We’re back in Florida again this week, digging into the Official Statement for the State of Florida’s $369 billion Department of Transportation (FDOT) documentation for disclosure of climate risk. As a Full Faith and Credit offering the climate risks for Florida overall sit over the top of the risks for FDOT specifically, and that primarily means hurricane risk in terms of shocks to the local economy generally, and to Pledged Gas Taxes specifically.
The quick overview of Florida’s main risk profile first, calling on our prior analysis of the risk buried in the state’s Florida Hurricane Catastrophe Fund (FHCF). Climate change alone is predicted to increase Florida’s annual probable property losses from hurricanes by $13B, or 42%, by 2030. Conservative, linear assumptions for property value growth layered onto climate change effects, probable hurricane Value at Risk (VaR) in 2030 totals $66B, or a 106% increase versus 2020.
Approximately 52% of the residential property value at risk for Florida is insured so commensurate risk will flow through to the FHCF, requiring stronger investment performance to maintain historical liquidity for the fund, and potentially draw on common financial pools for economic activity that would also be impacted by hurricane events. A study published just this week on the financial impacts of hurricanes to municipalities is evidence of the the balance sheet pressure the State of Florida will face if it has to backstop municipals within the state. Florida does not have sensitivity to the ad valorem tax components other states do, but it therefore relies more heavily on unabated economic activity. Accordingly, the 48% of property risk that is not insured will pressure discretionary income and in all likelihood flow through to expenditure on local travel, and on other key sales receipts. At a broader level, there is inherent risk to GDP overall, and GDP has a very different risk profile across the state than property value to each peril. A higher proportion of GDP risk is insulated from coastal risk than property, driving up hurricane flood risk (i.e. from precipitation events like Harvey, Florence and even Eta) as well as inland flood as relative GDP impairers.
There is zero mention in the OS that climate change is a thing anywhere in the document and just the usual generic language essentially saying “events could happen but we can’t quantify the risk” is on page 24 of the OS. A combination of broadly believed basic science and our prior analysis of the FHCF should have put those notions to bed. The OS has a good amount of boilerplate language regarding climate risk, and hurricane risk specifically. In fact, exactly the same language for “Environmental Risk Factors” can be found in the OS for Florida International University that we looked at in more depth last week. Does anyone think that the risk factors for debt serviced by state gas taxes and the state overall are the same as risks for a higher education housing development project in the greater Miami area? With respect to gas tax receipts specifically, we can go even deeper using some pretty basic logic from public documents.
Tourism accounts for a significant proportion of the overall economy, and tourists account for a significant portion of vehicle miles traveled and associated gas consumption. In fact, a 15 year study showed that 11-12% of total vehicle miles travelled in Florida were from visitors to the state. This need to be coupled to hurricane impacts. For example, visitors to Florida decreased by 1.8 million people (5%) in the four months after Hurricane Irma according to Visit Florida’s own numbers. As a general rule, Florida’s peak tourism occurs immediately after hurricane season so that’s a bad combination in a climate changing world. Irma was a one time event and such impacts compound when multiple events and/or severe events impact higher leverage areas. As an example of the latter, the year after Hurricane Katrina visitor numbers to New Orleans dropped by 65% year-on-year. Just as Property VaR to hurricane has geospatial nuance based on concentration of property, tourism concentration is also geospatially distributed. Total county spending by visitors in 2018 ranged from a high of nearly $30 billion in Orange County to a low of $800,000 in Liberty County, with much of the higher spend cohort also being the highest risk hurricane areas. All these numbers focus on just the tourism component, but the local resident population also decreases in the aftermath of hurricane events, especially the higher severities. We’ve discussed in past analyses e.g. for Bay County FL, that population (including temporary residents there in recovery efforts) dropped for multiple years in the aftermath of Hurricane Michael, taking their fuel consumption and broader household spending with them. As such, the increasing cadence and severity of hurricanes will impact state revenue streams through lost tax receipts at the gas pump directly as well as broader receipts at the Full Faith and Credit levels.
A few other data-driven observations that can’t be gathered from the Official Statement:
- No discussion of carbon transition risk in the OS, which is curious for a revenue stream relying initially on automotive fuel tax receipts. Fuel-efficient vehicle regulations struck some headwinds in the last 4 years, but could be expected to reaccelerate. At the top end of this, EV registrations grew in the state in both 2018 and 2019 (despite decreasing nationally during the same period) and at current pace will top 100,000 vehicles in 2 to 3 years and grow from there.
- Florida has a much higher percentage of car commuters, and commutes with one person in the vehicle, than the national average. As such, fuel tax receipts will be more greatly impacted by persistent trends in remote and at-home work, which have only been accelerated by COVID-19. In addition, Florida is only at the 13th percentile among all states for percentage of commute times less than 20 minutes (see the workplace demographics tab in risQ’s UI), further enhancing lost fuel tax revenues versus other areas from lost commuters.
- In a broader recessionary environment, Florida’s vehicle use has been more negatively impacted than nationally according to the FDOT’s own data. For example, between 2008-2012, Florida’s vehicle miles dropped 3.6% compared to 0.8% nationwide.
For Florida, the exercise required is to both look at risks – climate and otherwise – to the specific revenue streams securing a specific series of bonds like gas taxes, but then also stack the those bonds that eventually draw from and are backstopped from the same state-level wallet that relies on GDP and consumer spending. Climate risk, and the increasingly growing risks associated with hurricanes, represent such large, stacked and inevitably growing burdens.