Garett Jones  

Supermajority Rule: Buchanan, Tullock, Bernanke

Last Night's Debate on Health ... Tiebout Progressivity: That Di...
Most supporters of democracy embrace the ideal of majority rule.  The great Swedish economist Knut Wicksell started from a different ideal: Unanimity.  If your idea is so great, shouldn't we all agree to it? And even if it's only a great idea for 51% of the people, if it's truly great then the 51% should be able to buy off the support of the other 49%.  

Under unanimity rule, everybody wins, every time.  Pareto efficiency in economics jargon (or is it sociology jargon?).  

It's yet another example of the Coase Theorem: Regardless of who owns the property right--here, the right to stop change--if people are good at bargaining, then we should be able to get to an efficient outcome.  

Buchanan and Tullock, in the Calculus of Consent, wisely started from the Wicksell position and dialed back slightly (check out the legendary chapter 6 here, Alex's PowerPoints on the topic here).  In real life, some people are grumpy holdouts--maybe they're bad bargainers, maybe they're just oddballs whose well-being we don't place much weight on.  In cases like that, even if a deal is worth doing, you probably won't insist on 100% agreement.  

But you'd probably still want more than a mere majority: So especially on big issues, a society that wanted to really make almost everybody better off would make government decisions with supermajority rule.  

There's another reason to support supermajority rule: Credibility.  If you live by the one-vote majority, you die by the one-vote majority.  And when actions today depend on beliefs about future government actions, credibility matters a lot.   

In the realm of monetary economics, credibility has become an obsession in recent decades. The rational expectations revolution made monetary economists worry about how beliefs about the future influence behavior today.  And the problem of credible commitment by politicians and bureaucrats--usually known as the problem of "time consistency"--brought a more explicitly political edge to the issue of beliefs.  

Investors, savers, CEOs--all care about the likely future more than they care about the promised future. Promises are cheap, after all.  

I'm guessing the issue of credibility has kept Bernanke awake nights over the last four years.  The full story of his lobbying and cajoling and persuading other members of the Federal Open Market Committee (FOMC) remains to be written, but he appears to have succeeded at creating credible expectations that the Fed will stick with his policies whether he's reappointed as Fed chairman or not.  

I could show you market based evidence of that, but I'd rather show you the mechanism that really illustrates Fed credibility: The Bernanke supermajority.  

Below is a chart from the FOMC's economic projections report, a regular report that deserves your attention.  It shows the number of FOMC participants (voters and nonvoters) who think 2012, 2013, 2014, 2015, or 2016 is the right year to start raising short term interest rates: 


Looks like they've got a solid supermajority for 2015.  Of course, these are people forecasting their own future choices, and the best choice will change as the data change.  But (setting aside the issue of which FOMC members are allowed to vote in a given year) there's a filibuster-proof supermajority behind extended easing.  

Someday, we may find out what Bernanke's own preferences have been in recent years.  But in any case, since credibility is hard to gain and easy to lose, it's worth sacrificing the best policy in order to get a credible policy.  And the credible policy isn't the policy that squeaks through by one vote. 

Comments and Sharing

COMMENTS (8 to date)
John Thacker writes:
And the problem of credible commitment by politicians and bureaucrats--usually known as the problem of "time consistency"--brought a more explicitly political edge to the issue of beliefs.

As in, for example, the problem of whether $700B in voted Medicare cuts as part of a bargain to add $1T in other coverage subsidies will actually be allowed to occur, or whether they'll be suspended when they start to bite like the Sustainable Growth Rate, right? ;)

Bernanke's performance in persuasion is all the more impressive, I think, given that so many macroeconomists (let alone ordinary people) seem to be against his policy. Though I am more than willing to say he may be right.

Ken B writes:
the 51% should be able to buy off the support of the other 49%.
But of course that's a different proposal. "I will support funding for X" and "I will supoort funding for X if there is funding for Y" are not the same.

Plus of course we get other interesting possibilities. 51% suppoprt X. Then on a second vote to pay the losers, 51% agree. But who votes on the second proposal anyway? Just the 51 % aye-sayers? Or all the voters, again?

Floccia writes:

Some one said that judging from 12 man juries that the required super majority to pass a law should be 1/12 or about 92%.

Meximum Liberty writes:

Many states have sunset laws, that automatically discontinue some identifiable set of laws (usually related to some agency that it creates).

Does anyone know of any suggestions to tie the period before sunset to the level of supermajority in the legislature? For example, if a bill passes by 90%+, give it a 20-year sunset period. If it passes by a bare majority, it lasts two years.


Thomas Boyle writes:

This touches on something I've been thinking about for a while.

Libertarians differ from anarchists in believing that some government is good and necessary, but that government "tends to go too far" in some way.
This raises the question of how much government is "enough". Everyone has their own opinion: how do you operationalize it?

Qualitatively, libertarians agree that it is appropriate for government to enforce laws against murder and fraud. Note that there is also a very high level of social agreement that murder and fraud are bad things.

This suggests that, in a democracy, there is at least a large overlap between libertarianism and a philosophy that calls for supermajority support for government action. How big a supermajority? That's less clear, but each individual's answer would almost provide a metric for "how libertarian". Thus, "100%" is anarchist. "2/3" would be fairly moderate libertarian. What about 50%? It's mainstream, but looking at the effect it's had almost everywhere, it's pretty communitarian. In fact, "majority rule" may very well be the definition of communitarian.

Interestingly, in civil rights (but not economic rights) large parts of the country have begun to accept a moral argument that, in effect, means that if something like 10% of the population wants a particular liberty, they may have a moral claim to that right (gay rights, for example), even if the majority's instinct is "ick, no!" - but if the number is much smaller than that (polygamists, perhaps, or flag burners) they may reasonably have their liberty claims denied, more or less on the grounds of being "weird" or "extreme". (Flag burning is protected in the US by the Supreme Court's interpretation of the First Amendment, but in most democracies it would not be.)

Taking this approach, would suggest that a number closer to 80-90% might be a practical limit in a libertarian democracy.

What about the asymmetry, that we might see a majority favoring passage of a law reducing the scope of government, but a rule requiring an 80% vote to pass it? Clearly, supermajority requirements must specify a default direction (i.e., it takes only a vote of, in this case, 20% to reduce the scope of government, but 80% to increase it), but there are situations where it may not be clear which is the default (laws that have mixed effect, for example).

Of course, the Constitution (at least the Bill of Rights) is based on the idea of near-unanimity, as opposed to fleeting legislative majorities. It was designed to make change difficult.

Michael Makovi writes:

What I would want to see, is a super-majority required to pass a law, and an equivalent super-minority to repeal a law. E.g., if it takes 75% to pass a law, it ought to take only 25% to repeal.

This answers Rothbard's objection, that a super-majority rule's effectiveness depends on the starting point. If you begin when there are no laws yet, fantastic. But if you begin with a situation of mountains of federal regulations, a super-majority rule to both pass and repeal will freeze us in place. My super-minority for repeal solves this.

Besides, I think it's just logical. If it takes 75% to pass, then it means that at passage, 25% is enough to block passage. If 25% is enough to block passage, shouldn't it be enough to repeal as well?

Kopfschlaeger writes:

The Coase theorem assumes no transaction costs. One major transaction cost in bargaining in a legislature is that posed by the policies of a well-organized party, whose reluctance to bargain is based on goals unrelated to the purposes of their participation in legislation. For example: prevention of the dominant party or the executive to have any substantial success.

Comments for this entry have been closed
Return to top