Sheila Bair, formerly the head of the Federal Deposit Insurance Corporation, says we should link Congress's pay to performance. Bair would pay part of elected officials' compensation in 10-year Treasury bonds and release that compensation only when they hit certain performance benchmarks. It's an intuitively appealing idea, but it would not work well.
The first problem with Bair's proposal is the idea of paying Congress in long-term bonds, so that they would lose money if bond yields rise. Bond yields are an ambiguous indicator: They can rise because of increasing worries about inflation or default, but they can also rise because of improved expectations for real economic growth.
Right now, the most likely reason that Treasury yields would rise is the economic recovery becoming stronger. We certainly don't want a pay structure for members of Congress that would make them less eager for a rapid recovery. Payment in Treasury bonds would also reward Congress for further reductions in long-term inflation expectations, which would also be undesirable.
You could fix those problems by paying Congressional bonuses in cash, but other problems would remain. Bair would withhold performance pay until benchmarks are met for labor force participation and gross domestic product growth. In general, rises in GDP and labor force participation are signs of improved well-being.
But if you start to tie congressional pay specifically to those measures, you encourage Congress to find policies that "improve" the measures, whether or not that actually makes the public better off. We could boost labor force participation by handing out government make-work jobs to the unemployed. We could raise GDP by gutting environmental regulations.
In other words, performance pay could have the same negative effects that we sometimes see in publicly traded companies: Lawmakers would be discouraged from taking the long view on policies like carbon regulation.
Tying pay to economic indicators would also politicize the generation of those indicators. Decisions about how to measure the labor force and the economy are not always clear cut, and statistical agencies could come under pressure to produce "good" numbers, undermining confidence in economic data.
Finally, Bair proposes linking a part of Congress's pay to a voter option: Voters would approve the release of bonuses if they were satisfied with the job elected officials did. But we already have a mechanism like this: elections.
If we shouldn't tie congressional pay to economic performance, what should we do to improve policy production? I have one idea: Reform the organizing rules of the Senate. Changing the rules so that legislation and nominations could pass the Senate on simple majority votes would make it easier for Congress to respond to economic challenges.
The problem with Congress isn't that its members don't care whether the economy grows. It's that institutional paralysis and supermajority requirements prevent it from taking action to grow the economy. This is a problem that is better addressed by changing the institution's rules than changing its members' incentives.
This column does not necessarily reflect the opinion of Bloomberg View's editorial board or Bloomberg LP, its owners and investors.