Bryan Caplan  

Voting, Vote-Selling, and Externalities

PRINT
Am I a heterodox Keynesian?... A Fertile Criticism of Happine...

Philosopher Michael Sandel asked Greg Mankiw whether people should be allowed to sell their votes. Mankiw's answer:

[T]he standard argument for unfettered voluntary exchange does not apply because there are externalities. That is, when one person sells his vote to another, that transaction may affect unrelated third parties through the electoral process.
On reflection, though, the sale of the vote is not the source of the externality. The externality comes from voting itself! This is precisely Bastiat's argument for restricting the franchise to the well-informed:
Because it is not the voter alone who must bear the consequences of his vote; because each vote involves and affects the whole community; because the community clearly has the right to require some guarantee as to the acts on which its welfare and existence depend.
The lesson: If you're afraid of vote-selling, you should be afraid of voting as well.


Comments and Sharing





COMMENTS (15 to date)
Brandon Berg writes:

Besides, vote-selling is inherent in democracy. The principal means by which politicians procure votes is by promising to steal money from some taxpayers and give it to others. With an honest market in votes, people would have to use their own money, which could conceivably improve outcomes relative to the status quo.

Mike writes:

I found Mankiw's posting very interesting. Too bad he doesn't want to hear a compliment thrown his way. Should he be reading this blog I thought I would share the following with him which accepts his challenge to use this approach in a variety of useful ways. Anyway, the following three scenarios pretty much illustrates the phenomena of earmarks pretty well.

Enjoy ...

A better example would be to introduce two public goods that were identical except for the beneficiaries. The first illustration shows why an earmark in isolation will not prevail but multiple earmarks will. The following summary table of the benefit-cost analysis shows the impact of log-rolling with earmarks.


Scenario #1: Andy tries to get his self-serving earmark projects funded but must get a majority vote.

Andy Ben Carl Public

Cost -3 -3 -3 -9
Benefit 8 0 0 8
Net benefit (cost) 5 -3 -3 -1
Vote (1=yes;-1=no) 1 -1 -1 1,-2 [failed]

Scenario #2: Andy and Ben conspire to get their respective self-serving projects passed at the expense of Carl.

Andy Ben Carl Public

Cost of A -3 -3 -3 -9
Cost of B -3 -3 -3 -9
Benefit of A 8 0 0 8
Benefit of B 0 8 0 8
Net benefit (cost) 2 2 -6 -2
Vote (1=yes;-1=no) 1 1 -1 2,-1 [passed]

Scenario #3: Now visualize this process on a grand scale where all three have their own earmarked projects and each “interest group” only myopically sees the benefit of their beneficial earmark and does not appreciate the cost to them of other earmarks. This will benefit the representatives who are “bringing home the bacon,” at the expense of the myopic voters. This benefit will be called “Myopic benefit.”

Andy Ben Carl Public

Cost of A -3 -3 -3 -9
Cost of B -3 -3 -3 -9
Cost of C -3 -3 -3 -9
Myopic Cost of A -3 0 0
Myopic Cost of B 0 -3 0
Myopic Cost of C 0 0 -3
Benefit of A 8 0 0 8
Benefit of B 0 8 0 8
Benefit of C 0 0 8 8
Net benefit (cost) -1 -1 -1 -3
Net myopic benefit (cost) 2 2 2 6
Vote (1=yes;-1=no) 1 1 -1 3,0 [passed]

p.s. Greg its sad that you had to do what you had to do and I totally understand. Just remember there are hoards of people like myself that appreciate what you bring to the table. I only wish people could be more civil and appreciative.

Mike writes:

Oops

Upon further checking the net myopic benefit should have read 5 5 5 15

Sorry about that

Gary Rogers writes:

I think the more important point that Bastiat makes is that if law is what it should be, the question of selling votes resolves itself. He argues that we can have a ruling class where a few plunder the many, a democracy where everyone votes to plunder everyone else or keep the law as it should be and no one plunders anyone. Only the third choice gives us the liberty we desire.

Bastiat's vision is amazing. Thanks for the link!

ivan writes:

So you should restrict voting because there are externalities involved? Does not sound like a libertarian argument to me. If anything externalities should not lead to restrictions upon voluntary exchange.

conchis writes:

You might argue that vote-selling merely allows the Coase theorem to function, such that the typical externalities from voting are more likely to be internalised. Unfortunately, the effect is likely to be somewhat selective internalisation (depending on who can pay), and you could easily set up a prisoners' dilemma type model to generate the result that, while it may be in individuals' interests to sell their votes (defect), they would collectively do better if they didn't. The reason this could occur is precisely because of externalities that aren't taken into account in the choice to sell your vote.

If you're afraid of vote-selling, you should be afraid of voting. But that doesn't mean you shouldn't be more afraid of voting and vote-selling together. Or that there are better alternatives to voting.

In addition, there may actually be a reasonable argument that there are additional externalities in vote selling: it reduces the independence of the signals sent by voting, which is one of the conditions for things like the Condorcet jury theorem to go through.

Unit writes:

Currently entire groups of people sell their votes in exchange for government favors. Yes the costs are diffused but the benefits are also somewhat ill-defined and go help moderately large groups. So if we were to allow vote-selling, I don't see why any one individual would have an interest in buying up votes, because government favors are rarely directed at just one individual. What could happen is that organized groups like unions or other associations would buy entire blocks, but how is this different from what is already happening?

Matt writes:

Corporations buy and sell votes by definition. Why are the externalities ignored for corporations?

Bastiat may want to restrict the franchise, but the end result, at best, would be more efficient vote selling.

I would acknowledge democracy to be a vote selling process, and make it as easy as possible for individual citizens to sell their vote.

Vote selling should assure the most economically efficient government, over long integration times.

Dr. T writes:

Ivan said:

So you should restrict voting because there are externalities involved? Does not sound like a libertarian argument to me. If anything externalities should not lead to restrictions upon voluntary exchange.
I don't view this as a libertarian/non-libertarian issue. One of the biggest problems with democracy is that the votes of stupid or uninformed citizens count just as much as the votes of the intelligent and informed citizens. Few other aspects of life work that way (well, perhaps among union workers where the slackers earn as much as the industrious). I see nothing but benefit when electorally clueless citizens sell their votes to well-informed citizens. The problem is that the votes also could be purchased by those who would directly gain (money, power, status) by the election of a bad (for the rest of us) candidate. I see no way around this problem, so I reluctantly oppose direct vote selling.

I don't remember where I first read about this idea, but I support voter tests at the polls. Voters who fail to answer simple questions about the candidates or referendums cannot vote in the relevant elections. So, if I know about the presidential candidates but not about state senate candidates, I can vote for the former but not the latter. This would have the added benefit of weakening party affiliations: there is less likelihood of winning just by being in the locally dominant party. The voters must know about your positions and how they differ from your opponents'.

greenish writes:

The question posed was

If you economists are so in favor of voluntary exchange, would you extend that conclusion to letting a person sell his right to vote to another person?

but it is not exactly a choice between allowing and disallowing a voluntary exchange: it is impossible to sell a vote, with no need for any sort of coercive power, as long as no proof of individual votes are ever generated. Certainly a market will be generated by changing this situation, but it seems silly to suppose that economists (free- or of any other kind) are in favor of creating any and all markets possible.

richard writes:

matt,

what do you mean corporations buy and sell votes "by definition?" I'm pretty sure that's your opinion.

General Specific writes:

I hereby refute a few misunderstandings. Nay, they refute themselves:

"One of the biggest problems with democracy is that the votes of stupid or uninformed citizens count just as much as the votes of the intelligent and informed citizens. "

becomes

One of the biggest problems with free markets is that the purchases of stupid or uninformed consumers count just as much as the purchases of the intelligent and informed consumers.

and

"I don't remember where I first read about this idea, but I support voter tests at the polls. Voters who fail to answer simple questions about the candidates or referendums cannot vote in the relevant elections."

becomes

I don't remember where I first read about this idea, but I support consumer tests at the shopping mall. Consumers who fail to answer simple questions about the products or services cannot purchase in the relevant market.


Both voting and behavior in free markets have externalities.

I look forward to the day when the intelligent high IQ people can care for everyone else. Utopia here we come.

greenish writes:

Both voting and behavior in free markets have externalities.

Sure, and both tap water and sewage have impurities.

Graham Lawlor writes:

If all exchanges that had externalities were banned, there would be no exchanges. Externalities are unavoidable. Any extant government has externalties too. I challenge Mankiw to come up with a system that does not.

Graham Lawlor writes:

I just read Mankiw's original post. He's delusional there too. Vote trading has nothing to do with the externality in the example he offers. The initial bill could say "tax C and split the revenues between A and B" or "tax everyone $3 and pay for something that A and B, but not C gain from". You don't have to have vote trading for that kind of externality to be possible, it's in the nature of majority voting. Harvard professor, eh?

Comments for this entry have been closed
Return to top