Not all politicians only seek power. Even when given alternative options similar to cases 1 and 2, most voters still support power-seekers. Why?

Expand full comment

Politicians don't do what you suggest because politics isn't about social change, it isn't about keeping promises, it isn't about doing the people's will. Politics is about the politician -- It is about obtaining *power*, then keeping the *appearance* that one is doing "something" for as long as necessary.

Your hypotheses about why politicians do not back up their promises with risk for them certainly sound very nice, but you neglected to mention the most obvious one, the one most congruent with the facts: politicians are hypocrites (many of them Social Dominant / psychopaths) -- they know they make grandiose promises and they have no intention of fulfilling them once they are elected.

Sometimes the most obvious answer is the correct one. Sometimes a cigar is just a cigar.

Expand full comment

Another possible reason for not doing (2) is that unless the politician is a single-issue fringe candidate, voters will often like parts of their platform but dislike other parts. If there's a promise you hope a politician will break, a binding commitment could discourage you from voting for them.

Expand full comment

I don't know that I'd call that a Condorcet cycle involving all four possibilities -- rather there are two Condorcet cycles that share an arm.

Expand full comment

I have a feeling that the main reason these ideas are not implemented is simply their “kookiness”. Candidates who propose radical changes like these get written off.

A politician could campaign with very little emphasis on this aspect of their platform, and then hope that their increased fidelity to their promises would get them reelected. This isn't a great strategy for a newcomer, but you might expect it would have succeeded *somewhere*.

Expand full comment

The only institutions of personal authority in the present U.S. political system that I can think of are judges and presidential vetoes. The former are often not elected democraticallyFixed that for you. It really varies from place to place whether judges are appointed or elected.

Expand full comment

Pre-nomination political commitments are designed to hold the base which nominates the candidate; after nomination in a two party system, the middle is forced to choose between the decisions of the base, which sought the committment, and the other party, which has its own base, and may or may not have commitments on the same issues..

To illustrate:

The way this question has been framed has been to ignore the interests of the middle, which does not make its choice until later. The middle prefers no commitment, but is forced to make a binary choice.

Unfortunately choices are made in an incomplete set of options.

For example, Candidate A will commit to never raise taxes, but is strangely silent on whether to balance a budge or run a huge deficit.

Candidate B commits to balance the budget, and is silent about taxes and spending.

Candidate A is also silent about what is a tax...is a copay for medicare a new tax on seniors, is it a cost shift rather than a tax, is mandating state spending by the federal governement which shifts costs back to the state a tax increase at the state level; is a private mandate a tax increase?

The point is: you get fooled just when you think there is a commitment or promise, because there is a vast open set of non-commitments which get you back to the starting point or accomplish the same thing, or give you something you would never have chosen in the first place if it was set against the thing that was "promised".

Expand full comment

Clearly, Scott Adams was wrong about that.

Expand full comment


Exactly. Politicians go out of their way to /not/ commit to things precisely because they do not need to compete X*Y up to the credible guarantee level.

Now if we had Knightian, perfectly indistinguishable candidates in every respect (same blow-dried hair, same southern twang) but for their promises (X) and credibility (Y), we might observe creative forms of guarantee.

Expand full comment

For some reason, I thought of this quote:

"Remember, you can't be wrong unless you take a position. Don't fall into that trap." - Scott Adams, Dogbert's Top Secret Management Handbook

Expand full comment

Cool - I didn't know about those three real examples. It would be great to see a survey asking ordinary folks why they don't want to vote for those parties.

Expand full comment

Perhaps voters are used to thinking that they think about non-meta issues?

We can separate politics into the election/ policymaking process, and the actioning process. As you say, voters focus on the election/policymaking process (more valuable signaling opportunities) and tend to lose interest afterward. The actioning process is far less useful for signaling opportunities and as voters are rarely involved in the actioning process, they demand fewer assurances of their politicians during this process.

Expand full comment

Voters would reject a candidate whose campaign focused on such meta issues, and prefer to support a candidate who would better help them signal their particular positions on non-meta issues. - Robin HansonThis seems like it would make the biggest difference. Voters are used to thinking about non-meta issues, so there's a strong status quo bias towards them.

I think lack of awareness is probably a strong reason. Most voters don't follow their candidates for long enough after an election to be aware of whether they kept or broke their promises, so politicians have very little incentive to keep them. The Presidential election might be the one exception. I'm reasonably certain Politifact's Obameteris going to be an important talking point in the 2012 election.

Expand full comment

Another possible reason is that the position as elected official is not very attractive with these restrictions. Perhaps no credible candidate is interested in that job?

Promise #1 has a few problems.

➢ Being on that jury seems like a real pain! Lots of unpaid work and difficult decisions. Part of why we elect representatives is to not have to go through that.

➢ Given that the elected representative gets to frame the issues, he still has a lot of influence over the "verdict"

Expand full comment

So politicians promise X with probability Y, and the question is why they don’t take steps to increase Y, in order to increase X*Y.

That's very strong idealization over what I see in the real world. In reality, politicians rarely promise anything concrete enough that you could even say what "X" is with any certainty.

If you go to a typical electoral campaign website, you won't see a list of plainly stated promises. Instead, you'll see vague and emotionally charged statements of ideological affiliation, whose practical implications are mostly left to the voter's imagination. Why this is so, should be pretty clear when one observes the reality of the modern political system that I described in my above comments.

Expand full comment


I would also challenge the assumption that people vote on what a politician PROMISES, but rather vote based on what a politician DOES.

I promise is made before an election; VOTES are actually made based on performance--that is, how a politician ACTED during office.

One is prospective; the other is retrospective.

I would argue that we vote RETROSPECTIVELY and should all but ignore promises--because we want, under some circumstances, for promises to be broken.

In other words, even if a politician kept all his promises, the retrospective look at whether they should have kept them will lead politicians to be kicked out of office for keeping their promises--deservedly I might add--because, if they keep a promise when, under the facts, they should have broken it.

So, in the end, what we do is give directional signals to elected officials and vote retrospectively based on their conduct.

Not consistent with the promise hypothesis.

Expand full comment