The following op-ed by Harvard Law School Professor Mark Roe ’75, “What happens when the government enters the ring?” appeared in the July 21, 2009, edition of Forbes.

Bernanke and Paulson are still taking heat from Congress for pressing Bank of America’s Lewis into going forward with the Merrill Lynch purchase, a deal that shackled Bank of America with significant losses. And Bank of America’s Lewis took considerable heat from its shareholders for not telling them how bad Merrill looked at the time of the purchase.

Eventually, the Treasury put another $20 billion into Bank of America and documents now indicate that the government raised the possibility of ousting the bank’s senior management if the deal had not gone through.

Several core transactions in the financial crisis have the government in a dual role, as simultaneously being a regulator and a market-like player. It’s as if the referee in a sport started fielding his own team. Even first-rate refs doing their job well, and as fairly as they can, can distort how everyone else plays the game, once the referee becomes a player too. This problem also emerges when the governmental regulator becomes a market player too, as was the case three times in the past year: with Bank of America’s purchase of Merrill Lynch, when the government was standing behind Bank of America as a vital lender; with Morgan’s purchase of a failing Bear Stearns last year with the Fed and the Treasury brokering the deal; and with Chrysler’s rescue via government loans.

A standard objection to the government as market player–as, say, an owner of companies like GM and Chrysler–is that it’s a bad manager. It wastes resources, makes mistakes and misallocates capital. It’s insulated from market incentives.

But recent evidence suggests it might not be so bad as a manager. And when the government meddles with or replaces failed managements–viz. the American auto industry–it’s not replacing America’s most admired management teams, but its worst. The bar for it to clear is not all that high.

The government’s goals are usually seen as the bigger issue. Rather than profits, the government-as-owner seeks to maintain employment or another nonprofit goal. Sometimes these further sensible social policy. But because it isn’t focused on profits, the government often puts capital where it’s less effective in the long run. These reservations to the government as market player are standard.

There’s a third issue with the government as market actor, one that’s potentially as insidious as any of the others, but less vivid.

When the government enters the market it doesn’t leave its governmental powers behind. It brings its governmental muscle into a market where the players have understandings, norms and rules, all of which a muscular government acting like an ordinary market player–but really still a government with strength, power and weight–can disrupt.

First, when the government enters a market, private players become disoriented when dealing with a player that appears as a market peer, but one with huge resources, one that can act unpredictably and one often with nonstandard goals that can disrupt the markets it’s trying to fix.

Second, when the government is a market player, it doesn’t leave its capacity to make the rules behind. It can disrupt market rules, precedents and practices, and that in turn can disrupt normal market flows. We have seen this three times in the past year: in Chrysler, Bear Stearns and Bank of America’s purchase of Merrill Lynch.

Consider Bank of America’s purchase of Merrill Lynch. The deal followed the usual merger sequence: first, the boards agreed to the merger. Later, shareholders got to approve or veto the deal, and thereafter the deal closed. When Merrill’s deeper troubles became apparent last December, the U.S. Treasury–Bank of America’s principal financier via the U.S. Treasury’s bailout loans–asked Bank of America to go forward with the deal anyway.

In normal securities law terms, Bank of America was obligated to tell its shareholders what it knew about Merrill’s deteriorating conditions. In normal contract terms, the bank probably could withdraw from the deal because Merrill had deteriorated too much and its contract allowed it to withdraw in the event of a “materially adverse change” in Merrill. But it neither told its shareholders nor withdrew from the deal. Government policy disrupted normal understandings of what gets disclosed to shareholders and what gets done in a merger deal.

If afterward, the market players revert to old form, no damage done here. But if the understandings change–if other think: Bank of America didn’t have to tell its shareholders, why should we?–the transaction could produce lasting damage.

The same sort of problem was in play when JP Morgan Chase acquired Bear Stearns last year. There, speed was critical to the government and of less concern to the private players. Under normal corporate law rules, shareholders would have had to approve the deal, but shareholder approval–a check on the board’s decision making–would take more than just a weekend. So the deal went forward without a shareholder vote.

In the resulting shareholder suit, Delaware courts ducked the issue, rather than risk disrupting federal policy (or re-doing its rules for all corporations or finding a technicality). It didn’t want to collide with the Federal government. One might suspect that if JPMorgan Chase and Bear Stearns had acted similarly without the government involved, courts would have acted differently.

In the Chrysler bankruptcy, the Treasury made rescue loans that no private lender would make–loans that few expect the government to fully recover. It acted like a lender, but not with a lender’s normal concern of making money on its loan. Capital markets could not tell who was getting what from whom in the deal. Real marketplace lenders in a bankruptcy expect to make money, but that was secondary to the government decision on the terms on which to lend. The bankruptcy promised UAW’s retirees more than the financial creditors. Normally that signals a severe priority problem. Surely the UAW got a present paid for with the government’s money, lent on a nonstandard, noncommercial basis for policy reasons. But capital markets players fear it also got a present out of the creditors’ money. Capital markets players assume the worst and then say they’re wary of lending to similar firms.

This third problem is potentially insidious–distortion or destruction of normal market understandings in order to implement government policy. In a market economy, the government may find that it can best reach its policy goals by acting like a market player, by saving Merrill Lynch when telling Bank of America to buy it and financing the purchase, by turning Bear Stearns over to Morgan, and by lending to Chrysler. Market structures are already in place, so it can flood one channel or another with money to make its policy work.

And its policies can seem more legitimate because it mimics market actions. Or it may want sometimes to obscure its policies–by rendering opaque whom is getting bailed out and how much, or how weak a company really is. Structures, rules and practices then emerge that weren’t designed to work well when the government is in the room get distorted.

The government’s muscle as a market player disrupts understandings, norms and patterns that usually govern commercial transactions. That’s what happened in three of the major government-funded restructurings in the past year. Sometimes regulators have no choice but to enter the market themselves to implement policy, but there are costs to doing so.

Mark J. Roe is a professor at Harvard Law School.