ECON 369 - Voting (b)

Lecture Six - Public Choice Continued: Voting Rules and the Outcomes of Direct Majority Voting

Sources: Government Failure by Gordon Tullock, Arthur Seldon and Gordon Brady, Median Voter Model sources - Roger D. Congleton, Randal Holcomb, others (see in text).

Small groups can make decisions on the basis of consensus, for larger groups this is much harder if not impossible due to the transactions costs involved.

The historic solution to this problem was Autocracy (dictatorships). And in some cases, still is.

We can place voting rules on a continuum, from everybody getting a vote, to only one getting one.

Example: Hard to think of examples where this is actually used except in smaller circles.

· Super-Majority – more than a simple majority is needed to win (for example 75%) (everyone gets one vote).

Example: Some states have this on proposals to raise local taxes. Proposing Amending the U.S. Constitution (2/3)

· Majority Rule – more than 50%, the most widely used rule (everyone gets one vote).

· Plurality Rule – whoever gets the most votes wins (everyone votes once).

Example: The system is often used to elect members of a legislative assembly or executive officers and most elections in the United States.

· Dictatorship – Rule by one person. Often by a larger “elite”, or ruling class.

· “One man, one vote”, is how we typically think of voting. In actuality, there are many other ways to vote that are used in practice.

· Equal Voting – everybody gets the same vote, each person’s vote counts the same

· Threshold Requirements – everybody gets the same vote, but not everybody can vote

· Approval Voting, we can vote for everything that we like. What or who wins depends on what is being voted on. One winner? Etc.

· Exhausted Voting – we vote multiple times, each time we eliminate an option

· Weighted Voting – all stakeholders can vote, but the vote is proportional to the stake held.

· Point voting: rather than a single vote, voters are given an allotment of point votes that they can allocate across different alternatives.

Sometimes known as the Borda Count Method (but there are other names). Borda was an 18th century French mathematician.

Because it sometimes elects broadly acceptable candidates, rather than those preferred by the majority, point voting is often described as a consensus-based electoral system, rather than a majoritarian one.

You might have a candidate that is ranked second by almost everyone -- two other candidates that split the top vote. In majority vote, one of the top candidates wins. In the point system, the guy who pretty much everyone likes - but is not top choice - could win. This seems more "consensual" then the minority having someone in office they really dislike.

1. What about Vote trading among voters: you vote for mine and I’ll vote for yours.

Efficiency as defined as the Marginal Willingness to Pay (a proxy for the benefit received - MWTP) is greater than the Marginal Cost.

Assume each project's marginal cost = $3,000. Each person pays a per capita share of the project or $1,000.

Welfare analysis (assuming the sum of the MWTP vs. Marginal Cost) - it would be “efficient” to do both projects, since the MWTPs are $3,200 (total).

BUT BE CAREFUL – all of this assumes some kind of measurement of utility (it is only hypothetical).

Inefficient: But suppose Tom’s MWTP for both projects is $0. Neither project is efficient because the sum of voter’s MWTPs is less than the cost. If Dick and Harry engage in vote trading, we have an inefficient outcome. In other words, vote trading can approve projects such that the MWTP, if we could measure it, is less than the marginal cost.

Given our public choice assumptions – what are some of the predicted outcomes of majority voting?

The Voting Paradox (Condorcet’s paradox): Inconsistency in political choices – even when individual voters are consistent.

Consider three people choosing among three alternatives. Their preferences are given in the table:

Using simple majority voting in which the alternative with the most votes wins, we put A against B and then put the winner, which in this case would be A, against C. C would win.

But suppose we had started by putting B against C, in which event B would win, and the B against A, in which event A would win.

Or suppose we had started by putting C against A, in which event C would win and then C against B with a win for B.

The voting paradox is that any one of the three outcomes can be reached by a simple majority voting procedure, depending on the order in which the alternatives are considered.

Ms. Middle (could possibly win because other voters outside of the group might vote them in)

The group would want to put Ms. Middle up against Ms. Worst first -- and they all vote for Ms. Worst! Then Ms. Worst goes up against Ms. Best -- and Ms. Best is more likely to win!

A very popular tool used by public choice economists especially to analyze local government activities is the median voter model.

The model was built on the foundations of Black (1958), Downs (1957) and Bowen (1943). At the beginning of the 1970s the model was accepted almost without question, but by the end of that decade it had been criticized in a variety of ways (and abandoned by some).

Median voter: a person whose most preferred outcome in an election on a single issue occupies the midpoint of voters’ opinions.

This is because the model was seen as being realistic only under certain conditions. So some have equated this model to models like “perfect competition” – very unrealistic in most if not all cases.

According to economist Randall Holcombe it depends upon how the model is interpreted:

2. Under general circumstances the demand for public sector output in a democracy is the median voter’s demand.

#1 says that what is produced by government is what the median voter wants. Is that true?

#2 says that what is actually demanded from the government is what the median voter demands - but it is not necessarily actually produced by the government. Is that true?

We have to consider The Median Voter Theory: When the voting paradox is ruled out -

Suppose the median voter has a single preference and half the remaining voters prefer a quantity equal to or greater than Qm and the other half prefer a quantity equal to or less than Qm. Qm is the only quantity that can command majority approval over the other. alternatives.

Is this efficient? Probably not. The MWTP (marginal willingness to pay) for the voters to the left is lower than the cost. The MWTP for the voters to the right is greater than the cost. What is the likelihood that these two even out?

Normal distribution of voters – most voters in the middle, there is not a great deal of polarization, and policy/politics will not change much unless a large number of voters change their position.

For example, it seems plausible that voters could do this if they can clearly place political candidates on a left-to-right continuum, but this is often not the case as each party will have its own policy on each of many different issues.

This assumption predicts that the further away the outcome is from the voter's most preferred outcome, the less likely the voter is to select that alternative.

The median voter theorem implies that voters have an incentive to vote for their true preferences.

But remember, that isn’t always the case. As Buchanan pointed out, voters might vote based on values, not preferences.

For example, as they gear up for the election, Republicans and Democrats are operating on the premise that turnout will be low and the outcome will be determined by partisan activists. Therefore – the median voter will not determine the outcome – if they are right.

Sources: Government Failure by Gordon Tullock, Arthur Seldon and Gordon Brady, others (see in text).

Vote-trading or in this case, Logrolling, usually occurs with representatives. Let's elaborate on this theory some more: (as per Tullock)

Implicit logrolling: measures that different politicians favor can be incorporated in one piece of legislation and a single vote taken on the bundle (“Tullock's bundling”).

Regardless of whether it is explicit or implicit, logrolling occurs because most laws have differential effects on different groups and parts of the country.

Logrolling can be “desirable.” If a project will benefit some city a lot and would have a relatively modest cost to the national taxpayer (again, assuming some hypothetical measure of value - and assuming that one believes redistribution of wealth via the government is moral).

But usually logrolling is talked about with respect to the harm that it creates.

Logrolling leads to almost all projects being voted through, including those whose net value to society is negative.

There can be so much logrolling (and so many bills passed) that even the “beneficiaries” of passed legislation (say, people in one state) may, on net, be worse off because of how much they have to pay in federal taxes to support the other legislation passed (that benefits other states).

Logrolling partly explains such public-sector programs as agricultural subsidies and other government spending that benefits a very small group at a large cost spread among national taxpayers.

Assuming the voter is very badly informed (rationally ignorant and maybe also rationally irrational) one could readily anticipate a great amount of inefficiency from logrolling. Because voters are normally particularly badly informed about legislation that affects them very little.

Some people regard vote trading as immoral. The reason being that when legislator’s vote trade their public behavior does not square with his or her private values. They vote for bills that they really don't value.

Does bundling all these projects together in one gigantic bill make it moral? After all, then they would be voting for something they actually agreed with.

This ignores the procedure by which the bundling takes place. Trades take place in committee before the final bill is ever developed. So the “immoral” act takes place prior to the bundling.

As noted, most voters are not informed, especially about projects that don’t affect them very much.

The concerns that voters hold intensely are far more important to the members of Congress than those about which voters have only minor feelings and certainly more important than those of which voters are not aware.

As we have discussed: Politicians as Vote Seekers: Politicians compete for votes by promising to work in the legislative body for programs that please a majority of the voters in their districts.

Therefore, most political battles are likely to take place over proposals that are of primary interest to small groups. Members of congress wishing to be reelected will take careful account of issues and bills that strongly affect small minorities, whether it is a reduction of transfers to them, an increase in the taxes specific to them, or a special tax exemption.

The special interest groups are more likely to vote – but also more likely to give the politician support, financial and otherwise, for election purposes.

Furthermore: the marginal benefit to politicians of attracting more votes declines as the number of votes they already have increases. Once they have enough votes to win an election, the benefits of attracting more votes is low. Why announce a new policy that might alienate existing voters?

This leads us back to the median voter model as it applies to representative government. In order to please as many groups as possible, or not alienate as many as possible – the politician will move towards the middle of the road, trying to please as many special interest groups as possible.

A representative is less likely to be middle of the road if his constituency is smaller and he has a few strong special interest groups that overlap in some way. For example: environmentalists and animal rights groups would overlap; property rights advocates and the right to bear arms groups.

What if two politicians espouse the same platform on some issue? What’s their strategy then?

1) “Go negative” and label the opponent “too liberal” or “too conservative” depending upon which voters the politician is trying to attract. Stake out the middle ground and represent the other as “too extreme.” Or go negative on a more personal level - dig up some dirt!

2) Choice of political party is important. It is like a “brand name.” So voters (being rationally ignorant or rationally irrational) might vote simply on party grounds. So the movement towards “independent” can be explained by this. Saying the other two parties are “extreme.” The problem is, a third party can usually only attract voters from the “tail” of the distribution. But if there are enough people who are “independent” then they can really make a difference in an election.

Why? Term limits are likely to reduce the frequency of logrolling, because a politician whose term is about to expire and who cannot get reelected has little incentive to cooperate with a logrolling effort.

If his optimal strategy is to cheat on his last vote, then other politicians may try to cheat him on his second-to-last vote, and this reasoning can be extrapolated indefinitely - often resulting in non-cooperation on present logrolling efforts.

2. If every voter (politician) is given a fixed number of votes to allocate among the issues that she sees fit, this would formalize the logrolling process and allow for a limited trading of votes - but only so long as individuals have votes remaining to trade.

This formalization would prevent indefinitely large voting blocs from forming, and it would also lead to a tradeoff between individuals investing all their votes into their preferred issue and trying to garner other votes by devoting their own votes to other issues.

Lobbying by special interest groups and PACs (Political Action Committees – group of private lobbyists) make public choice very difficult.

He pointed out that, when a relatively small number of people are heavily affected by a collective activity, organizing is in their interest.

1. Individuals in the group will either benefit a good deal if the political action is in their favor or be injured a good deal if it is against them.

2. Because there are only a few of them, transaction costs to organizing are relatively low.

On the other hand, if the legislation affects a large number of people but represents only a small amount to each of the group, the converse applies.

Let’s say that the proceeds of a tax of 5 cents levied on every citizen of Durango are to be given to this class for writing some ideas on passing or not passing a new housing development. We (the class) would be very interested in getting this bill passed – we would benefit a lot but each citizen would only pay 5 cents, so it wouldn’t be worth it for them to spend much time against it.

It might – but there is still a large probability it won’t as the transfer is written – a per head tax. So perhaps we need to figure out how to hide the tax (maybe call it a “fee” instead of a tax – tie it to the sale of something)?

Even though this might pass but how can we increase the probability that it does?

Another example: At one time in the U.S. there was a tariff to protect the manufacturers of chin rests for violins. Only one company employing four or five people made the chin rests. Each violin purchaser had to pay 3-4 cents more so they didn’t lobby against it. But the manufacturer did lobby for it.

What argument was given: potential unemployment of the four or five people he employed to do the chin rests.

A small tax on violins to provide a pension for the employees of the company would have failed because, although economically more efficient, it would have been too obvious.

Conclusion: there is a political aversion to direct transfer payments – and this means that protected projects are much less efficient than they would be otherwise. The cost to the people who incur the cost is higher and the return to the people who receive the benefit is lower - than if a simple transfer were politically feasible.

So a lot of resources are spent “hiding” taxes or making them less obvious in some way.

You will most likely hear a public interest argument from a lobbying group. Public interest arguments normally require that the project itself be designed in such a way that the direct transfer is hidden from the public eye.

An old example is the farm program in the U.S. Most citizens don’t have a clue they pay more for vegetables, etc. because of the farm program. The cost of the program far outweighs the benefits the special interest group receives (or at least most economists think so – again, they actually can’t measure the costs/benefits – but they see how much farmers gain and how much waste there is in the program and make the assumption).

The argument: Reduce the federal budget while making sure that the cuts fall predominantly on the projects of special interest beneficiaries. This would make everyone better off because, although almost everyone would lose some kind of special privilege, the cost of all of the special privileges held by others is greater than the benefit received from any one special privilege than an individual may have.

AGAIN - In other words -- people are actually paying MORE for their special project (because of the taxes they are paying for all others) than they are benefiting from it. But because the taxes are not obvious, they don’t know this.

Here’s the problem: although I would be better off if all special privileges were eliminated, I would be even better off if all were eliminated except those that benefited me specifically!

Hence we have the prisoner’s dilemma. Each voter will continue to vote for special privileges for themselves – if I don’t and everyone else does, I lose. If I do and everyone else doesn’t, I win.

Voter	MWTP for Project A	MWTP for Project B
Dick	$500	$2,000
Harry	2,000	500
Tom	700	700

	Voting Paradox
Mr 1	Mr 2	Mr 3
A	B	C
B	C	A
C	A	B