Bryan Caplan  

Ron Paul vs. the Median Voter Model

PRINT
Thoughts on Spain... Arbitrary Intervention...
I'm already devouring Brian Doherty's Ron Paul's Revolution.  Brian's such a great writer and such a careful historian - a true treasure.  To me, though, the biggest puzzle isn't how Ron Paul can command the loyalty of 5% of the U.S. population, but how Paul manages to repeatedly win re-election in a seemingly normal congressional district.

The Median Voter Theorem usually works pretty well in the real world.  Why does it fail so badly for Ron Paul?  Please show your work.



COMMENTS (33 to date)
Chris writes:

Ron Paul is closer to the median voter in his district than any of the Democrats that have run against him. So, the only thing he needs to do is win the GOP primary.

Because turnout is so low in primaries (around 30% for both parties), you at most need to get 15% of your district's support to get your party's nomination -- and that's assuming you need a majority rather than merely a plurality against a fragmented opposition.

In a low-turnout environment, it's easy to get 15% support if you're long-established incumbent with money, organization, name recognition, and a track record, and your opponent is starting from scratch.

Piotr Pieniążek writes:

I think that at the beginning of his career he had a beginners luck, but now he has a comparative advantage of long and conspicuously honest track record, so that his views (so radical from the mainstream perspective) given the above aren't decisive even for the median voter.

Martin writes:

Look at the wiki-link: Ron Paul out-raised his opponents two to one or more.

Let's modify the median voter model, let's say that the candidate can choose to spend funds and to choose his or her position to win an election. The sole objective of the candidate is winning and there is some cost - to the candidate - for changing position; both the position and funding are exogenous.

By choosing a position, the candidate, comes closer to some candidates at the price of distancing himself or herself from other candidates.

By spending money, the candidate can move certain voters closer without distancing others.

A candidate who has more money is more likely to stay where he or she is and is more likely to win the election than the same candidate without.

I'd say that more funds are a pretty good explanation.

Greg G writes:

Let's see, we have a congressional incumbent who is very likable, who gets a lot more publicity than his opponents and who raises a lot more money than his opponents. Somehow he gets re-elected regularly.

And we have, in the same person, a Presidential primary candidate who loses badly despite being popular in his home district.

This is an anomaly that requires a special theory to understand it?

Ken B writes:

British parliamentary history has many examples of 'oddball' members being returned. Sometimes the voters in a district will vote to keep an oddball just because they know he is one, and value the idea.

Socialist senator Bernie Sanders kinda defies the median voter theorem too.

david writes:

He raises money outside the district to campaign within it, diverting non-median support elsewhere into one area.

Tom writes:

A couple points.

First, Paul's most, shall I say, outre policy positions are simply unimportant and irrelevant as an individual congressman. A desire to abolish the Federal Reserve as a congressman is about as relevant as a desire to learn how to fly. Paul caucuses as a Republican and doesn't vote Democrat. That's more important. "Dr. No" on 420-1 votes are less important than party-line "yes" on 223-210 votes, and if your voters value reputational independence, "Dr. No" votes can be positive.

Second, everything I've read is that Paul's office is very good at constituent services. Congressional districts are small enough that people are able to contact their congressman and expect action, and small enough that if the congressman is responsive enough, his responsiveness becomes well-known in the community. This isn't particularly different from Paul's pre-congress service as an ob/gyn.

Disclaimer: family who lived in Paul's district, one of whom wondered idly about whether Paul's outre positions might make him vulnerable to somebody more straight-line R.

Kevin writes:

I think there are two non-exclusive ways to attack this question...

1) Ron Paul narrowly won a primary against a candidate who had recently switched parties in 1996 and since has, and this is just my impression from the wiki page you linked to, faced Democratic challengers who appeal less and less to the median voter of the district. Maybe over-time being the district that elects 'the crazy libertarian guy' becomes it's own form of signal upon which the median voter makes their votes.

2) Maybe the question shouldn't be whether the median voter of Texas' 14th district is different from other median voters in other districts, but whether Paul is really different (in practice) from the median politician in congress from the perspective of his constituents. Suppose we take a look at earmark spending by district in Texas (link at bottom). How different from the median politician is Ron Paul in practice vs. Ron Paul in rhetoric?

http://endingspending.com/fixing-the-budget/earmarks/map-2011/#rep:410/state:TX

Alex Godofsky writes:

In addition to all of the posts above, the median voter theorem needn't say that such a result always happens, only that it usually happens.

Tom writes:

He delivered most of the voting age children of the district. He's well known in the Religious and charitable communities.

I suspect he's popular on a personal level, like a small town mayor.

Collin writes:

Three reasons:

1) Any reasonable Republican would have anyway and all he had to do was win the primary. He is an excellent speaker so he has an advantage.

2) Libertarians, Ron Paul, do an awful job of winning support from Democrats. They spend time complaining about licensing signaling instead explaining in terms of simple kitchen table issues. Tell people our health care system keeps birth control pills behind the counter and cost you (or insurance company) money.

3) He is way too obessessed with the Federal Reserve and should focus more on wars as the root of big government. I think blaming all financial failures on the Fed, gets very old. If we could cut military by $200B then I bet lots of Dems would cut other services $300B.

Collin writes:

Three reasons:

1) Any reasonable Republican would have anyway and all he had to do was win the primary. He is an excellent speaker so he has an advantage.

2) Libertarians, Ron Paul, do an awful job of winning support from Democrats. They spend time complaining about licensing signaling instead explaining in terms of simple kitchen table issues. Tell people our health care system keeps birth control pills behind the counter and cost you (or insurance company) money.

3) He is way too obessessed with the Federal Reserve and should focus more on wars as the root of big government. I think blaming all financial failures on the Fed, gets very old. If we could cut military by $200B then I bet lots of Dems would cut other services $300B.

Floccina writes:

Maybe people in his district think:
He is a likable guy and has only one vote 435 in the house so why not vote for him, what harm can he do.

BZ writes:

Tom is right. I spoke with a lady at a Ron Paul fundraiser in Texas City many years ago and asked her why he was so popular here. Her reply: "The man delivered all our sons and daughters!". I got a chuckle out of that.

happyjuggler0 writes:

Some possibilities:

1) The Median Voter Theorum may be wrong. Inconceivable!

2) Does the Median Voter Theorum predict 100% success? There are 535 members of congress....

3) Incumbents have a massive edge getting elected due to tons of free publicity (aka media reporting) as well as perennially entrenched supporters.

4) Ron Paul is anomalously not a lying politician. Many people find that refreshing.

5) What others said.

Mentha Trecenta writes:

isn't that normal for a celebrity politician in a small district? all german major politicians (Merkel, Lafontaine) also have their small districts, from which they are elected again and again directly to parliament. i once read an article about an obscure social-democrat, who wanted to defeat Merkel in her Stralsund district, hopelessly, needless to say. i got the impression that people are simply flattered by a celebrity courting them, regardless of what their general views on federal policy may be. the median-voter-theorem makes no predictions about preferences, so it may well be that a celebrity could shift the preferences of the electorate in a particular district, without affecting national preferences at all, due to the lack of the intimate celebrity effect nationally. QED ;)

Jonathan Bechtel writes:

1). He has a small district and is well connected in religious and charitable circles and has delivered lots of babies over 30 years. I believe his district is very small, and works more like a small town.

2). His outlier opinions on national issues do little to effect the opinions of voters who care about local issues.

3). Concerning local issues, I'd guess he's pretty responsive. Responds to voter requests, organizes different services, etc.

4). His outlier opinions make him very popular with particular groups, and thus the beneficiary of lots of funding, which allow him to outspend the opposition on a local level

5). He comes from an old fashioned district, and is closer to the median voter than the typical democrat

6). "incumbent momentum"...once you're there long enough people like to keep you there out of habit, familiarity, and perhaps a sense of pride.

7). The median voter theory is an average, and is not necessarily a good predictor for a specific district, where the political conditions are more heterogeneous

8). Ron Paul has always made aggressive use of earmarks....his libertarian, laissez faire scruples have never applied locally the same he they do federally

9). The median voter model actually holds up well with Ron Paul...it's just the mechanics of his district are bit different than usual.

Assuming people vote based on issues that are easily visible to them and allow group affiliation to heavily influence their opinions, Ron Paul isn't an anomaly at all.

A good ol' boy who's very culturally connected and aggressively provides public funds for different works in a responsive manner re-inforces the theory, not refute it.

A lot of these may have some validity.

The "small district" theory, however, can't be correct. All Districts are roughly equal in population.

Perhaps overlooked is that Ron Paul's launch in the 1970's was when Goldwater-Reaganesque values held more sway in the GOP than they do now, and Paul definitely aligned with that camp and supported Reagan in 1976. That probably helped propel him to office the first time, when his views were more commonplace in the GOP.

After his break from office, he would have had, in 1996, some name recognition and previous experience to bank on. After that, there are the advantages of incumbency offered by earlier commenters.

Jeff writes:

Some good ideas in the previous comments.

I'm not 100% certain this is true, but I'll throw this out there anyway, and if someone wants to refute it, have at it: there are fewer median voters in his district than you might think. Texas' (albeit brief) history as an independent republic left a mark on the culture of native Texans that survives to this day in the form of a nationalist streak as Texans that you won't find in say, Ohioans or Floridians, for obvious reasons. This nationalist streak leads voters there to view Washington with a more suspicious eye than the median voter. Paul's anti-beltway rhetoric merely plays to the nationalist impulses (remember the Alamo!) that many Texans are brought up with.

It's not that voters there are really more libertarian-leaning than voters elsewhere, but enough of them still view us Yanks as enough of an "other" in the tribalist sense to nudge them in a more libertarian direction.

Vake writes:

As an obstetrician-gynecologist with 4000 births under his belt, he probably delivered half the voters in his district.

Bryan Willman writes:

This may be a case like single elimination playoffs - they do NOT assure that #1 meets #2 in the finals.

So it could very easily be that the 10 democrats in the district pick a "pure" democrat over the electable one, and the "pure" democrat would get shellacked by any reasonable republican.

What's more, I've only seen one or two truly odd ideas attributed to Ron Paul. Some of his unusual ideas (like, let's not spend horrid numbers of lives and staggering sums of money to achieve basically nothing fighting pointless wars) will appeal to both the left and many elements of the right.

("Strong defense" and "squander people and money in Iraq" don't actually go together.)

Finally, what I've seen of him, he comes off as sane and smart. Might disagree, but he's neither stupid nor mentally ill.

In WA we had one recent election where John Candidate was opposed by a person whose legal name was, really, "GoodSpace Guy". I'm sure that most any "not observably mentally ill" candidate can defeat "GoodSpace Guy".

Have the democrats been running "GoodSpace Guy of Texas" or "Nobody" (real name) or "Nice Guy" aginst Ron Paul?

Bryan Willman writes:

Further, if the wiki article about the median voter theorem is correct, then the theorem is basically useless in the real world.

Political views expressed on one dimension? There's only one real issue at all in the entire election? For a federal post? Surely you jest?

R. Pointer writes:

Bryan, it can be extended into a multi-dimensional policy space with little change in predicted outcomes. See Veto Players by G. Tsebelis.

My take is the dominant issue for voters is status quo. That is why most Congressmen get reelected at such high rates. It takes a pretty big calamity for citizens to get upset with the status quo and most politicians with a voting record/constituency work can position themselves properly for the next election.

So because he is an incumbent he has an automatic advantage that allows him to associate himself with the SQ.

@Vake and others: I don't know the exact size of the district, but ALL Congressional districts run into the several hundreds of thousands. If parents have fond memories of dealing with Dr. Paul, that can steer some of the vote overall. But Paul couldn't have "delivered" more than a small percentage of voters in the District.

Lewis writes:

He represents the median position for the core of his single issue voters. No illegal immigration. Fiscally conservative. Good family values.

Also important to note until 4 years ago he was rather obscure. Name recognition kicks in at some point with a R next to his name.

Bob Knaus writes:

Easy. "Yellow Dog Democrat"

Origin of the phrase is a district so solidily Democratic that they would elect a yellow dog if that were the party candidate. Used to be many districts like this in the South. Same model can apply to any party.

rpl writes:
I don't know the exact size of the district, but ALL Congressional districts run into the several hundreds of thousands. If parents have fond memories of dealing with Dr. Paul, that can steer some of the vote overall. But Paul couldn't have "delivered" more than a small percentage of voters in the District.
You are correct. Paul's district had over 650,000 constituents, as of the 2000 census, which means even the impressive total of 4000 births is a trivial fraction of the district's population.

What's interesting to me is that Ron Paul seems to have created a narrative in which he lives in a "small" district (there really is no such thing) wherein he is the family doctor for nearly all of the constituents. Who wouldn't vote for such a pillar of the community? It's complete bunk, of course, but it's well enough crafted that even people who are way above average in their engagement with politics (e.g. commenters on blogs like EconLog) are taken in by it.

To me, this observation is almost enough to bury the Median Voter Theorem. At least in TX-14, policy considerations are way down the list of factors in the election, almost to the point of irrelevance. The outcome of the election would seem to be more of an exercise in marketing. The main question left is whether TX-14 is typical in this respect or an outlier.

Brian Shelley writes:

Since I have lived in his district for a long time, figured I would chime in.

Bob Knaus is probably the closest. In parts of Texas, this one included, a large majority simply will not vote for a Democrat, period. It doesn't matter their positions, (D) = evil.

The area is heavy with petrochemical complexes, agriculture, mixed class suburbia, and evangelical. He is pro-life, anti-taxes, pro-gun, and anti-regulations. His other positions just didn't matter enough to people in the district to vote him out.

The only time that I thought he might be in trouble was when the lines moved to include a lot of NASA employees and he voted against the funding.

I suspect that he was originally able to get elected when the Republican party was more comfortable with a less aggressive foreign policy.

Trevor H writes:

I have just a little to add to the many good comments. For all of his party heterodoxy, Ron Paul has always been a reliable vote against abortion which is a very important issue to many in these parts. And unlike most libertarians he also passes the local litmus test of talking frequently, publicly and convincingly of his commitment to Jesus.

He just has to win the primary anyway. The counties surrounding Houston rarely elect Democrats to anything.

Ron Paul gets them the earmarks:

http://ricochet.com/main-feed/Ron-Paul-s-Earmarks

dha writes:

1. People vote for parties, not individuals. Ron Paul doesn't matter much to what the GOP will do on net, at least as a Senator. When he runs for President people pay more attention.

2. People in his district probably don't view him the way internet libertarians do. Have you ever seen a Ron Paul Congressional (not Presidential) ad? All people see is he is a doctor, in his old military uniform, and he believes in God and wants to overturn Roe vs Wade.

J.D. writes:

Could it have anything to with the fact that he's been delivering babies in his district for decades, and thus receives a lot of voter loyalty from grateful parents and grandparents?

bh writes:

Was Ron Paul on the Republican primary ballot in Texas? It might be interesting to see how his district voted for him in that contest.

Comments for this entry have been closed
Return to top