Potential oligopoly: A deep dive into competition in homeowners insurance

As the Oklahoma homeowners insurance market has come under intense pressure, Insurance Commissioner Glen Mulready has said that the market is highly competitive, citing the total number of insurers in the market — about 100 licensed to write policies and 50 actively doing so.

In October, in response to a critical letter from Attorney General Gentner Drummond, Mulready offered additional evidence of a highly competitive market: federal standards.

Oklahoma Watch’s investigation into the history of Oklahoma insurance legislation and measures of market concentration — that is, how economists quantify competition — revealed the shortcomings of those tools to evaluate competition in the Oklahoma homeowners market.

Firm Count, CR4, and HHI

Mulready has cited the number of active insurers in Oklahoma as evidence of competition in the homeowners market, a firm count system.

However, firm count is a preliminary, basic indicator of competition that can fail to account for whether market power, and not simply market concentration, is on the rise.

In other words, a higher number of firms is a good initial sign, but a market with few firms may  be competitive nevertheless, and one that has many firms but just one with dominant market power — as is the case of State Farm in the Oklahoma homeowners market — can falsely suggest the presence of healthy competition.

Beyond simply counting firms, there are two main measures of market concentration or competition: the four-firm concentration ratio, known as CR4, and the Herfindahl-Hirschman Index, or HHI.

CR4

CR4 is calculated by adding together the percentage of market share of the top four companies in a particular market. A score below 40 indicates low concentration or a competitive market; above 70 is highly concentrated and likely a monopoly. Scores between 40 and 60 indicate moderate competition, indicating a likely oligopoly. An oligopoly is a market dominated by a small number of firms with significant market shares, as in Oklahoma’s homeowners market, where the top four companies hold more than 60% of the business.

Like firm counts, CR4 can provide misleading data by failing to account for an unequal distribution of market share. For example, a market whose top four firms each control 10% of the market would be clearly competitive and would score 40 on the CR4 scale. A different market whose top four firms controlled 34%, 2%, 2%, and 2% of the market would be noncompetitive, but would still score 40 on the CR4 scale.

CR4 is a historical measure with weak sensitivity to rapid changes in how a market might adapt to technological innovations. Insurance in general, and property and casualty in particular (including homeowners insurance), is a rapidly evolving industry on a number of metrics that may leave a CR4 evaluation unable to measure true competition.

HHI

HHI compensates for some of CR4’s shortcomings, but retains shortcomings of its own.

HHI is calculated by adding the squares of the percentages of market share of all of a market’s firms, producing a number between 0 and 10,000. In 2023, the Department of Justice adjusted the numbers so that an HHI score below 1000 indicates a competitive market; above 1800 is highly concentrated, and risks noncompetitive behavior. Between 1000-1800 is moderately concentrated, again edging into oligopoly territory.

HHI is intended for the evaluation of mergers and adherence to antitrust laws; it is not perfectly suited to insurance, which enjoys conditional immunity to antitrust laws. However, even outside the insurance industry, HHI is not considered a particularly good proxy for true competitiveness, though it is widely regarded as the best tool available.

Like CR4, HHI may be too simplistic, and suffers from the same inability to evaluate the full impact of a market with a single dominant player. Counterintuitively, lower prices in health insurance have sometimes been found in markets that have higher HHI scores.

The Data Don’t Match

Oklahoma Watch’s CR4 and HHI calculations relied on market share data from the National Association of Insurance Commissioners. Oklahoma Watch also ran calculations on data from the insurance commission’s annual reports.

The two data sets don’t match.

For example, in 2022, the insurance commission measured the market share of Allstate at 6.071%; the NAIC number for Allstate was 9.48%. In 2024, the insurance commission measured the market share of State Farm at 28.6%; the NAIC number was 30.43%.

Economist Birny Birnbaum, executive director of the Austin, Texas–based Center for Economic Justice, offered an explanation for the discrepancies.

“I think the reason is that the OK report shows individual company market shares while the NAIC shows insurer group market shares,” Birnbaum said in an email.

Counting companies instead of groups can have a profound impact on CR4 and HHI results. For example, imagine three separate companies forming a single insurer group — effectively one organization — that control 26%, 2%, and 2% of a market. The HHI evaluation of the group would add the percentages together and square it: 900. However, if the insurer group is counted as individual companies, the HHI impact is reduced to 684.

According to the 2023 DOJ merger guidelines report, a shift of 100 points on the HHI scale is considered significant.

In Oklahoma, for example, Allstate Property and Casualty Insurance Company offers homeowners insurance, as do National General Insurance and Encompass Insurance, which are both Allstate subsidiaries.

“Counting individual insurance companies instead of insurer groups is clearly the wrong way to count the number of competitors in the marketplace,” Birnbaum said in an email.

A Potential Oligopoly

Oklahoma Watch’s investigation revealed that the Oklahoma homeowners market, on either the CR4 or HHI scales and using either NAIC or insurance department data, is moderately concentrated. That is, a potential oligopoly with a few companies dominating the market share.

The calculations also revealed a significant gap between the NAIC and insurance department data.

On the HHI scale, the average gap over ten years between the NAIC and insurance department data was 207.36 points. On the CR4 scale, the average gap was 9.35 points. In 2024 alone, the gap in HHI was more than 300 points.

In both HHI and CR4, the insurance department data suggests a more competitive market than the NAIC data.

In an email to Oklahoma Watch, Mulready confirmed that the annual reports counted companies, not insurer groups. Using the example above, that means Allstate, National General and Encompass are counted as three separate companies. When considering the competitiveness of markets, the department will additionally look at group data, Mulready said. He denied that there was a meaningful shift between companies and groups.

“When running those numbers, the group indexes do change but not in a significant manner that changes the competitive market scoring for them,” Mulready said.

For Birnbaum, the amount of change was not important.

“Whether the difference is big or small is not relevant,” Birnbaum said in an email. “The fact that the OID uses HHI based on companies and not groups reveals either a profound lack of understanding of competition or an intent to deceive or both.”

Oklahoma Watch (OklahomaWatch.org) is a nonprofit, nonpartisan news organization that covers public-policy issues facing the state.

This story is part of our investigative series on homeowners insurance rates in Oklahoma. You can find all the previous stories at https://oklahomawatch.org/series/hail-no-why-homeowners-insurance-is-so-expensive/