We’re conscious of making the climate risk data and analyses we deliver correspond to both historical damage and rational climate change models. Just as critically, we’re aware that relating climate risk to more familiar metrics such as property value, population and economic activity are key to incorporating it into assessments of municipal bond issuers. With that in mind, the output below shows population change versus Flood risQ Score for all inland counties.
While we’re not yet claiming causation, there is more than ample evidence for statistical correlation. While that sinks in, and keeping that in mind, lets move to this week’s Preliminary Official Statements…
This week’s climate risk non-disclosure distinction goes to:
Glendale Elementary School District No. 40 of Maricopa County, Arizona — School Improvement Bonds, Series B
Maricopa County’s Glendale Elementary School District 40 (GESD40) is located just 15 miles northwest of Phoenix, AZ. Eyeballing the risk profile of GESD40, the District appears to be taking on water and, worse yet, the District’s calculation of its climate risk appears well underestimated and under-disclosed. The Bonds are general obligations of the District, payable from ad valorem taxes levied against all taxable property within the GESD40. The Series’ maturity date schedule runs from 2022 out to 2036.
Calculating risk is straightforward — one need just apply the formula hazard x exposure x vulnerability. risQ’s value-at-risk outputs are a direct function of modeling the impacts of naturally occurring hazards on property tax base and GDP exposure. As we show below, GESD40’s value-at-risk numbers are high when compared to other school districts in Maricopa County. But, let’s start with vulnerability component. Besides the influx of funding trickling down from the Federal level through the State of Arizona, and into the coffers of Maricopa County1, GESD40 appears strikingly more vulnerable than its encompassing and municipalities — the City of Glendale and Maricopa County.
Student enrollment in GESD40 has either declined or stayed the same in every year since 2014. Enrollment has decreased over 22% from a 10-year high of 13,989 students in 2014 down to a 10-year low of 10,811 in 2020 (pg. 166). Property taxes in its General Fund have declined each year since 2017, down 5% in 3 years (pg. B-15).
Municipal officials blame declined enrollment on increased competition from charter and private schools, the lack of affordable housing, and low birth rates following the 2008 recession.2 And, now schools in the district are starting to close — the District currently comprises 17 schools (+1 online), but due to enrollment declines, 2 schools (Imes and Sine) are set to close in 2021-22, and 3 additional schools (Coyote Ridge, Desert Garden, Bicentennial) are set to close in 2022-23.
Assuming we consider enrollment changes as a proxy for population change, the population within GESD40 appears to be shrinking. We’re also seeing lousy residential demographic and economic data coming out of GESD40:
- High percentage below poverty line of 29% — 84th percentile statewide
- Low median household income of $38,775 — 25th percentile statewide
- Per capita income of $16,994 — 3rd percentile nationally
- High percentage of ‘stressed renters’ of 41% — 91st percentile statewide
- Lack of health insurance among adults aged 18–64 years of 25% — 99th percentile nationally
At the same time, the surrounding area appears to be signaling robust economic growth:
- Substantial population inflows to Arizona3 and Maricopa County — Arizona only lags Texas and Florida in terms of total population growth, and only Idaho and Nevada in terms of percentage population growth (2018-2019)
- Robust property value increases in Maricopa County — from 2012-2018, real estate values increased 58%, 95th percentile nationally
- Healthier economy in the City of Glendale — median household income of $53,965, 83rd percentile statewide, versus GESD40’s median household income of $38,775, 25th percentile statewide
- Note: population growth over the past 10 years in Glendale City (~7%) is lagging well behind both Maricopa County (~14%) and the State of Arizona (~12%).
Fitch claims4 that GESD40 “benefits from its location in Maricopa County … whose broad and diverse economy and employment base represent the hub of the state’s overall economy.” Hm. While that may be the case for other areas of Maricopa County, that does not appear to be the case for GESD40. As a matter of fact, the opposite seems to be happening — other areas of Maricopa County appear to be pulling population and investment growth from Glendale and GESD40 (Glendale has the lowest median household income out of the list of Maricopa cities/towns highlighted below).
Lack of affordable housing, alternative school options appear to be eroding the desirability and exacerbating the vulnerability of GESD40. GESD40 is a loser in many socioeconomic categories when compared to its neighbors; however, the District is a big ‘winner’ in one crucial category that could exacerbate its socioeconomic woes: flood risk.
risQ projects that the cumulative property value at risk in GESD40 by the year 2030 is 42%, ranking in the 91st percentile nationally and the 96th percentile statewide. These loss projections are driven by one hazard in particular, inland flood (property value-at-risk by 2030 is 18.2 for inland flood, and only .01 for wildfire).
A 100-year flood event equates to a cumulative property value-at-risk of over 21%, and a whopping 48% GDP-at-risk. These numbers will only continue to worsen under climate change and as the NFIP gradually starts hiking flood insurance premiums.GESD40’s 2036 debt is currently priced at 2.05%, and 2030 debt at 1.3%. Two weeks ago, Fitch published a AA+: Outlook Stable rating for GESD40.
Armstrong County, PA (risQ Score 1.9; Highest Peril Score 3.8 (Flood)): 88th percentile overall for counties nationally but 93rd percentile for inland flood, but you won’t find any mention of flood risk in this Preliminary OS. That said, there are all the indicators that the risk is having tangible impacts. FEMA buy-outs of serially flood-prone properties has begun. The population has consistently decreased by 0.8% annually over the last 10 years according to both data in the OS as well as our data that is more recent. Contrast this with Butler County that sits immediately to the west and similarly placed on the outskirts of Pittsburgh. Butler has half the flood risk and has seen population marginally increase over the same period. The vast majority of the flood risk to residential property is uninsured, driven by disparities between FEMA’s SFHA designations and the greater extent of flood risk that exists, as well as low overall takeup rates even for properties that are SFHA. The flood risk toxin has not fully manifested itself across into property value, with assessed and market value still both marginally rising, but in a generally appreciating real estate market over the same period, the growth is tepid.
North Kingstown Public Schools, RI (risQ Score 2.7; Highest Peril Score 4.2 (Flood)): As seems to be common with Rhode Island issuer, North Kingstown Public Schools – the obligor inside the Rhode Island Health and Educational Building Corporation issuance – isn’t messing around with their climate risk disclosure. On page 5 of the OS, and as the first section in “Risk Management”, you’ll find an extensive discussion of climate change and reference to a plethora of coastal flood actions plans already under way to address those risks. The discussion also references the town’s participation in FEMA’s Community Rating Survey program, a national initiative and database we also check when evaluating a given issuers efforts in mitigating risk. We wish more issuers would 1) engage in this program; and 2) take the time to disclose it in their OS. All that said, the town could be doing more, given that their CRS rating is only a 9 (on a 1-11 scale where 1 is best). This rating is more akin to still being in the planning stage versus genuine execution and enables only a 5% saving on flood insurance for residents. In addition, while the discussion of risk from sea level rise is commendable, the district is 94th percentile nationally overall and half of the risk is actually from inland flood as opposed to coastal flood. The county of Washington, RI as a whole has seen its share of NFIP flood claims at $7,600/capita/year over the last 20 years and, despite this, a good 80% of the flood risk to residential property remains uninsured. The risk disclosure in this OS is definitely better than the vast majority, so credit where credit is due, but still some brickbats to go with those bouquets.
Midland Public Schools, MI (risQ Score 0.4; Highest Peril Score 1.3 (Flood)): This one’s a bit different as the risQ Score is far from alarming. Some might recall that the greater Midland area had an extreme flooding event in May of 2020 driven by a combination of a dam failure and an extreme precipitation event. We in fact did a full analysis of this event in the context of infrastructure failure overlaid with climate events to gauge how common such an occurrence might be. For those of you that missed it, the images below overlay the age (left) and height (right) of Michigan dams (excluding the Upper Peninsula) with risQ’s cumulative property VaR for Michigan counties in 2029. For clarity, the Edenville and Sanford dams that failed are circled in red. Both the Edenville and Sanford dams are relatively old and relatively high. While not in the pre-1920 bucket, both commenced operation in 1925, so not far from it. Edenville, in particular, is a relatively large dam in the context of Michigan and has a nameplate storage in the top 2% of all dams nationally. Gladwin County, sitting immediately north of Midland, has relatively high flood risk and, given Edenville’s age, size, and location was the wrong dam in the wrong place and failing at the wrong time for Midland.
We raise this again here as the OS from Midland Public Schools fails to mention the flood even once. That’s remarkable given the severity of damage and disruption it caused. The observant amongst will note, however, that the school system’s enrollment dipped by 4.3% for the 2020-2021 school year versus the stable enrollments experienced for the prior 5 years. Coincidence, or do you think there is something to this population vs flood thing we keep talking about? Regardless, the lack of mention of last year’s flood is just a massive oversight in terms of disclosure.
1In June 2020, Gov. Doug Ducey set aside $370 million in federal CARES Act funds for Enrollment Stabilization Grants to help protect schools’ budgets against an anticipated decline in enrollment due to the COVID-19 pandemic by ensuring that schools received at least 98 percent of their previous year’s state funding.