Michigan bans price optimization for property and casualty insurers under a law that took effect immediately upon signing on September 21, 2026.
Senate Bill 1013, sponsored by Senators Jeremy Moss, Mary Cavanagh, Stephanie Chang, Jeff Irwin, and Mallory McMorrow, amends the state's insurance code to classify price optimization as an unfair method of competition and an unfair or deceptive act or practice. It is now Act 98 of the 2026 Public Acts.
So what exactly is price optimization? In short, it is the practice of setting premiums based on how much a policyholder is willing to tolerate paying - rather than on actuarially justified factors tied to risk of loss or expense. The new Section 2027a of the Michigan Insurance Code draws a clear line: if the pricing input is not about risk, it does not belong in ratemaking.
The statute spells out three specific prohibitions. Insurers cannot factor in the likelihood that a customer will shop around, switch carriers, or cancel. They cannot estimate a policyholder's willingness to pay more than other policyholders for the same coverage. And they cannot use any measure of price elasticity of demand, whether for an individual or a group.
The law also defines what counts as "activities that result in insurance policy turnover" - the behavior insurers are now barred from predicting. That list covers shopping with other insurers for a lower premium, canceling before term, failing to renew at renewal, and complaining to the insurer or its agent.
Michigan is far from alone on this front. The bill's lead sponsor, Senator Moss, cited roughly 20 other states that already outlaw the practice. Michigan's own Department of Insurance and Financial Services had already largely banned price optimization through a March 2024 regulatory bulletin, but the new law codifies that position directly in statute - giving it the permanence of legislative authority rather than relying on regulatory interpretation alone.
For P&C carriers writing business in Michigan, the compliance question is immediate: any rating algorithm that factors in policyholder retention behavior, willingness-to-pay estimates, or demand elasticity needs to be scrubbed before the next filing cycle.