Economics 849 Midterm

Prof. Bryan Caplan

Fall, 2005

Part 1: True, False, and Explain

(10 points each - 2 for the right answer, and 8 for the explanation)

State whether each of the following six propositions is true or false.  In 2-3 sentences (and clearly-labeled diagrams, when helpful), explain why.

1.  Suppose you receive a phone call from the General Social Survey asking whether you favor or oppose using quarantines to prevent bird flu.  Counting yourself, there are 2001 people in this survey.  Each of the other respondents has a 51% chance of supporting the quarantine and a 49% chance of opposing it.  There is an 10% chance that politicians will do whatever the majority of poll respondents want; otherwise they ignore the poll.  You only respond to the poll if you have at least a .1% chance of changing policy.

T, F, and Explain:  You will respond to the poll.

TRUE.  The probability that your response changes policy is .

2. T, F, and Explain:  The existence of negative externalities is always inefficient.

FALSE.  Negative externalities are only inefficient if the SB of their existence exceeds the SC.  The externalities that remain after the imposition of an optimal tax are efficient; so are the externalities that remain after optimal Coasean bargaining.  In fact, the existence of infra-marginal externalities is efficient even if nothing is done about them.

3. The U.S. produces almost all of its own gasoline, and national production is highly inelastic.  Gas taxes are included in the price of gas.

T, F, and Explain:  If Virginia cuts the state gas tax, the price of gas in Virginia will not noticeably change.

FALSE.  This would be true for a national tax cut; with highly inelastic supply, the (price + tax) would have to remain roughly constant to clear the market.  However, it is not true for a state tax cut, because Virginia's tax cut would siphon off gas from other states.  This illustrates the general rule that supply and demand are more elastic for a sub-market than for the overall market.

I gave partial credit to students to observed that it demand were also highly inelastic, suppliers and demanders would share the tax cut.

4.  Delli Carpini and Keeter show (Table 4.1) that the variables that affect political knowledge usually work across the board.  For example, education predicts greater knowledge in all six of the categories they study.

T, F, and Explain:  The same is true for gender; women get lower scores in all six categories of political knowledge.

TRUE.  The coefficient on gender is negative for all six categories, although it is not statistically significant for the gender sub-category.  (I gave full credit to students who gave this answer but said FALSE on the grounds that the statistically insignificant coefficient could easily equal 0).

5.  T, F, and Explain:  Log-rolling and the “Miracle of Aggregation” are two solutions to the problem of social intransitivity.

FALSE.  Log-rolling is a solution to social intransitivity, because bargaining allows people to express preference intensities/willingness to pay, and willingness to pay is always socially transitive.  But the Miracle of Aggregation has nothing to do with this problem.  It only says that the preferences of the well-informed guide policy.  It is quite possible that the preferences of this sub-group are socially intransitive.

6.  Suppose law is determined by a single legislative house, but the interpretation of the law is decided by one Supreme Judge.  Legislators know that the Supreme Judge is more liberal than the median legislator.

T, F, and Explain:  Equilibrium policy might actually become more liberal if the Supreme Judge resigned and let the legislature appoint a replacement.

TRUE.  As explained in lecture, if legislators KNOW that judges will misinterpret their policies in a liberal direction, legislators will respond either by refusing to pass legislation, or by passing legislation that is more conservative than the legislators actually want.  If the Supreme Judge resigned, the legislators would replace him with a Judge who shared their median preference, and could then approve new legislation without worrying that it will be "highjacked" by the judiciary.

I also gave full credit to students who observed that with only one legislative house, judges in the Cooter model actually have no discretion.  This is correct unless there are transactions costs of passing "clarifying" legislation.

Part 2: Short Essays

(20 points each)

In 6-8 sentences, answer all of the following questions.

1.  Suppose the number of murders is completely insensitive to criminal punishment.  Could it still be Kaldor-Hicks efficient to punish criminals?  Explain your answer.

[The second sentence of this question should have read "murderers," not criminals, and most students answered it that way.  Students could get full credit for either answering the intended or the literal question.]

Reading the question literally: Even if murder were insensitive to punishment, it does not follow that all crime is insensitive to punishment, so there could still be valuable deterrence benefits of punishment.  Furthermore, even if murder were insensitive to punishment, it does not follow that the type of murder is insensitive to punishment, so there could still be some efficient gains.  Finally, even if no aspect of murderers' behavior changed due to punishment, the public could still enjoy direct revenge benefits.  If the public gets \$10,000,000 in satisfaction from having a criminal executed, and the criminal would pay only \$1,000,000 to save himself, then execution is the Kaldor-Hicks efficient decision.

2.  Suppose a third party (e.g. the Greens or Libertarians) hires you as a voting theory consultant.  The party wants strategic advice on the best way to push policy in its preferred direction.  What would you recommend?  Explain your reasoning.

There are several routes to take.  One important one is to offer to support the party closer to yourself IF they take more extreme positions.  Dropping out irrevocably shifts the median position AWAY from your own; dropping out conditionally pressures the party closer to you to become more extreme – albeit at higher risk of losing altogether.  Another strategy would be for the extremists to donate money to the major party they dislike the least in exchange for more extreme policies.

Other ideas:  Third parties could devote their energies to changing the position of the median voter through education, campaigning, etc.  Or, they could donate all their money to the OTHER third party in order to help them siphon off voters from the major party!  This would pull the median voter in the desired direction.

3.  Suppose local governments acted like perfectly competitive firms.  Which of the rationales for redistribution discussed in class, if any, would be most likely to guide local policy?  Explain.

With perfect competition, most forms of redistribution will be impossible, because the losers could costlessly switch their location to escape the burden.  Redistribution as investment or insurance would fail unless the local government were as efficient as the market's investment and insurance options.  Even redistribution to correct externalities has a serious problem: Given perfect competition, paying young, able-bodied males not to commit crimes attracts infinitely more young able-bodied males to your locality.

However, there are at least two exceptions to this rule.

First, taxing people who create negative externalities and redistributing to people who create positive externalities would work.  Instead of giving welfare to the homeless, you could tax them extra.  They could threaten to leave, but in equilibrium other localities would want to tax them as well.  This is no different from a nightclub making men pay a higher cover charge.

Second, you could redistribute more flexibly if you imposed residency requirements.  People might prefer to live in an area where the local government helps the homeless, but not if this just attracts more homeless people.  Limiting public assistance to people who have lived in your locality for a couple of years largely solves this problem.