Are Eliot Spitzer's Insurance Lawsuits Hitting the Wrong Targets?Published: October 19, 2005 in Knowledge@Wharton
On September 15, New York State Attorney General Eliot Spitzer indicted eight former executives from Marsh & McLennan Companies -- Marsh itself was not charged -- for their part in an alleged bid-rigging scheme involving insurance brokers who got insurance providers to submit inflated quotes that made customers think they had received the best price for their policies, when in fact they hadn't.
For Spitzer, bid-rigging is just part of the problem with the insurance industry. He says that another practice, contingent commissions, also artificially inflates the price of commercial insurance. Contingent commissions are extra payments, above ordinary commissions, that insurers funnel to brokerages in return for more business. Despite the fact that contingent commissions appear to be legal, Spitzer has linked them to kickbacks and is using his clout to get insurance brokerages to stop the practice.
It is appropriate for Spitzer to investigate bid rigging, says J. David Cummins, a professor of insurance and risk management at Wharton. But he draws the line at eliminating contingent commissions. "Contingent commissions can help keep property-casualty and other markets efficient," says Cummins, the executive director of Wharton's S.S. Huebner Foundation. "The practice may actually level the playing field by giving buyers and sellers equal access to vital market information."
This past summer, at the World Risk and Insurance Economics Summit in Salt Lake City, Utah, Cummins and Neil A. Doherty, also a professor of insurance and risk management at Wharton, presented a paper titled The Economics of Insurance Intermediaries that details their thinking on the subject.
In their paper, Cummins and Doherty suggest that contingent commissions are not the real problem with pricing in the insurance industry. Instead, there may simply be too few insurance brokers operating at the top of the industry. When a small group of dominant brokers acts as gatekeepers -- allocating a limited supply of insurance among companies seeking coverage – there is a greater potential for pricing abuse, says Cummins.
Collusion Replacing Competition
"Under an ideal insurance brokerage distribution model, insurance brokers will get requests from companies and will consider the buyer's coverage, loss history and other issues," he says. "The broker will then solicit bids from a variety of insurance providers." But the low bidder does not necessarily always win, Cummins adds. And that can actually be a good thing.
He notes that the price of the commercial insurance policy is only one of several criteria that corporate buyers consider. Other issues include the breadth of coverage, the risk management services provided, and the insurer's reputation for claims settlement and financial strength.
But according to Spitzer's complaint, pay-to-play schemes distorted the market's pricing mechanism. The state attorney general contends, for example, that each insurer that was in on the contingent commission scheme had certain "favored" business it was to receive from Marsh, Aon and others in exchange for extra payoffs. Of course the insurers would recoup their payoffs by passing them on to commercial customers in the form of higher prices.
If the market was transparent, the clients would soon realize that they were paying top dollar for coverage, and could simply migrate to other insurers. But Spitzer says that Marsh and other brokers would periodically require the non-favored insurers -- ones who were not pre-selected by the brokers to provide a particular policy -- to submit fake, inflated quotes. This way the company purchasing the coverage would think it had received a competitive quote. The "non-favored" insurers went along with the scheme because they knew that sooner or later it would be their turn to be a "favored" insurer, and the other companies would cover for them in a similar way.
The stakes were significant. In 2003, $800 million of Marsh's $1.5 billion in net income was derived from contingent commissions, reports Cummins.
Spitzer came down hard on the insurance brokers. In March 2005, he reached an $850 million civil settlement with Marsh, a $190 million policy restitution agreement with Aon and a $50 million one with Willis North America over a series of allegations concerning contingent commissions and other issues. Agreements have not yet been reached with ACE, American International Group, Zurich American Insurance and Liberty International Insurance.
But Cummins says that despite the appearance of a taint, contingent commissions -- at least when they are not accompanied by kickback or bid-rigging schemes -- actually play an important part in promoting competition among insurers.
The process begins with the way that insurers develop prices, generally basing them on information from their own risk analysis or from analysis provided by intermediaries. Cummins says that if the insurer believes it has good information about the level of risk for a particular insured, its price quotes are likely to be competitive. But if the insurer is unable to develop complete or accurate information, it may offer a non-competitive bid, or simply decline to offer a quote. That's because of the fear of a condition known as "the winner's curse," a fear that may be eliminated through contingent commissions.
The winner's curse, a phenomenon well known to people who are involved in auctions, refers to a situation where parties are bidding for something when the value of the item is uncertain. In a case like this, says Cummins, the winning bid is likely to be above the true value, and the winning bidder is usually the person who overvalues the item by the biggest margin.
Breaking the Winner's Curse
"This means that the winner will often rue his success -- giving rise to the term 'winner's curse' -- and will serve as a cautionary flag against bidding on things we do not quite understand," he explains. "The takeaway is either not to bid, or to submit a conservative bid when you are not sure of the value. In the commercial insurance market, this means that price quotations may vary considerably in their competitiveness."
But a qualified intermediary can increase price and quality competitiveness by providing the insured with access to a wider range of possible insurers. And contingent commissions, particularly those based on profit, may actually stimulate competitive bidding by aligning the interest of the insurer with that of the intermediary, and by giving the insurer greater confidence in the selection of risks and in the information provided by the intermediary.
"This can help to break the 'winner's curse' and encourage insurers to bid more aggressively," says Cummins. "Additionally, commissions based on volume can help insurers to achieve better risk diversification, and enable them to realize scale economies by spreading fixed costs of establishing and maintaining agency relationships over a greater base."
He says that the idea that contingent commissions can actually benefit the policyholder follows from a model previously developed by Michael Rothschild and Joseph Stiglitz, which showed that when policyholders know their own level of risk, but the insurer does not, market failure will occur if the insurer tries to offer insurance at the average price to both high and low risks.
"At the average price, high risks are subsidized and low risks pay premiums that are actuarially unfair," says Cummins. "As a result, low risks buy less insurance than high risks or drop out of the market, and market failure occurs – premiums will fail to cover claims and the insurer will withdraw from the market."
Rothschild and Stiglitz, he says, show that the insurer may be able to overcome the informational problem and create a viable market by offering a menu of policies, some of which appeal to low risks and others to high risks. Buyers will thus "self select" into a set of policies which enable the insurer to cover its costs.
The insurance industry certainly sees nothing wrong with the way it does business.
"Overall, the property/casualty insurance industry has historically been a highly competitive business, in terms of both the numbers of firms and the products, services and prices they offer," according to a statement from the Insurance Information Institute, a New York City-based trade group. "Commercial insurance prices fell in nine of the past 12 years and are projected to fall once again in 2005 -- a direct reflection of the competition in the industry." Cummins estimates there are about 1,000 property/casualty insurers operating in the U.S.
Is Consolidation the Real Problem?
But if the insurance industry is so competitive, what was behind the illegal bid-rigging? Cummins traces the cause to a high concentration of power at the top of the brokerage industry, noting that two firms, Marsh and Aon, together account for about 50% of the domestic commercial lines property-casualty brokerage market. In all, the top 10 insurance brokers service about 80% of the U.S. commercial brokerage market.
"A lot of merger and acquisition activity in the 1990s -- spurred on by a booming stock market -- led to a concentration at the top of the brokerage market," according to Cummins. "Too much market power leads to abuses. One solution may be to reduce the market power of the largest brokers by breaking them up into smaller divisions."
Cummins suggests that stockholders might actually benefit from such a move, since research indicates diversified firms tend to trade at a discount on the stock market, in comparison with ones that have a sharper focus. "Divestiture often increases the overall value of the firm to shareholders," he says. "Therefore, reducing firm size by divestiture would add value for Marsh and other shareholders, and improve both confidence and efficiency in the insurance markets."
According to the Insurance Information Institute, "as a result of the Attorney General's investigations, several leading insurance brokers and a few insurers have announced that they will not be making contingent commission payments going forward. Others are examining their compensation structures, and have yet to determine whether they will seek to retain incentives tied to the overall financial success of their broker relationship."
Meanwhile, while the September indictments against the ex-Marsh executives indicate that Spitzer will remain vigilant against what he sees as excesses in the industry, some industry observers thought the state attorney general might be ready to show more flexibility when it comes to contingent commissions. According to the Alexandria, Va.-based National Association of Professional Insurance Agents, in a January 31 speech at the National Press Club in Washington, D.C., Spitzer declined to endorse an industry-wide ban on contingent commissions.
According to the association, Spitzer said that the way they were used by Marsh and others was not appropriate, but that "in other parts of the [insurance] industry, they may be appropriate."
But Spitzer spokesman Brad Maione appeared to dash water on those hopes. "In the months that followed those remarks, three of the largest brokerages -- Aon, Marsh and Willis -- have reached settlements with the attorney general that addressed contingent commissions," he noted this week. "I wouldn't say it's appropriate to infer that he's making any changes at this time."