Prof. Bryan Caplan

Econ 410


Week 7: Empirical Accuracy of the Median Voter Model

I.             The Median Voter Model Versus U.S. Constitutional Structure

A.           At least to a rough approximation, U.S. political opinion is one-dimensional.  That dimension is liberal-conservative.

B.           Liberal-conservative preferences are probably almost always single-peaked.

C.           So the Median Voter Theorem looks highly relevant, right?

D.           But the issue is more complicated because most countries - especially the U.S. - are not run by simple majority rule based on one big election.

E.           Instead, the U.S. federal government has three branches, each with complicated election rules; and under the federal government there are state governments, each with complicated election rules of their own.

F.            Brief review:

1.            The U.S. president is whoever gets a majority of electoral votes.  In most cases, 100% of a state's electoral votes go to whichever candidate got the most votes in that state.  Electoral votes are not quite proportional to population.  (See attachment).

2.            Each state has 2 senators regardless of population.

3.            Each state has House representatives proportional to population.  These representatives are elected by districts into which each state is divided.

4.            Supreme Court justices are appointed by the president but must win 2/3 approval from the Senate.

G.           To become federal laws, bills need either:

1.            House majority, Senate majority, and Presidential cooperation (signature or failure to veto)

2.            2/3 House and 2/3 Senate

H.           Supreme Court can find laws "unconstitutional" with a majority vote.  Justices stay until they retire or die.  Congress has the power to change the number of justices, but this hasn't happened for about 100 years.  (One attempt in the 1930's flopped badly).

I.             The Median Voter Theorem, taken strictly, does not predict that the preference of the median U.S. voter will prevail under these circumstances.  Instead:

1.            The U.S. president will cater to the median electoral vote.

2.            Senators will cater to the median voter in their state.

3.            Representatives will cater to the median voter in their district.

4.            Supreme Court's activity depends on the median justice, who was a compromise between the President and the Senate at the time of his appointment (possibly 50 or even 60 years ago!).

II.            The Case for Simple Models: Do Constitutions Really Matter?

A.           Strictly speaking, then, figuring out the exact prediction of the MVT in the U.S. political system is really complicated.

B.           Perhaps, however, all of these constitutional rules are just "bells and whistles" that cancel each other out.  Then the political system would act "as if" it were run by the median voter.

C.           Is this crazy?  Briefly surveying political systems around the world, it is rarely easy to show that policy differences stem from constitutional differences.  Variations in voter preferences are, in contrast, obvious.

1.            Ex: France versus the U.S.

D.           Even dictatorships frequently pay attention to public opinion.

E.           When a political minority manages to foil majority wishes by cleverly using constitutional rules, those rules frequently change.

F.            Moreover, as a practical matter, starting with simple models is often a more productive way to analyze real-world problems, even if the simple models are less realistic.

G.           So: Is the behavior of the U.S. federal government consistent with a simple median voter story?  If not, how so?

III.          U.S. Government Policy: An Overview

A.           There are widespread misconceptions about the numbers on taxation and spending.  Let's start with some basic facts. 

B.           For the federal budget in 2015, expenditures are comprised of roughly:



Social Security




Domestic Discretionary




Net Interest


Income Security




Other Retirement/Disability




Offsetting receipts


C.           Main facts to note: payment for the old add up to 41% of the budget, over twice spending on defense.  Payments for the poor come out to something like 18%.

D.           For the federal budget in 2015, revenues are comprised of roughly:



Individual Income Taxes


Payroll Taxes


Corporate Income Taxes


Excise Taxes/Customs




E.           Main facts to note: most taxes come from the items you see listed on your paycheck - income taxes, social security taxes, and Medicare-type taxes.

F.            There’s also lots of regulation: Environmental, worker safety, drug safety, anti-competitive behavior, labor...

G.           Is all this what the median voter wants?

IV.          Does Policy Match Public Opinion?  What Are the Unpopular Policies?

A.           Starting with the budget: Social Security and Medicare remain extremely popular programs; the military is also usually well-regarded.  The remaining items are more contentious. 

B.           Broadly defining "welfare" as Medicaid and "Other Means-Tested" spending, we get 13% of the budget.  But:

1.            Few people want to actually abolish these programs

2.            Medicaid also pays for middle-class nursing home residents who have run through their personal savings.

C.           The national debt is unpopular, but repudiating it would be even less popular.  So "net interest" ultimately has voter support.

D.           That leaves 25% of the budget for "non-defense discretionary" and "other" spending.  Some of this spending is "waste.  Waste is unpopular.  But outside of isolated examples of $500 toilet seats, what spending do a majority of Americans agree is wasteful?

E.           Turning to spending: It is surprising that income and SS taxes are such a large percentage of the budget.  But insofar as business "passes on" corporate and other taxes, do a majority of Americans really want significant changes here?

F.            Regulation is more complicated.  Are there majorities in favor of weaker (or stronger) environmental regulation?  Worker safety?  Drugs?

G.           Challenge: What policies exist that a majority of American voters oppose?  Consider all the clichés of politics.  Do any hold water?

1.            Relatively weak gun control?

2.            Foreign aid?

3.            NAFTA?

V.           Application: State-Level Policy

A.           There have been a number of empirical studies of state-level policy.

B.           Main findings: Variations in degree of liberalism are strong predictors of variation in state policy.  When public opinion is liberal (as in NY), policy is liberal; when public opinion is conservative (as in Colorado), so is policy.

1.            It is hard to convincingly show that public opinion and policy match each other 1:1, but the evidence is suggestive.

C.           Median vote story: In spite of the complexities of state constitutions, state-level democracy more-or-less goes along with the preferences of each state's median voter.

VI.          Adjusting for Bargaining

A.           As we have seen before: If people first vote, then bargain, policy need not match the median voter's preferences.  Minorities with intense preferences can "bribe" the majority.

B.           Recall the Mean Voter Theorem: With fully efficient bargaining, policies conform to mean voter preferences issue-by-issue - regardless of constitutional rules!

C.           Does this shed any light on the survival of apparently unpopular policies?  Are minorities implicitly bribing majorities?

D.           Ex: Freedom of religion?