to catch up with the surrounding district as its teachers age. It could take decades for a district's teachers to become 'normally distributed' across the experience scale, and I suspect that thing like early retirement incentive programs just delay that normalization (ie - our district will again have a batch of young teachers as the retirees are replaced).

      The Ohio Revised Code already specifies the minimum pay for teachers across the state, but it's a joke: the top end of this pay scale in the law is less than the starting salary in most school districts around here. I know of only one school district who pays anywhere close to the state minimum (Bettsville), and they frankly should be shut down and merged into one of the healthier surrounding districts.

      But where would the money come from to place those Bettsville teachers on the higher paying wage scale of a healthier district?

      That gets to the ugly truth about our so-called public school systems. They're really private school systems operated for the benefit of the families who can afford to live in the community. We pay into a pool administered by the State that redistributes that money so that the poorest school districts get just enough to prevent causing sufficient civil outrage such that the invisible community walls are brought down.

    2. Within the community, there is a constant struggle between the teachers and the taxpayers to keep pay levels high enough to avoid teacher job actions which have a negative impact on property values, and to keep taxes reasonable - also to prevent an erosion of property values.

      I don't disagree with comments regarding competition between school districts. There are some very talented teachers getting laid off these days just because of the way the union contracts work. For example, the administration can decide to eliminate all teachers in certain roles which can only be filled with specific licensures. The easiest example of this is foreign language instruction. If a district eliminates say French from its course catalog, all the French teachers are going to get laid off, as there is slim chance that a French teacher has licensure in another subject area. This could result in a very talented French teacher being put on the street and subsequently recruited by several school districts.

      But the more likely situation right now is that we have tons of teachers getting laid off at the same time tons of new education graduates are hitting the market, and it's tough for them to find jobs. So 700 of them show up at our job fair for maybe 50 jobs. Most of those 700 would show up also for a job fair if your district put one on.

      But there are plenty of unemployed teachers who aren't so mobile. My kid is one of them. Not only is her field relatively narrow (music ed), but she's not relocatable as her husband has a great job where they are. So she applies for every opening within a reasonable driving range, but they're few and far between. While there were folks at our job fair from the four corners of Ohio, I bet 90% were from the Columbus metro area.

  12. (ctd) On a side note, I do want to discuss the value of the STRS plan at some point. I believe you're making a false conclusion in many of your posts as to the value of the pension. I see the following as serious problems:

    * The value of the pension to young and future employees (BIG problem)
    * The gross disparity in value to teachers of different ages
    * The fairness of the pension to those teachers who enter it honestly
    * The method in which benefits are calculated
    * The false assumption that many make that the district's contribution is really being given to the current teachers pension "accounts".
    * The unrealistic rate of return STRS is hoping to get on their assets
    * What will likely happen when the plan doesn't achieve that rate of return

    Net, this is a key, we are really NOT giving the younger teachers a good plan. *IF* they play the system honestly and work a full career, it won't be any better than social security and a 401(k) with an average company, assuming that the pension plan actually regains solvency. Most likely, it will be worse than you get in the private sector.

    HOWEVER, I honestly believe that there are ways to improve the system. You should be able to actually provide a better pension to employees, based on an EASILY obtainable rate of return at a significantly LOWER contribution rate from the district, if you simply reform the plan properly. I'm dismayed that the state chose NOT to reform things well, simply because I believe it is fairly likely that many of those teachers working today and in the future will end up getting a very poor deal as a result. More later, though. :)

    1. Not sure what false conclusion you're accusing me of making. I think I'm pretty much in agreement with each of the points you listed, even if I haven't spoken explicitly about them.

      In particular, the point about how different the system is for teachers of different ages. Teacher pay accelerated rapidly during the latter half of the careers of the teachers retiring now. Added to that was the kicker for those retiring with 35 years of service, which will certainly be eliminated in the new law.

      The compound effect of those two things puts teachers right now in what is probably the all-time sweet spot for STRS benefits. I told a retiring teacher friend of mine that her benefits had a present value on the order of $1 million - that it was just like STRS handed her a check for $1 million on the day she retired, and told her to live the rest of her life on that money.

      And her lifetime contributions will pay for only a fraction of that. Lots of it will come from the contributions made by the teachers still working, who will never see retirement benefits anywhere so generous.

      This is another one of those gut-wrenching, soul-searching conversations which need to be had: who's going to take the haircut to keep STRS solvent? Teachers already retired? Teachers just starting their career? Teachers about to retire?

    2. Paul - my comment about false conclusions was meant not to imply that you aren't correct about the value of the benefit to older teachers - but simply to point out that when we all discuss this, we *MUST* (in my opinion, at least) discern the difference between the value to current, new, and old teachers. Most people don't have a clue about this, and when the pension is labeled as "generous" in general, it grossly mistreats the younger workers who, as you point out, will NEVER get ANYWHERE near as generous of benefits, even if the pension plan does return 8% as they hope. I believe your posts on STRS can often lead the reader to false conclusions on the value of benefits, simply because they often are written in such a way that all the teachers are treated as a homogeneous mass, which they aren't. If I misinterpreted your meaning, I apologize.... but I do want other readers to understand this difference.

      Personally, I just want a fair system for all... and I think we can do it, if we simply throw out old models of doing business. The current one is terrible, though. Another example of its issues: I got a pension buyout notice from a former employer. It included links for a website to calculate the lump sum required under federal law to be able to (combined with rates of return equal to a composite corporate bond rate) afford monthly income equal to the pension upon reaching eligibility under that pension plan.

      I turned around and used the same webpage to calculate the lump sum required under that federal law to provide a benefit equal to what my wife has earned through STRS over 10 years of service. The lump sum was about 25% BELOW what she has paid into the system so far. Not employee+employer contributions, but employee (10%) only. That's seriously screwed up, IMO, and largely the result of the points I mentioned above. Their proposed solution only exacerbates some of these problems.

    3. They are certainly not a homogenous group -I absolutely agree with you. That's the reason I point out that the teachers who have retired in the past decade have fallen into a 'sweet spot' which is unlikely to be repeated, and it can blind one to the situation of the teachers who came before them, and those who will come after. That's your point, and I agree with it.

      To apply another overused analogy - there has been a perfect storm. The teachers who retired a decade or so before did so at much lower final average salaries. Many retired before the accelerator was applied in 1999 that granted retirees with 35 years of service a pension equal to 88% of their final average salary. These older retirees - many who are 70 years or older now, will collect perhaps half the pension of a teacher who retires now, before the 35-year accelerator is eliminated.

      Nor can the system support the pension promises still being made to teachers who are many years from retirement. The assumption of investment return is still overly optimistic, in my opinion, and will not generate the income necessary to support even the new benefit plan being considered by the General Assembly.

      Regardless of the finer points of this debate, the question remains: who takes the haircut?