Property and Casualty Insurance Just Split in Two: What New York's Rate Bill Gets Wrong

What does the New York bill actually do?

The bill leans almost entirely on a mechanism called prior approval. In plain terms, a carrier cannot file a rate and start using it. The regulator has to sign off first, or a statutory clock has to run out, before the rate can take effect. New York already applies this to certain lines, and the bill extends it to commercial property, commercial general liability, and personal residential property.

At minimum, that builds a delay between filing a rate and being able to charge it. The regulator becomes a gate every commercial filing has to pass through. The bill pairs that gate with a transparency requirement: once a filing is effective, the filing and its supporting data become open to public inspection, and if a rate rises, the insurer has to print the size of the increase plus a written explanation of the main rating factors right on the bill.

There is one more wrinkle worth flagging, because it shapes the whole debate. The bill contains exactly one section that mandates an actual forced rate cut, and it applies only to homeowners. If a homeowners carrier is consistently profitable beyond a set benchmark, the state can step in and order a reduction. Commercial property and commercial liability get the disclosure rules and the implementation delay, but not the forced cut. As Grace put it on the show:

Why did Florida's prices fall without a mandate?

Florida is the case study for the competing theory, the supply side fix. The argument is simple: insurance is expensive because litigation is expensive, so fix the courtroom and the price follows. Florida ran exactly that experiment. It shortened the window to bring a claim, ended inflated medical damages that never reflected what was actually paid, and closed the rule that made it nearly free for a plaintiff's attorney to sue an insurer.

The result showed up fast in the data. Nuisance litigation collapsed within about a year. Once that pressure left the system, the math under the carriers changed, and the price followed the cost driver down. Nobody ordered USAA to cut the check. That is the core of theory one: you do not need to regulate the price, you need to fix the thing making it expensive, and the price takes care of itself.

New York is reaching for theory two: do not wait for carriers to feel generous or for courts to catch up. Make the regulator approve the rates and force the math into public view. The Advocate app data has a strong opinion about which theory fits commercial lines.

What does the platform data show on a New York multifamily building?

To test the bill's premise fairly, the team set a control rather than comparing a pre-war walk-up to a brand new high rise. They pulled a single New York multifamily archetype: pre-war construction built around the mid 1930s, roughly five stories, sitting about three miles off the coast. Holding the building constant means anything left in the spread is about the market, not about wildly different assets.

The measure is rate online, the cost to buy $1,000 of coverage, which normalizes pricing across policy sizes. Looking at the middle of the market, the 25th to 75th percentile where most buyers actually sit, two very different pictures emerged.

Commercial property is the calmer story. The middle of the market runs about $1.49 to $4.32 per $1,000 of coverage, roughly a 2.9x spread. Wide, but not alarming for this industry.

Commercial liability is where the wrinkle lives. The typical policy is actually cheaper than property at the median, around $1.76, but the spread is enormous: roughly $0.75 up to $5.16, about a 6.9x range, more than double the property spread. The average liability rate of about $5.90 sits roughly 3.4x above the median, so most buyers pay a little while a brutal tail of very expensive policies sits up top.

Is a wide price spread the same as an unfair one?

Not automatically, and this is where the bill's premise starts to wobble. A wide spread is only a transparency problem if the price cannot be explained. Run the factor analysis on the property side and most of the spread is explained by real risk factors: construction, location, and how the coverage is structured. That is risk based pricing doing exactly what it should. A spread that is wide but fully explainable is not an opacity problem at all.

The one place the spread is genuinely hard to explain by the building alone is liability. On the same controlled archetype, the model pulls property pricing down and pushes liability pricing up at the same time, comparable risk underneath both, opposite directions. That is not noise. That is the market making a decision.

So what pushes liability up? The single largest factor is location: population density, the local loss environment, and above all the local litigation environment. What juries in that area tend to do, and how large settlements tend to run, matters more than anything about the physical structure. It is bigger than any building component in the entire analysis.

Isn't a coastal building really a catastrophe story?

That is the assumption almost everyone reaches for, and the data says no. This archetype sits about three miles off the coast, and when the catastrophe factor is isolated on its own, it is real but modest, nowhere near the driver people assume. The catastrophe exposed line, property, is the one that is softening. The line that is genuinely hardening, liability, has almost nothing to do with distance from the coast.

Sharpen it one level further and the hardening is not even spread evenly across liability. It concentrates specifically in umbrella and excess coverage. Umbrella is climbing while commercial property just posted a record national decline. This little New York archetype is not a local quirk. It is the whole national market tied up in one building.

How does this connect to the K-shaped market?

This is "Joe's K" made literal. In an earlier episode, guest Joe Zuk argued the insurance market has stopped pricing neatly by line or geography and started pricing by risk quality cohort, and that if you simply stack property and liability next to each other they trace a K by themselves. Here it is, not a theory on a whiteboard but one newer building type pulled straight from transaction data: property heading down one arm, liability heading up the other, comparable risk underneath both.

One honest caveat: the two segments are comparable but not an identical book. The property set skews a little larger and the liability set a little smaller, though rebuild cost per door is almost the same on both sides, around $180,000. The magnitude shifts a little depending on how you slice it, but the direction does not. Property down, liability up, location as the driver. That holds.

So what is the honest scoreboard on the bill?

Put it together and the asymmetry is hard to miss. The bill puts commercial property under prior approval, but the data shows commercial property is already correcting downward on its own, with no mandate at all. So on property, prior approval regulates a problem that is actively fixing itself, and worse, it adds lag. You put a delay on a market that is finally repricing in the buyer's favor, which can slow the relief already on its way.

Meanwhile the line that actually hurts, commercial liability and specifically umbrella, is driven by location. The bill does subject liability to prior approval, so it at least touches the painful line, but it does not fix it. Prior approval does not change the cost driver, it delays the filing. Forcing a carrier to publish a high price and sit on it for 60 days does nothing to the local litigation environment, which will not change in two months. Disclosure does not reach the courtroom.

There is a politics footnote, offered as context rather than a verdict: plaintiffs' attorneys have been a meaningful force in Albany for years, which is part of the backdrop for why a legislature might reach for rate regulation rather than tort reform. Tort reform is the lever that would actually touch the liability driver, and it also happens to be politically hard to pass. So the legislature reaches for the tool it can move.

Worth stating plainly: this is not law. It is one member's bill, referred to committee, with a companion bill in the Senate, and with the session wrapping up the most likely outcome is that it stalls.