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. State enterprises often lose money year after year, and need subsidies to stay in business.
T, F, and Explain: If the reason for the losses is that state employees are overpaid relative to their abilities, the subsidies are Kaldor-Hicks inefficient.
FALSE. Over-paying people relative to their abilities is just a transfer. Thus, suppose you would have broken even paying people their abilities. Efficiency-wise, this is no different from over-paying by $1000 and suffering $1000 in losses.
2. Suppose that the electorate is evenly divided (p=.5) and there are 100,000,000 voters other than yourself.
T, F, and Explain: It is instrumentally rational to vote if the victory of your preferred candidate will bring you $13,000 more in present value than the victory of his opponent.
[I should have given you a MC of the time to vote, but given the way the numbers worked out, this had no effect on anyone's answer.]
FALSE. Using the probability of decisiveness formula, p=, which simplifies to p=, or 1-in-12,533. p*$13,000 is therefore $1.03, well below the value of anyone's voting time.
3. Suppose that a malevolent government wants to maximize the deadweight loss of taxation subject to a maximum revenue constraint. It is only able to tax two markets, both with linear supply and demand curves.
T, F, and Explain: To achieve this end, the government must use the Ramsey rule, taxing both goods at a rate inversely proportional to their elasticities.
FALSE. The Ramsey rule minimizes deadweight loss. To maximize it, you must raise 100% of the tax from the more elastic good. If they are equally elastic, you should still put 100% of the burden on one or the other.
4. Suppose that the Median Voter Theorem is true.
T, F, and Explain: Because the MVT implies platform convergence, it is impossible for any one vote to make a difference, regardless of the size of the electorate.
FALSE. It is still possible to change outcomes by changing the median position! Simply jump from one side of the current median to the other. This is particularly clear with small electorates: if voter bliss points for spending are $0, $100, and $1000, the person with a bliss point of $0 can change the outcome by changing his bliss point to $200, raising the equilibrium platform from $100 to $200.
[Most answers argued that when p is small, voters are more likely to be decisive. Correct, but if platforms converge, even a decisive voter seemingly has no effect - they give the election to a different politician who does exactly the same thing.]
T, F, and Explain: The more strictly this clause is interpreted, the greater the effect of voter ignorance.
TRUE. The theory of optimal punishment shows that you can adjust for voter ignorance by adjusting punishment by a probability multiplier. If there is a cap on permissible punishment, then it may be impossible to control agents.
6. Local and state governments are often accused of being much more corrupt and receptive to rent-seeking than the federal government.
T, F, and Explain: This is exactly what the Tiebout model predicts.
FALSE. The Tiebout model predicts that local and state government, facing competition from other localities and other states, will be tightly constrained to do what voters want. Otherwise, citizens will leave corrupt jurisdictions in favor of honest ones. The federal government faces much less pressure of this sort.
Part 2: Short Essays
(20 points each)
In 6-8 sentences, answer all of the following questions.
1. Give an example where doing nothing to correct a genuine public good problem is second-best efficient. Defend your answer.
Preventing people's feelings from being hurt is a genuine public good. A first-best policy on free speech, then, would impose Pigovian taxes on hurtful speech, so that people only say mean things if they value saying them more than people disvalue hearing them. In practice, though, this would be an incredibly burdensome system. Getting the taxes right would be almost impossible; who knows how much anyone really suffers from hearing mean things or gains from saying them? Enforcement would be very costly, and the end result would probably be that all sorts of socially valuable speech is curtailed. It is therefore plausible that the do-nothing policy of free speech is a second-best efficient policy.
2. The Miracle of Aggregation shows that politicians have to win the support of a majority of informed voters to win. But empirically, as Delli Carpini and Keeter find, voter knowledge is not randomly distributed. Assuming selfish voters, discuss one empirically important way that voter ignorance would change equilibrium policy.
Delli Carpini and Keeter show that males are more politically informed. Thus, the median informed voter is male, even though the median voter is female. If males vote selfishly, then, politicians will cater to males, and eschew sexual harassment laws, gender discrimination laws, and other measures that act against male interests. Less-informed voters, disproportionately female, will not know enough about politicians' voting records to punish them for their pro-male stance.
3. The logic of the
Ramsey Rule applies to redistribution as well as taxation. If you want to give away a fixed amount of
income at the minimum deadweight loss, you have to give more money to groups
who change their behavior less, and less money to groups who change their
behavior more. Could the relative
generosity of the
would argue that
[Many answers focused on the elasticities of the products different groups were likely to buy.
But recall that the precise benchmark for inefficiency is change in behavior relative to a lump-sum transfer. Poor people will spend more on inelastically demanded items, but they would have done the same with a lump-sum transfer.]