New York Attorney General Eliot Spitzer launched a complaint against the insurance brokerage arm of Marsh & McClellan last November with his usual flamboyant press release accusations of “widespread corruption.”

Unsurprisingly, Marsh recently settled in the usual way — by writing a big check and doing whatever Mr. Spitzer asked. That included replacing its chief executive officer with an old friend of Mr. Spitzer’s, Michael Cherkasky.

There was no trial, of course. Writing in Slate, Daniel Gross noted “Spitzer doesn’t like taking cases to trial. Instead, he has developed a more powerful tactic: He exploits the threat of stock declines and business losses to force industries to change.… He didn’t simply indict. He issued press releases.”

Trial‐​by‐​press‐​release circumvents truth and justice. No judge ever separates “findings of fact” from fictional prosecutorial accusations. The accused never get to face their accusers (usually competitors). And no jury is ever asked if Mr. Spitzer’s complaints have been proven beyond reasonable doubt.

In the Marsh & McLellan case, there is ample room for reasonable doubt, regardless of the settlement. There is, first of all, the sheer implausibility of Mr. Spitzer’s claim that sophisticated purchasing agents at the nation’s largest corporations could be routinely overcharged by one of many insurance brokers without competing brokers spilling the beans and corporate buyers either catching on, shopping around or simply bypassing the middleman (as many do).

More to the point, the evidence in Mr. Spitzer’s formal complaint indicates many seemingly incriminating statements were taken out of context, as in the Merrill Lynch case (as I proved in “Spitzer’s shakedown,” archived at cato​.org).

Paragraph 55 of Mr. Spitzer’s complaint, for example, claims, “Marsh asked ACE to refrain from submitting a competitive bid because Marsh wanted the incumbent, AIG, to keep the business.” This was followed by a quote that seems to suggest ACE “could get to $850,000 if needed” — the same bid as AIG.

Yet the next sentence of that letter from an ACE executive was suspiciously omitted. It said, “Apparently both Marsh Atlanta and Client are extremely unhappy with AIG, and if we can put offer on table at $800,000 we’ll get it.”

Paragraph 66 quotes grievances Marsh regional managers had with the New York office, Marsh Global Booking. One asked, “What are the rules on pricing — are we to quote our numbers or what MGB [Marsh Global Booking] wants us to quote?”

In Mr. Spitzer’s complaint omitted, however, the next sentence: “We get more price‐​driven deals from them [MGB] than anyone. The local offices don’t push us for price the way they do.” This regional manager’s actual pricing doubts were about being pushed to lower prices — for price‐​driven deals.

Paragraph 63 claims Munich‐​American Risk Partners disclosed to a client “contingent commissions that were being passed on to the client.” But that is not what the cited letter says. That letter from Risk Partners actually complained that commissions could not be passed on and were therefore burdensome. The writer thought Munich and Marsh should “find a way to share in the financial impact rather than having Risk Partners (not clients) share the entire burden.”

Paragraphs 64 to 66 echo gripes from Munich about not getting much business from Marsh. Yet undisclosed parts of this same letter admits this was because Munich’s bids were too high: “Zurich is currently very aggressively pricing umbrella business.… We currently stand almost no chance of competing with Zurich.… Their pricing approach seems to be just what MGB is looking for.… Pricing still drives that operation (MGB).”

Again, the complaint was about Marsh selling insurance too cheaply. Munich complained too much of Marsh’s insurance business was going to Zurich — a company that paid no contingent commissions to Marsh — because MGB liked Zurich’s aggressively low pricing approach.

Assertions about high prices in the Spitzer complaint are just unverifiable hearsay, with no exhibits at all, about high prices nobody paid. Paragraph 59 says: “Typically, Hartford’s underwriters were told to price the quote or indication 25 percent above a particular number, and that by doing so Hartford need not worry that it would get the business.” If Hartford “typically” quoted too high a price, then Hartford need never “worry about” selling any insurance at all. Why would Hartford go along with that?

Similarly, the complaint asserts, “A cast of the world’s largest insurance companies… have paid hundreds of millions of dollars for Marsh to steer business their way.” Yet Marsh could not steer business to one company without steering it away from another. Why didn’t the others complain? The only evidence proposed of such “steering” is in paragraphs 70 to 74, concerning the Greenville South Carolina School District.

There, the choice among four bids came down to ACE and Zurich. ACE paid contingent commissions to Marsh but Zurich did not. Yet Zurich won the bid. If that was steering, Marsh needs driving lessons.

Mr. Spitzer does not claim the highest bid won in this or any other case. Yet the complaint nonetheless asserts without proof that “clients purchased insurance at prices higher than they would have paid.”

In short, to make his case with the press — which is the only place these complaints are ever aired — Mr. Spitzer relied on deception, concealment, suppression and false pretense.

This is not to say none of the evidence appears damaging to Marsh. Some accusations and comments look suspicious. But it is hard to interpret them without seeing accusers challenged in a courtroom. And so many of Mr. Spitzer’s key complaints appearing to be misquoted or unverified make the overall complaint seem unequal to close scrutiny by a judge or jury.

Question: Why do companies like Marsh & McLellan rush to settle if they’re not guilty? Answer: The Martin Act.

The 1921 Martin Act gives New York’s AG unique power to smear a company, industry or profession in the press and to threaten to unleash interminable class‐​action suits unless the companies act contrite and write a big check to Albany.

Writing in Legal Affairs last June, Nicholas Thompson explained the Martin Act empowers the New York attorney general “to subpoena any document he wants from anyone doing business in the state.… People called in for questioning during Martin Act investigations do not have a right to counsel or a right against self‐​incrimination.… To win a case, the AG doesn’t have to prove that the defendant intended to defraud anyone, that a transaction took place or that anyone actually was defrauded. Plus, when the prosecution is over, trial lawyers can gain access to the hordes of documents that the act has churned up and use them as the basis for civil suits.”

Eliot Spitzer is the best gift to trial lawyers since asbestos. It would hardly come as a surprise he can count on the mass class‐​action lawyers’ financial gratitude when he runs for governor. Unfortunately, the economy of New York City cannot prosper on fat legal fees alone.

Every Spitzer blitz has been followed by job cutbacks at the affected firms — 3,000 in the case of Marsh, with similar job losses at numerous banks, brokerage houses and mutual funds accused of assorted unproven, vague transgressions.

All this has taken its toll on the nation’s financial center. Ever since Eliot Spitzer began attacking financial and insurance industries, the New York City unemployment rate has been remarkably high — rising from 7.9 percent in 2002 to 8.4 percent in 2003, and still 6.2 percent at the end of 2004. The latest year‐​to‐​year job growth was only 0.9 percent in the city, about half the national average of 1.7 percent.

These economic consequences of raising the cost and risk of doing business in Manhattan bring up interesting political questions: Can anyone win New York’s governorship without winning New York City? Can New York City afford Eliot Spitzer?