Blog Post < Previous | Next >
Guy
Measure A $517 million TUHSD Bond is way too much
Summary
On March 5th, we will vote on Measure A to approve a $517 million bond to shore up the Tamalpais Union High School District (TUHSD) infrastructure. This bond measure needs a 55% majority to pass.
This is an odd time to issue such a large bond when TUHSD's student enrollment is expected to contract rapidly from 5,070 in 2022 down to 3,920 by 2030, and ultimately 2,807 by 2054, at the end of the bond's term. For more info on this subject check the Enrollment Trend section.
Redwood and Tam account for about 74% of the overall cost of the projects. And, the funds are spent on mainly vanity architectural endeavors such as a luxurious cafeteria, a ceramics building, an auto shop, a common area, etc. These architectural showcases will do little to further students' academic and career prospects. For more info on those topics check the 'Cost of the projects' and 'The merit of the projects' sections.
TUHSD issuing so much debt ($517 million) will dramatically increase its financial leverage. And, it will similarly reduce its capacity to ever raise debt for true future urgent structural restorations. For more info check the 'Bond rating financial considerations' section.
As a community, we support education. But, we should not support the construction of architectural showcases at exorbitant costs that will have little impact on improving students' future. We should vote No on Measure A and force the TUHSD to come back with a revised structural improvement plan that is necessary and at a reasonable cost.
Content:
- Enrollment Trend
- Cost of the projects
- The merit of the projects
- Understanding the bond structure
- Bond rating financial considerations
- Americans with Disability Act (ADA) costs
Enrollment Trend
Before embarking on an analysis of this Bond measure, it is critical to understand the TUHSD student enrollment trend. TUHSD forecasts that student enrollment will decline by - 22.7% from 5,070 in 2022 to 3,920 by 2030.
I extended this student enrollment forecast out to 2054, the estimated maturity date of the $517 million bond. I did that by leveraging my Marin County long-term projection using the Marin County natural growth rate projected by the California Department of Finance. For more information on this model, you can check my earlier article Bay Area Long-Term Population Projection
When extending the enrollment projection, as specified, TUHSD enrollment decreases by - 44.6% from 5,070 in 2022 to 2,807 by 2054.
The above projection is benign. The table below indicates that enrollment is estimated to decrease by a slightly lesser amount over 24 years (-1,113) vs over the first eight years of the forecast (1,150).
Given Marin County's aging demographics, the above forecast may very well underestimate the student enrollment contraction over the next several decades. For more information on Marin County's aging demographics, you can check my earlier essay on the topic: Marin is as old as Italy if not Japan
Given the rapid prospective decline in enrollment, the $517 million Bond appears challenging to justify. Between 2024 and 2054, the average student enrollment is 3,430. The $517 million Bond corresponds to $151,000 per average enrolled student. This may be among the highest cost per average enrolled student of any k-12 school bond measure.
Cost of the projects
Project priority
Before evaluating the cost of the TUHSD projects, it is important to understand the meaning of the TUHSD priority semantics. They classify the projects in three different priority categories as defined below.
- Priority 1 projects represent 63.6% of total cost.
- Priority 2 projects represent 33.7% of total cost.
- Priority 3 projects represent 2.7% of total cost.
From a prioritization standpoint, there is nothing wrong. TUHSD focuses mainly on what they consider "Must-Dos" and "Should-Dos."
Cost of projects by school and priority level
TUHSD estimated that its total project costs rose from $393.6 million in 2022 to $517.1 million in 2023. However, TUHSD did not update the specifics for each school project. I did update such cost estimates within the table below by just increasing all earlier cost estimates by 31.4% so the sum of the project costs would increase from $393.6 million to $517.1 million.
As shown above, Redwood and Tam account for 73.9% of total costs excluding project contingency and other miscellaneous expenses. This is in line with their student enrollment which amounts to 72.7% of the total in 2022 and rises to 75.2% by 2030.
Some have argued that Archie Williams gets shortchanged. That is the case. Even though its enrollment is forecasted to decline more rapidly than the other schools, by 2030 it still accounts for 20.7% of overall enrollment; but, it receives only 10.8% of the adjusted funds (total costs excluding project contingency and other miscellaneous expenses).
The merit of the projects
As mentioned the project costs are concentrated on Redwood and Tam. So, let's focus on those two schools, and understand what we are actually paying for.
Redwood
The $194.5 million will allow the development of:
- A cafeteria
- A ceramics building
- A common area
That's a ton of money for an infrastructure that will do little to further student's college and career prospects. Who can make a living as a ceramist?
Tam
The $144.1 million will allow the development of:
- A music building
- An auto shop
- Classroom building for math, science, and photolab
The massive expenditure on 2 out of the 3 above items is questionable. Spending tens of $millions on an auto shop? Tam students are earmarked for college at the UCs and CSUs and private universities of similar caliber. They are not earmarked to become auto mechanics. A auto mechanic path should be housed at the College of Marin.
Spending tens of $millions on a new music building vs. sticking with the existing music facilities will not do much to enhance Tam students' college and career prospects.
Conclusion for this section
As reviewed, the majority of the Redwood and Tam projects are architectural vanity endeavors.
TUHSD needs to revise Measure A and come back to voters with projects that make sense given TUHSD's rapidly declining enrollment. And, such projects should enhance students' academic prospects instead of being vanity projects as if TUHSD was competing with Ivy League campuses with $billions in endowment funds; It is not, it is a public high school district.
Understanding the bond structure
The $517 million total bond issuance is subdivided into three series that will be issued as follows:
- Series A, $200 million, issued in 2024
- Series B, $158.5 million, issued in 2026
- Series C, $158.5 million, issued in 2028
The three TUHSD bond series will amortize as shown on the graph below.
When you stack up the three bond series, you get the following overall bond amortization profile.
You can streamline the above graph to focus on the overall amortization of the total $517 million bond issuance.
As shown above, the principal outstanding never exceeds $490 million even though the total bond issuance is $517 million. This is because Series A and B have several semi-annually scheduled principal payments before Series C is issued.
Bond rating financial considerations
The table below discloses a summary framework of how Moody's, one of the main bond rating agencies, rates school districts.
For background on the above, refer to the Moody's methodology.pdf attached at the end of this essay.
Economy. Weight 30%
Following Moody's parameters, the TUHSD would get an Aaa rating on both Resident Income and Full Value per capita. They are both high compared to most other communities. However, it would get a Ba rating on the Enrollment Trend reflecting the rapidly declining enrollment.
Financial performance. Weight 30%
Even after the bond issuance, TUHSD's Available Fund Balance Ratio and Net Cash Ratio should both exceed the 25% threshold to meet the Aaa standard.
Leverage. Weight 30%
The issuance of $517 million in long-term debt will have a profound impact on the TUHSD leverage, as specified by Moody's relevant financial metrics.
I first measured the impact of the bond on the Long-Term Liabilities Ratio.
Next, I translated the Long-Term Liabilities Ratio into corresponding Moody's ratings using their own ratio thresholds.
As shown above, by the time the TUHSD would issue the Series C in 2028, the mentioned ratio would have risen from 117% at the onset to nearly 450% by 2028. At such time, the corresponding rating for this ratio level would be Baa. Over time, as the TUHSD pays down the bond, this rating would progressively rise back to Aaa by the time the bond matures. But, the climb would take a long time.
Next, I looked at the Fixed Cost Ratio.
And, I translated this ratio into ratings over time using Moody's thresholds.
The Fixed Costs Ratio rises over time because the semi-annual bond principal repayments also rise over time. And, such principal repayments are included in the Fixed Costs. As shown, this ratio's corresponding bond rating declines over time (the higher this ratio, the lower the ensuing specific rating for this ratio).
Aggregating the above, I derive the following Moody's Scorecard rating estimates:
Given the above, I arrive at an overall rating of single A for the TUHSD. My calculations used a few reasonable assumptions. Moody's calculations will differ somewhat from mine. But, directionally I am confident that my estimates are directionally representative. This is because the dominant factor in those ratios calculations is the bond series cash flow. And, I had good information on such cash flow.
As reviewed, the TUHSD prospective demographic and financial profile has several weak spots including a rapidly contracting student enrollment and rising leverage due to issuing an enormous amount of debt (the $517 million bond). These weak spots account for 40% of the overall Moody's rating scorecard.
Based on my projection of the TUHSD financial condition over the life of the bond, especially the leverage, it will have little financial flexibility left to ever raise other material bond issuance for true urgent infrastructure improvements.
Moody's or other bond rating agencies may still give the TUHSD an Aaa rating regardless of the above figures. This is because:
- I don't know that the rating agencies will bother estimating the impact of the bond on the financial condition of the TUHSD over the entire 30-year term of the bond as I have.
- They may give a pass to the TUHSD on the declining enrollment, and not look beyond just a few years of decline.
- Rating agencies are notoriously lenient. Remember they are the ones who assigned Aaa ratings to mortgage-backed securities stuffed with subprime mortgages during the Housing Crisis. This leniency is due to an inherent conflict of interest whereby rating agencies are paid by the bond issuers. I don't think this conflict of interest has ever been fully resolved.
Keep in mind that the TUHSD could receive a rating lower than Aaa, yet the bond could still receive a higher rating of Aaa. This is because bonds that are repaid by earmarked parcel taxes are deemed safe independently of a school district's financial condition. In this regard, it is informative to look at the prospective debt service coverage (DSC) for this bond.
Bond Debt Service Coverage (DSC)
The DSC is equal to annual revenue from the parcel tax divided by bond principal + interest due in that year.
To come up with a reasonable DSC scenario over the 30-year term of the bond, We have to come up with:
- an interest rate assumption.
- a long-term yearly growth rate of the overall assessed value for the TUHSD territory.
Formulating the rate assumption can be straightforward. 30-Year single A municipal bonds are currently trading at 3.80%. Given the Federal Reserve ongoing tight monetary policy, such rates are relatively high. By the time, the TUHSD issues such bond series, rates are more likely to be lower rather than higher. Additionally, the TUHSD may well receive an Aaa rating on such bond instruments. That would further reduce the rate paid by about 0.40 percentage points. To give us a margin of safety, we will use a bond rate assumption of 4.00%.
Regarding the assessed value growth rate, let's first look at the history of such a growth rate going back to 1990.
As shown above, currently the TUHSD total assessed value is just above $60 billion. Below, see the yearly change in such assessed value.
The yearly growth in assessed value distribution over the 2014 - 2024 period is shown in the table below. The table shows the Min, Max, and several percentile levels in between. The median (50th percentile) is 5.5%.
To be reasonably conservative, let's use a yearly assessed value growth rate of 4.5% at the 15th percentile level. The resulting Debt Service Coverage over the entire term of the bond is shown below.
Regardless of the rate & growth assumptions used, the DSC ratio typically falls much below 1.0 in 2026 and 2028 because of the principal due on the Series A and B bonds. This is not a problem as the TUHSD should have an adequate reserve fund to cover this temporary gap. For bonds repaid by parcel tax, the rating agencies view any DSC greater than 1.10 as adequate. The TUHSD handily meets that threshold beyond 2028.
By the time, the TUHSD issues the bonds, if the rates and growth are much more favorable than assumed, the TUHSD can always lower the parcel tax. But, it has the discretion to raise the parcel tax back up to $30 per $100,000 of assessed value.
ADA costs
Bay Area school construction costs are huge
To begin with school construction costs in the Bay Area are huge. Todd Lee, President of Greystone West Company, an engineering firm working with TUHSD communicated to me that such costs run at about $1,000 to $1,200 per square foot. This fits with the recent Measure G Bond to upgrade Mill Valley Middle School which was associated with a $1,045 per square foot construction cost.
Outside the Bay Area, school construction costs are way lower. According to Save Our Schools the average cost to build a school in the US ranges from $200 to $500 per square foot. Another specialized source, Levelset estimates the average cost nationally at $359 per square foot. These figures are high as they are in line with current home construction costs in Marin. You can check that out for yourself at Bluehammer. One would think that school construction costs per square foot should be way lower than for homes, given that homes have a far more complex infrastructure (more bathrooms, counters, rooms, electricity & plumbing). So, why would school construction costs per square foot in the Bay Area be more than 3 times as much as the national average or home construction costs in Marin? But, you have not seen anything yet. Let's focus next on ADA costs.
Are ADA estimated costs rational?
Increased estimates in ADA costs are a material reason why the overall costs of the TUHSD projects rose by 31% between 2022 and 2023. In the most recent cost estimations, ADA costs come in at $42 million, or over 8 times more than the earlier estimate of $5 million!
Todd Lee from Greystone West indicated that the most recent ADA cost estimates are mandated by regulation to represent 20% of the hard costs of the projects. And, 20% of about $210 million in hard construction costs comes in at $42 million. Todd Lee provided me with supporting documentation of this 20% mandate (see attached at the end of this essay ADA Regulation pdf). This document describing the 20% mandate mentions exceptions for financial hardship such as projects becoming financially unfeasible. However, the language of the document is so obfuscating that it gives one no confidence that the 20% mandate could truly be avoided.
Next, I modeled what would be the price per square foot for ADA infrastructure associated with the overall $42 million ADA costs. I considered as inputs into the model:
- ADA toilet stalls (5 ft by 5 ft)
- ADA ramps (25 ft by 4 ft).
Given the specified inputs, if you build 50 ramps and 100 toilet stalls, the resulting ADA cost per square foot would be $5,600!
Next, I sensitized my inputs to consider scenarios ranging from 20 to 100 ramps and 40 to 200 stalls. The table below discloses the resulting ADA cost per square foot for all those scenarios.
The cost per square foot for different numbers of ramps and stalls ranges from $2,800 to $14,000! That leaves one speechless. I don't think it is feasible to spend that much money on ADA infrastructure.
Todd Lee/Greystone West indicated that my streamlined analysis foregoes many complications such as:
- paths from public transportation stops to campus
- Toilet rooms and locker room certification
- Lifts, elevators
- Ramps and associated retaining walls to accommodate grade changes
Well, I have the ramps and toilet rooms covered by even considering absurdly high unit quantities of both items. My excess quantities in both items could be substituted for constructing the other items mentioned by Todd Lee (elevators and paths).
In the end, I am still left speechless about the $42 million in ADA costs (vs. the earlier estimate of $5 million provided by PBK Architects, a firm well-versed in school construction). Todd Lee advances that PBK, even though it is a reputable firm, very much missed the boat on the 20% mandate. He is most probably right. Even though, I can well calculate that: $210 million x 20% = $42 million... that does not mean that ADA costs of $42 million make any sense. This may partly explain why Bay Area school construction costs are about 3 x the national average.
THE END