Econ 1000 - Unit 9

Market Failure

A situation in which the free market outcome does not maximize total social welfare. It is the usual justification for government intervention.

Government Failure

A situation in which total social surplus is decreased by government intervention in a market. Even when a market failure exists, it is naive to assume government intervention will automatically leave society better off (with a smaller deadweight loss).

Two Broad Causes of Government Failure

(1) The government fails to perform a necessary task efficiently. (2) The government fails to do only the tasks it should do, meaning it does more than it ought to.

Public Choice

The academic subfield that uses the tools and framework of economics to analyze issues that historically belonged to political science, such as voter behavior and how governments make decisions.

Economic Model of Voting: Setup

Three candidates (Donald, Joe, and Robert) seek an office, and voters have different ordinal preferences over them. With 3 options there are 6 possible voter types (distinct rankings). The model assumes everyone votes sincerely, according to their true preferences.

Voter Types in the Model

[i] R>J>D (20%), [ii] J>R>D (22%), [iii] R>D>J (6%), [iv] J>D>R (14%), [v] D>J>R (10%), [vi] D>R>J (28%). The letters stand for each voter's 1st, 2nd, and 3rd choices.

Voting Process 1: Plurality (Most Votes Wins)

All 3 candidates are on the ballot and the one with the most votes wins. D gets 38% (types v and vi), J gets 36% (types ii and iv), and R gets 26% (types i and iii), so Donald wins.

Problem with Plurality Voting

Donald wins with the most first-place votes, but he is also the person ranked last by the largest number of people (types i and ii, 42% of voters). So the outcome does not clearly reflect the "will of the people."

Voting Process 2: Runoff

All 3 candidates are on the ballot. If nobody gets 50%+, the top two vote getters have a runoff. D (38%) and J (36%) advance. In the runoff, J gets 56% to D's 44%, so Joe wins, a different winner than under plurality.

Voting Process 3: Pairwise Elections (Primary, Then General)

A primary between two candidates is followed by a general election against the third, using majority rule. Primary D vs. R, then general vs. J: D wins the primary, J wins the general, so J is elected. Primary J vs. R, then general vs. D: R wins the primary, D wins the general, so D is elected. Primary D vs. J, then general vs. R: J wins the primary, R wins the general, so R is elected.

Lesson of the Voting Examples

Each seemingly reasonable voting mechanism produces a different winner. This means the social choice is highly dependent on the voting rules, not just on voter preferences.

Voting Process 4: Beat Both Rivals in a Row

A series of pairwise elections where the winner is the first candidate to defeat both rivals in a row. The pairwise results are J>D, D>R, and R>J, so the process cycles forever and never selects a winner.

Pairwise Majority Results in the Model

Pairwise majority votes: J beats D (56% to 44%), D beats R (52% to 48%), and R beats J (54% to 46%). Each candidate beats one rival and loses to the other, creating a cycle.

Condorcet Paradox

A situation in which a series of pairwise majority votes over more than two options leads to a cycling of winners. First formalized in the 18th century by Nicolas de Condorcet. When it occurs, it reveals that social preferences are irrational (not clearly defined) because they are non-transitive.

Non-Transitive Preferences

Preferences that are internally inconsistent or contradictory. If A is preferred to B and B is preferred to C, transitivity would require A to be preferred to C. The Condorcet Paradox shows groups can violate this even though each individual voter is consistent.

Wafer Cookie Example (Transitivity)

If you prefer chocolate over vanilla and vanilla over strawberry, you should reasonably prefer chocolate over strawberry. Preferring strawberry over chocolate would seem irrational. Group voting can produce exactly this kind of inconsistency (J>D, D>R, R>J).

"Will of the People" and the Condorcet Paradox

When group preferences are non-transitive, the "will of the people" is a meaningless concept, so any attempt to find a voting system that reveals it is futile.

Seven Sources of Government Failure

(1) Informational problems, (2) cost of complying with government bureaucracy, (3) corruption and kleptocracy, (4) regulatory capture, (5) rent-seeking, (6) logrolling and rational ignorance, (7) deadweight loss from taxes.

Source 1: Informational Problems

When setting policy, government must choose a "scope" for its activity (for example, how much national defense to provide; the U.S. spent $863.0 billion on defense in 2024, and no one can easily say whether that is too little, just right, or too much). In markets, buyers and sellers have incentives to reveal their true benefits and costs. Outside markets they don't, so even well-intentioned bureaucrats can be misled when citizens misrepresent true benefits and costs.

Economic Calculation Problem

The argument that a system of planning can never achieve efficient outcomes, because planners lack the information generated by market activity. First made by Ludwig von Mises and later refined by Friedrich von Hayek.

Source 2: Cost of Complying with Bureaucracy

When governments impose regulations, market participants must spend resources to comply. The bigger the bureaucracy, the higher the compliance costs, and if they are very high they can outweigh any benefits of the regulation.

Permit Raj (India)

An example of excessive bureaucracy: a complex, irrational, almost incomprehensible system of controls and licenses under which everything needed approval.

Cost of Starting a Business (Ease of Doing Business)

Venezuela required 20 procedures and 230+ days at a cost of 211.8% of annual per capita income. New Zealand required 1 procedure and half a day at a cost of 0.2% of annual per capita income. This shows how heavy bureaucracy raises costs.

Source 3: Corruption

An environment in which regulations are not enforced and decisions are not made evenly and without bias. It leads to inefficient decisions. For example, in awarding a bridge contract, Contractor A has costs of $18M and Contractor B has costs of $21M, but B may win by paying bribes, so the less efficient firm gets the job.

Kleptocracy

An environment of extreme corruption in which government officials unabashedly seek personal gain at the expense of the public interest. The slides' example is present-day Russia, where a small group of people wholly above the law became rich beyond the wildest dreams of the tsars.

Source 4: Regulatory Capture (Setup)

Governments often regulate markets where producers have market power or where information is lacking, such as FDA approval of prescription drugs in the U.S. Ideally the agency would do a thoughtful cost-benefit analysis in the public's best interest, but the regulated industry has a strong incentive to influence the agency's decisions.

Regulatory Capture

A situation in which firms in a regulated industry influence a regulatory agency to the point where it enacts policies that serve the regulated firms' interests, even if they are not in the public's best interest. Regulators often identify more with the industry than the public, perhaps because people move back and forth between industry and regulatory jobs.

Source 5: Rent-Seeking

Attempts by people to manipulate government action or influence government decisions to make themselves better off at the expense of others. Resources spent securing these "rents" (surpluses) are essentially wasted.

Rent-Seeking Example: U.S. Peanut Trade Restrictions

Restrictions on imported peanuts greatly benefit U.S. peanut farm owners and workers, so they have a strong incentive to lobby to keep them. The restrictions are costly to consumers, foreign peanut producers, and U.S. producers of any goods that use peanuts as an input.

Rent-Seeking in Winner-Take-All Decisions

In a winner-take-all decision, like choosing where to host the 2036 Summer Olympics, competing groups spend huge resources lobbying to win. All the resources spent by the losers, and much of what the winner spends, are essentially wasted.

Estimated Impact of the U.S. Peanut Quota

About 397 jobs saved at a cost of $187,223 per job saved. Total costs were $74,372,531, or about $0.23 per U.S. resident. The benefits are concentrated on a few people while the costs are spread thinly over everyone.

Source 6: Logrolling

The process by which a legislator votes to approve one bill in exchange for favorable votes from other members on other bills. It is not necessarily bad, since it can let legislators express the intensity of their preferences, but it can produce inefficient outcomes when voters are rationally ignorant.

Rational Ignorance

Since becoming informed about public policy has high costs and low benefits for an individual voter, it is rational for voters to remain uninformed.

Logrolling and Rational Ignorance Together

Georgia voters learn about and recognize the local benefits of peanut protection but stay rationally ignorant of the costs of all the other trade protections that legislators from other states win by trading votes (such as machine tools, canned tuna, softwood lumber, and dairy). No single proposal would get majority support alone, but with vote trading and ignorant voters, all of them pass, which is clearly inefficient.

James Buchanan

Economist (1919-2013; Nobel Prize in 1986) who argued that politicians in a democracy often act to maximize their chances of reelection. This explains why a Georgia representative logrolls to pass peanut protections even if it means passing other inefficient programs.

Source 7: Deadweight Loss from Taxes

Government spending must be financed somehow, commonly by taxes. Almost all taxes distort costs and benefits and so influence economic choices, often inefficiently. A per unit tax requires buyers or sellers to pay a set amount to the government on every unit traded (for example, Georgia's gasoline taxes of $0.545 per gallon).

Tax Incidence

A measure of who bears the burden of a tax in terms of decreased welfare. It shows how much of the loss falls on buyers versus sellers.

Tax Example Setup

A market where a per unit tax of $T = $1 will be imposed. Before the tax, 5,000 units are traded at a price of $3.80.

Per Unit Tax on Buyers: Effect on Demand

At the point of sale, buyers are now willing to pay exactly $T = $1 less than before, so the demand curve effectively shifts down by $1.

Per Unit Tax on Buyers: Outcome

The new outcome is where the shifted demand curve meets supply. Only 4,200 units are traded. The price at the point of sale is $3.20, so sellers receive $3.20, and buyers pay the $1 tax on top, for a total price paid of $4.20.

Per Unit Tax: Welfare Effects

Consumer surplus falls by a+b+c (incidence on buyers). Producer surplus falls by d+e+f (incidence on sellers). Government tax revenue is a+b+d+e, which equals $1 x 4,200 = $4,200, less than $5,000 because tax is collected only on units still traded. The deadweight loss is c+f.

Why a Per Unit Tax Creates Deadweight Loss

The tax drives a wedge between what buyers pay and what sellers receive, so trade falls below the efficient quantity (from 5,000 to 4,200 units). The trades that no longer happen (areas c and f) would have benefited both sides, and that lost surplus is the deadweight loss.

Per Unit Tax on Sellers: Effect on Supply

At the point of sale, sellers must now be paid exactly $T = $1 more than before, so the supply curve effectively shifts up by $1.

Per Unit Tax on Sellers: Outcome

The outcome is where the shifted supply curve meets demand: 4,200 units are traded, sellers get $3.20, and buyers pay $4.20. The decrease in CS, decrease in PS, tax revenue, and DWL are all the same as when the tax was imposed on buyers.

Equivalence of Taxing Buyers vs. Sellers

It makes no difference whatsoever whether a per unit tax is legally imposed on buyers or on sellers. The quantity traded, the price buyers pay, the price sellers receive, the welfare losses, the revenue, and the DWL are all identi