← All explorations

Who pays for climate adaptation? Municipal bonds, insurance retreat, and the infrastructure financing crunch

Who Pays to Protect Your Town From Floods? The Short Answer Is: Nobody Has the Money

| 125 nodes · 445 edges
↓ .md ↓ .db Take this into your AI — the full analysis + graph as markdown, ready to paste into ChatGPT, Claude, Gemini or any AI.

Based on analysis of a 125-node, 445-edge knowledge graph mapping the connections between climate risk, insurance markets, municipal finance, and adaptation spending…

The Basic Problem

Imagine your neighborhood needs to build a better storm drain system because heavy rain keeps flooding your street. It costs $10 million. Where does the money come from?

Normally, your town has a few options. It could apply for a federal grant. It could borrow money by selling bonds (which is like the town taking out a loan from investors). It could rely on insurance money after a flood. It could use a loan program the state set up for environmental projects.

The knowledge graph analyzed here maps out all of these options — and finds that right now, every single one of them is blocked or severely damaged at the same time. That’s the central finding: it’s not that one door is locked. All the doors are locked, and they got locked by separate mechanisms that have nothing to do with each other.

Why Insurance Is the Linchpin

Before we talk about money for preventing flood damage, we need to understand what’s happening to money for covering flood damage after it happens. Insurance is where the story starts.

Insurance companies make money by collecting premiums from many people and paying out to the few who have losses. This works as long as they can accurately predict how often bad things will happen. Climate change is breaking that model. Storms are getting stronger in ways that don’t follow the old patterns. Insurers can’t price the risk reliably anymore, so many of them are simply leaving high-risk states — Florida, California, Louisiana. They’d rather not sell the product than sell it at the wrong price.

When private insurance leaves, homeowners get pushed to the insurer of last resort: state-run programs called FAIR Plans, or the federal flood insurance program (the NFIP). But these programs weren’t designed to cover everyone. They’re buckling under the load.

Here’s why insurance matters so much for the rest of this story: when people can’t get insurance, or can only get very expensive insurance, they sometimes stop buying it. When they stop insuring their homes, lenders get nervous. When lenders get nervous, home values start to fall. When home values fall, the property taxes that fund your town’s roads, schools, and fire stations start to shrink. And when the town’s tax base shrinks, the town can’t afford to borrow money for that storm drain system — and if it tries, it has to pay higher interest rates because investors think it’s a riskier bet.

The graph identifies this chain of events as the central mechanism: insurance retreat leads to falling property values, which leads to municipal fiscal stress, which leads to higher borrowing costs, which leads to less adaptation spending. And less adaptation spending means more climate damage, which means more insurance problems. It goes in a circle.

The Loops That Feed Themselves

The graph contains several of these self-reinforcing cycles. Think of them like a toilet that won’t stop flushing.

The FAIR Plan loop: When private insurers leave, everyone crowds into the state FAIR Plan. The FAIR Plan has to cover the riskiest properties at below-market prices, which means it keeps losing money. It raises premiums to compensate, which drives away more moderate-risk customers, leaving behind only the highest-risk ones. This is called “adverse selection” — the only people who stay in the pool are the people most likely to have expensive claims. The pool gets riskier, the premiums go up again, more people leave, and so on.

The credit information loop: Before investors buy a town’s bonds, they want to know how risky those bonds are. Credit rating agencies are supposed to tell them. But rating agencies are slow to incorporate climate risk into their ratings — partly because towns don’t have to disclose their climate vulnerabilities in any standardized way. So rating agencies don’t have the information. And because rating agencies don’t demand it, towns don’t provide it. Meanwhile, the actual physical risk keeps growing in the background, unacknowledged in the price of the bonds. This is a two-node loop that keeps the muni bond market ($4 trillion in total) carrying risks that aren’t reflected in prices.

The flood insurance pricing paradox: The federal flood insurance program tried to fix its own finances by charging people more accurate (higher) premiums based on their actual flood risk. This is called Risk Rating 2.0. The problem: people in the riskiest zones, who now face the highest premiums, are leaving the program. That shrinks the pool of insured people and increases the share of risky properties. The program becomes less financially stable than before. Accurate pricing is making the insolvency problem worse, not better. But inaccurate (subsidized) pricing keeps people living in dangerous zones and inflates property values in places that will eventually flood — creating an even bigger problem when prices eventually correct.

Why Solutions Are Weaker Than Problems

One of the less obvious findings in the graph is that the edges connecting one bad thing to another bad thing tend to have higher weights — meaning they’re stronger, more established connections — than the edges connecting solutions to problems.

Think of it this way: if someone drops a bowling ball and it starts rolling down a hill, the hill is very good at accelerating the ball. If someone puts a pillow in the way, the pillow slows the ball — but not as much as the hill speeds it up.

The highest-weight solution edges in the graph top out around 8 or 9. The problem-amplifying edges frequently hit 9 or 10. The most promising partial solutions identified — things like parametric insurance products (which pay based on a storm’s measured wind speed rather than individually assessed losses) and direct catastrophe bond markets (where large public utilities borrow directly from reinsurance capital rather than going through regular insurers) — are real, and they work in some places. But they work best for the largest, most sophisticated organizations. Small towns with fewer resources tend to have more basis risk with parametric products (meaning the payout doesn’t match the actual damage very well) and can’t access the direct capital markets at all.

The Strange Geometry of the Problem

A few connections in the graph are genuinely surprising.

The same federal mortgage guarantee system (Fannie Mae and Freddie Mac, called GSEs) that is absorbing billions in undisclosed coastal climate risk is also the system that blocks property owners from getting PACE financing — a mechanism where homeowners can finance climate-resilient home improvements through their property tax bill. The GSEs won’t buy mortgages on homes with PACE liens because the PACE lien takes priority over the mortgage. So the entity that is quietly absorbing the risk is simultaneously blocking the tool that would reduce the risk.

Anti-ESG legislation in several states — laws that prohibit state pension funds from considering environmental factors in investments — turns out to have a paradoxical effect on borrowing costs. By suppressing climate-related financial disclosures, these laws keep climate risk unpriced in bond markets. Climate-aware investors, who might otherwise invest in bonds from high-risk states if they could assess the risk properly, stay away. The result is that states passing these laws to protect their fossil fuel economies are making it harder and more expensive for their own towns to borrow money.

TIF districts — a tool where towns finance infrastructure improvements by capturing future increases in property tax revenue — have a structural vulnerability: they depend on property values going up. But the reason a town needs a TIF district for climate resilience is often that climate risk is threatening property values. The financing mechanism assumes success before it has been achieved.

The Root of Why Private Money Doesn’t Flow to Adaptation

All of the above exists against a backdrop of a more fundamental problem: climate adaptation is what economists call a public good. When your town builds a better levee, everybody benefits — including people who didn’t pay for it. Because the benefits are shared and can’t be easily charged back to individuals, there’s no revenue stream to repay a bond investor. Compare this to a solar farm, which generates electricity that can be sold, or a toll road, which charges users. Those generate cash flows. A levee, a cooler road surface, a wetland that absorbs storm surge — these don’t. Private capital markets are designed to finance things that generate revenue. Adaptation doesn’t.

This is why public grants and government loans have historically been the primary tools for adaptation finance. And this is why the simultaneous elimination of those programs — federal pre-disaster mitigation grants, clean energy revolving funds, Inflation Reduction Act funding, state green banks — is so significant. The private market can’t fill that gap by its own logic.

Bottom Line

The graph’s central structural findings, in plain terms:

Every financing pathway for climate adaptation is currently constrained by an independent mechanism. Federal grants were cut. State programs were defunded. Municipal borrowing is getting more expensive due to insurance stress. Private markets aren’t designed to finance adaptation. Each of these has its own cause.

Insurance retreat is the mechanism that connects physical climate risk to municipal finances. When insurance leaves, it’s not just a problem for homeowners. It cascades into falling property values, shrinking tax bases, and degraded municipal creditworthiness.

The feedback loops are stronger than the correction mechanisms. The self-reinforcing cycles — between insurance and property values, between credit ratings and disclosure, between pension exposure and municipal bonds — move faster and with more force than the partial solutions that exist.

The solutions that work best are available to those who need them least. Direct capital market access, sophisticated parametric products, and large-scale resilience financing are accessible to the biggest, most creditworthy issuers. Small municipalities — which often face the greatest relative exposure and the least capacity to absorb losses — have the fewest viable options.

The problem will not correct gradually. Because insurance retreat, property value declines, and credit rating lag interact in ways that suppress visible price signals, the graph structure suggests the reckoning will be abrupt rather than smooth — triggered by a specific event (a wave of municipal defaults, a GSE loss event, a mandatory disclosure ruling) rather than by gradual repricing.

The graph does not show who will pay. It shows that the mechanisms currently in place are not designed to produce a payer — and that the mechanisms pushing toward crisis are better connected and more powerful than the mechanisms pointing toward resolution.