Why Congress doesn’t work very well

The principal-agent problem

Years ago I bought a house that soon developed foundation problems. Inspection suggested that the previous owner had failed to disclose important information about the foundation and its deficiencies. My insurance adjuster confidentially recommended I hire a lawyer. A "barracuda" was the term he used for the ideal attorney in this case. I found and retained such a person and a lawsuit ensued. At my lawyer’s advice, we brought action against everyone involved: seller, real estate company, previous inspectors, firms that had done work on the foundation, the seller's insurance company, and my own insurance company (the last for not indemnifying me for my loss). The courts and the defendants themselves would figure out how to apportion responsibility.

As is common in such cases, my lawyer worked on a contingent basis. He would get a third of any settlement. This made sure that our interests were aligned. The bigger the settlement, the more he would make. If he did a lousy job and I received nothing, he would receive nothing and would be out his cost of preparing and conducting the case.

It was a strong case. He found damning evidence of a cover-up of previous inadequate work. One by one the defendants bought their way out of the lawsuit. I had told him at the beginning that I simply wanted to be made whole—that I wanted to receive the same amount I had paid for the house, ignoring the fact that houses in the neighborhood had risen in value in the time the lawsuit took. We got within about ten percent of the number I had specified. 

At this point my lawyer urged me to call it a victory and terminate the lawsuit. The additional cost to win the last ten percent wouldn't repay the effort.

Here I should say that I found my lawyer friendly but I didn't consider him a friend. Both of us had entered our agreement as a business proposition. And it was as a business proposition for him that he was suggesting we drop the lawsuit. He had other cases to get on to, and he didn't want to be bothered merely for his one-third of the remaining ten percent.

Holding out wouldn't cost me anything in cash, but there was the opportunity cost of not being able to sell the house while the lawsuit pended. Yet as a business proposition I concluded that it was worth carrying on.

At this point, it became clear that my interests and my lawyer's interests were no longer aligned. So I made a suggestion to bring them back into alignment. My lawyer should reduce his fee— that is, his share of the settlement—by the amount that would bring me up to the number I had specified at the beginning. I don't know if he had been expecting this, but he didn't bat an eye. He agreed, and the deal was done.

What I had encountered was what economists call the principal-agent problem. It occurs, at least potentially, wherever one party is making or influencing decisions about another party's assets. If you hire a manager to run your business, you encounter the principal-agent problem. If you sell a house, you have to be aware of the principal-agent problem. If you’re an author or a sports or entertainment star and you have a literary or negotiating agent, you run into the principal–agent problem.

Economist call it a problem because it confuses their economic theories. Owners of assets are assumed to make economically rational decisions regarding the use of the assets. But if the decisions are being made by non-owner agents, those agents might substitute their own rationality for the rationality of the owner principals.

Economists and people in business have devised various methods for aligning or realigning the interests of principals and agents. Lawyers can be paid by the hour rather than as a percentage of settlement. If I had made such an arrangement with my lawyer, he would have been happy to carry on the lawsuit. Indeed, he might have told me never to settle. Just such an arrangement underlies one of the plot hangers in Charles Dickens' Bleak House, where the lawsuit of Jarndyce and Jarndyce in a London chancery court goes on forever.

The principal-agent problem comes up in areas of life not strictly economic. Voters in democracies encounter it when they elect representatives to parliaments and congresses. In this case, the asset is political power, which voters hold as citizens in a republic but which they delegate to their elected representatives. 

Principals and agents are connected by contracts, explicit or implied. Business managers can be fired. Lawyers can be replaced. Political representatives can be voted out of office. But before such sanctions can be applied, agents can sometimes do serious damage to principals' interests. The legal expenses in Jarndyce and Jarndyce eat through the assets of the contested estate. Corporate CEOs can manipulate accounting to generate short-term profits on which their compensation is based, at the expense of the long-term welfare of their corporations.

Members of Congress can fail to follow through on campaign pledges. Prior to election, candidates take pains to identify their interests with the interests of their constituents. The more persuasive they are—that is, the more successful they are portraying the principal-agent problem as negligible—the likelier their election. But once sworn into office, they often conclude that their first priority is not what they have promised to voters but rather their own reelection. If this requires giving preference to well-heeled donors, that's simply a cost of doing political business.

Some states have addressed this problem by allowing early recall of elected officials. In extreme cases of malfeasance, officials can be impeached. But this is rare. Most states and most voters typically content themselves with waiting until the next regular elections.

Another remedy is a return to the days of direct democracy, when voters made decisions on legislation directly rather than through representatives. This tradition persists in some small villages and towns, especially in New England, its American birthplace. During the Progressive era of the early twentieth century, various states adopted the initiative and referendum, which mandate direct votes on proposed measures. Nearly all states require voters to approve constitutional amendments.

Direct voting eliminates the principal-agent problem, by eliminating the agents. But it becomes unmanageable if practiced too often. Which is why we are stuck with representative assemblies, including Congress at the national level. And why we are stuck with the principal–agent problem in politics. Greater skepticism on the part of voters might ameliorate the problem by making unrealistic campaign promises less successful. Voters might be quicker to refuse reelection to representatives who put their own interests ahead of voters' interests.

All of which is to say that to enforce better behavior on our political agents, we voters will have to behave better as political principals. In a democracy we get the government we deserve.