House Committee on Financial Services, Subcommittee on Oversight and Investigations

Deregulating Auto Insurance

I am pleased to appear before you today to discuss state regulation of auto insurance. As it turns out, the AEI-Brookings Joint Center on Regulatory Studies will release a major study of this subject in several months that was overseen by Professor J. David Cummins of the University of Pennsylvania. If the Subcommittee holds further hearings on this subject, I encourage it to seek testimony from Professor Cummins and others who participated in the study. In their absence, I will report some of its main findings.

Background and Summary of Testimony

The auto insurance industry currently collects about $120 billion in annual premiums, accounting for roughly 40 percent of overall property-casualty insurance premiums. As the Subcommittee is well aware, approximately half of the states have some form of prior approval over auto insurance rates.

The AEI-Brookings insurance study contains both a statistical analysis of insurance in all states as well as case studies of insurance regulation and deregulation in selected states, all authored by leading scholars in the insurance field.

The bottom line of all this analysis is very simple to state. Auto insurance is a competitive industry. It certainly is not characterized by monopoly, the traditional basis for price and entry regulation. Nor is the product so complicated that it requires government to set rates to protect consumers. Indeed, because it is what I would call a "plain vanilla" financial product—in large part because insurance policies have been standardized through forms regulation—consumers are easily able to use the Internet to shop for auto (and other types of) insurance. In facilitating price comparisons, the Net is making and will continue to make auto insurance—and the financial services industry more broadly—even more competitive.

In short, from an economic perspective, there is no basis for regulating rates. Furthermore, there is no evidence from either the AEI-Brookings study or in the academic literature of which I am aware indicating that either prices or profits in states that rely on markets to set rates—rather than regulation—are excessive

Experience Under Rate Regulation

What about the states that do regulate insurance? As part of the AEI-Brookings study, Professor John Worrall of Rutgers University examined the experience of New Jersey, while Professors Sharon Tennyson of Cornell and Mary Weiss and Laureen Regan of Temple University studied Massachusetts. In both of these states auto insurance rates are heavily regulated. The authors of these state case studies reached similar conclusions.

In both states, rates have been suppressed below levels that would obtain in a freely competitive environment. On the surface, this may look like a good deal for consumers, but closer study reveals deeper problems. For one thing, rate suppression not only discourages entry by new insurers, but encourages existing insurers to leave—which in fact has occurred in both New Jersey and Massachusetts. Meanwhile, many more of those insurers who remain operate only in a single state (either as standalone companies or subsidiaries of national firms that are formed to limit financial exposures to the parent companies). In Massachusetts, for example, in 1982 all top ten auto insurers in the state were national firms, but in 1998 this was true for only 3 of the top 10. A similar pattern has existed in New Jersey: five of the nation's top 10 auto insurers do not do business in the state. The net result from restrictive rate regulation is less choice for consumers among less diversified firms.

Less choice in regulated states manifests itself in another way as well. In his statistical analysis of insurance rates across states, Professor Scott Harrington of the University of South Carolina confirms that insurers in regulated states are less willing to voluntarily underwrite insurance, leaving a significantly higher fraction of consumers to buy their insurance in residual markets (where most states assign policy holders to insurers based on their shares in the primary or voluntary market). Again, Massachusetts illustrates the problem: roughly half of the state's drivers were forced to buy insurance in the residual market during the 1980s (reaching a high of 72 percent in 1989). The Massachusetts case study authors report improvements in the 1990s due to some reforms, but also observe that declining claims costs also made helpful contributions (as they did elsewhere, as I discuss later).

Furthermore, regulated rates are often distorted by political pressures in order to subsidize certain classes of drivers. The AEI-Brookings study found evidence that not only does regulation often suppress average rates, but distorts rates between different classes of drivers—keeping rates for high-risk drivers artificially low, while raising rates for lower-risk drivers. This cross-subsidization is accomplished directly through limits on rates in certain classifications or by channeling subsidies to higher risk drivers by keeping rates low in the residual market. The Massachusetts case study, for example, found that some high risk drivers receive subsidies as high as 60 percent, requiring some lower risk drivers to pay 11 percent more in premiums than they would pay in a competitive environment. Similarly, the authors of the South Carolina case study discussed shortly report that the residual market in that state ballooned under regulation to 42 percent of consumers in 1992, requiring significant subsidies from drivers in the voluntary market. By 1999, the state residual market facility had a cumulative deficit of $2.4 billion. Subsidizing high-risk drivers is hardly a desirable social or economic policy because it can lead to higher accident rates and loss costs (due to more ownership and driving by higher risk drivers).

What about the experience in California, which adopted one of the nation's best known regulatory regimes under Proposition 103 enacted in 1988? Professors Dwight Jaffee of University of California at Berkeley and Thomas Russell of Santa Clara University conclude that the harmful effects of regulation found by the authors of the Massachusetts and New Jersey case studies—exit of insurers, rising residual market shares, and rate suppression—did not occur in California. The major reason for this different result, however, is that in both absolute and relative terms, claims costs in California—especially liability costs—fell dramatically after Proposition 103 was implemented. Why did costs fall? Jaffee and Russell conclude that one reason was that Proposition 103 mandated a 20% "good driver" discount. But the more important factors, taken together, were more aggressive enforcement of seat belt and drunk driving laws, as well as the elimination in 1988 of third party lawsuits in the state against insurers for bad faith. Phillip O'Connor, former Insurance Commissioner of Illinois, has also recently testified to the fact that the most publicized part of Proposition 103—the 20 percent rate rollback—was never fully implemented (because of adverse court rulings).

In short, the California experience demonstrates that rate regulation need not produce deleterious results if other good things happen at the same time and if the regulatory regime is not that binding. But if there are upward pressures on costs, then almost by definition, rate regulation will result in rate suppression and the various negative consequences that flow from that outcome.

Experience Under Deregulation

In 1999, South Carolina substantially deregulated auto insurance rates (under legislation enacted in 1997) and began phasing out its subsidies. Professors Robert Klein of Georgia State University and his colleagues Martin Grace and Richard Phillips examined the limited data available since then and found some striking results. Before deregulation, South Carolina had an average of 59 insurers serving consumers, compared to almost 200 insurers in other Southeastern states. After deregulation, the number of insurers serving South Carolina roughly doubled. At the same time, the residual market facility in South Carolina has virtually disappeared—down to about 50,000 consumers, from a high of one million—because insurers now can charge rates based on risk in the voluntary market. Overall premiums have fallen, in part because claims costs have fallen (a result which may have been influenced by the increased use of risk based pricing).

Auto insurance has been deregulated in Illinois for over three decades (and indeed, the state is the only one in the nation without a rating law of any kind). In his study of this experience for the AEI-Brookings study, Professor Stephen D'Arcy of the University of Illinois finds that premiums in Illinois are in line with losses, that they change more frequently and in smaller increments than they do than in regulated states (as one would expect in a competitive market), and that the residual market barely exists in the state (at less than 1 percent of the market). Meanwhile, Illinois consumers have roughly twice the number of auto insurers (129) to choose from than those in New Jersey (67), where rates are tightly regulated. In sum, the Illinois experience is consistent with that of other states that have so-called competitive rating laws?laws that do not require prior approval?and the state accomplishes this result without having to divert scarce regulatory resources into monitoring rates (but can focus on solvency and market misconduct instead).

The experience from other industries where prices and entry have been deregulated also demonstrates that deregulation, by unleashing the forces of competition, helps drive out inefficiencies and thus leads to higher productivity and lower costs.10 In fact, there is evidence of significant inefficiency in the insurance industry. In another recent study, Professor Cummins and colleagues estimated that on average property-liability insurance firms could reduce their expenses by an extraordinary 32 percent if they were all highly efficient.11 Rate deregulation in the states where it still exists would help unleash competitive forces that would help realize these cost savings.

Conclusion

The economic case for eliminating rate regulation in auto insurance is overwhelming and compelling. Virtually all economists who have studied the industry over the last several decades have reached this conclusion. The obvious policy implication: auto insurance?indeed, all lines of insurance?should be governed by the market, just like other industries in our economy. Moreover, like other industries, insurance ought to be subject to the antitrust laws.

There are several roles for regulation, however: to monitor insurer solvency (so that consumers will be paid when covered events occur), to protect consumers from unscrupulous practices, and to help standardize forms for personal lines and to small businesses (so that consumers can easily compare prices). Eliminating rate regulation would free up resources within insurance departments to pursue each of these functions (especially solvency and misconduct regulation) more vigorously.


Robert E. Litan: Summary of professional background

Robert E. Litan is Vice President and Director of the Economic Studies Program and Cabot Family Chair in Economics at the Brookings Institution. He is also the co-director of the AEI-Brookings Joint Center on Regulatory Studies; co-chairman of the Shadow Financial Regulatory Committee; and co-editor of the Brookings-Wharton Papers on Financial Services (with the Financial Institutions Center at Wharton) and Emerging Markets Finance (with the World Bank and International Monetary Fund). He is both an economist and an attorney.

During his career at Brookings, Dr. Litan has authored, co-authored or edited 22 books and over 150 articles in journals, magazines and newspapers on government policies affecting financial institutions, regulatory and legal issues, international trade, and the economy in general. His most recent books include the forthcoming Beyond the Dot.coms (with Alice Rivlin) and The GAAP Gap: Corporate Disclosure in the Internet Age (with Peter Wallison).

Dr. Litan has consulted for numerous organizations, public and private, and testified as an expert witness in a variety of legal and regulatory proceedings. Among his various assignments, he has authored or co-authored a number of influential federal reports. Most recently, he co-authored two Congressionally-mandated studies for the Treasury Department on the role of the Community Reinvestment Act after the Financial Modernization Act of 1999. During 1996-97 he has served as a consu

Dr. Litan also has served in several capacities in the federal government. During 1995 and 1996, he was Associate Director of the Office of Management and Budget (where he was responsible for overseeing budgetary and other policies of six cabinet agencies). From 1993 to 1995, he was Deputy Assistant Attorney General, in charge of civil antitrust litigation and regulatory issues, at the Department of Justice. From 1977 to 1979, he was the regulatory and legal staff specialist at the President's Council of Economic Advisers. In the early 1990s, Dr. Litan was a Member of the Commission on the Causes of the Savings and Loan Crisis.

Dr. Litan received his B.S. in Economics (summa cum laude) from the Wharton School of Finance at the University of Pennsylvania; his J.D. from Yale Law School; and both his M. Phil. and Ph.D. in Economics from Yale University.