In 2010 and 2015, I did some data analysis to see which states had the most libertarians, based on Libertarian Party and Ron Paul election results. I’ve now done something similar for 2016.

Unfortunately, in 2016 we didn’t have a libertarianish Republican presidential candidate continue through every primary, and so we can’t use primary election results. However, Rand Paul did run for several months and collected campaign contributions, which we can use. In addition, we can use votes and campaign contributions for the Johnson-Weld campaign to try to see where the libertarians are.

The best way to do this would undoubtedly be to do issue surveys of enough voters in every state that we could estimate the percentage of voters in every state that take libertarian policy positions. But we just don’t have big enough sample sizes at the state level to do this right now. There are experimental, new methods that let us estimate issue positions at the state level with smaller sample sizes, but these methods are extremely time-intensive, and in any case we still don’t have consistent questions over time that would let us develop measures comparable over time.

There’s no one “right” way to do this, but here’s what I did – and all reasonable methods seem to yield similar results. I:

  1. took each state’s percentage of the vote for Johnson-Weld, campaign contributions to Johnson-Weld per capita, and campaign contributions to Rand Paul per capita;
  2. substituted national average values for home states (New Mexico and Massachusetts for Johnson-Weld, Kentucky for Paul);
  3. standardized the three variables to have the same mean and variance; and then
  4. averaged them together.

Substituting national average values for the home states seemed justified because these home states would otherwise be near the top of the rankings, even though none of these states seemed particularly libertarian in other elections when these candidates weren’t running. New Mexico was a below-average Libertarian state before Gary Johnson started running, Kentucky was a mediocre state for Ron Paul contributions in 2008 and 2012, and so on.

Without further ado, here is the ranking of states (and D.C.) by libertarians per capita in 2016, as best we can tell from these three measures:

state lp16_s john_s rand_s libertarians16
District of Columbia -0.2273757 5.063998 5.91443 3.583684
Wyoming 1.34241 0.7520453 2.782263 1.625573
Alaska 1.92475 0.8954656 0.3591455 1.059787
New Hampshire 0.4562405 1.881349 0.7760671 1.037885
Colorado 1.33397 1.46498 -0.1094947 0.8964849
Nevada -0.2358154 1.115038 1.282813 0.7206786
Washington 1.05546 1.029371 -0.0169397 0.6892969
Hawaii 0.1017729 1.514048 -0.1154285 0.5001307
North Dakota 2.211699 -0.9076071 -0.1790468 0.3750151
Oklahoma 1.815033 -0.6429195 -0.1350837 0.3456767
Idaho 0.4224817 0.1906986 0.2989358 0.3040387
Montana 1.679998 -0.4868041 -0.321723 0.2904903
Arizona 0.4056023 0.1295627 0.0997089 0.2116246
South Dakota 1.713757 -0.8959171 -0.3943799 0.1411533
Maine 1.258013 -0.5925905 -0.2984428 0.1223265
Oregon 0.9373039 -0.3396169 -0.2768817 0.1069351
Virginia -0.531205 0.7552064 -0.0225921 0.0671364
Connecticut -0.5396447 0.9314943 -0.1919439 0.0666352
New Mexico 0.1017729 0.1134186 -0.081947 0.0444149
Vermont -0.3370917 0.315843 0.0580951 0.0122821
Nebraska 0.8529069 -0.4625284 -0.3718339 0.0061815
Indiana 1.0639 -0.6398522 -0.4644283 -0.0134603
California -0.2020566 0.1975027 -0.069957 -0.024837
Texas -0.3708505 0.0970403 0.1393362 -0.0448247
Kansas 0.8782258 -0.57144 -0.4413628 -0.044859
Minnesota 0.2030493 -0.1859483 -0.2730781 -0.0853257
Massachusetts 0.1017729 0.0265284 -0.3852127 -0.0856371
Utah -0.1176594 -0.040376 -0.111288 -0.0897745
Iowa 0.1524111 -0.5752427 0.1136567 -0.1030583
Illinois 0.127092 -0.2596285 -0.2625065 -0.131681
Michigan -0.0248227 -0.3442523 -0.0487973 -0.1392908
Missouri -0.1345388 -0.5485766 -0.190781 -0.2912988
Maryland -0.6240419 0.1328795 -0.4675777 -0.31958
Georgia -0.4805668 -0.1231457 -0.4113307 -0.3383478
Tennessee -0.6662404 -0.2489252 -0.1508911 -0.3553522
Wisconsin -0.016383 -0.6375318 -0.4363715 -0.3634288
Florida -1.197942 0.0469722 -0.0559043 -0.4022913
New York -1.105105 0.1764863 -0.3696713 -0.4327633
Kentucky -0.6831198 -0.6866438 0.0554005 -0.438121
North Carolina -0.7253183 -0.3510917 -0.4069826 -0.4944642
Ohio -0.3792903 -0.6159415 -0.5193276 -0.5048531
South Carolina -1.062907 -0.3300098 -0.2762537 -0.55639
West Virginia -0.3455315 -0.9457347 -0.4400242 -0.5770968
Pennsylvania -1.029148 -0.4462026 -0.4482422 -0.6411975
Rhode Island -0.3539712 -0.8717933 -0.7971447 -0.6743031
Arkansas -0.8097153 -0.6793436 -0.6667849 -0.7186146
New Jersey -1.468012 -0.3178977 -0.4170856 -0.7343319
Alabama -1.273899 -0.6585236 -0.4357239 -0.7893822
Louisiana -1.459573 -0.6117339 -0.3441515 -0.8051527
Delaware -1.704324 -0.7962323 -0.2577564 -0.9194376
Mississippi -2.033473 -1.015875 -0.2154772 -1.088275

The “lp16_s” column is the standardized value of Johnson-Weld percentage of the vote, “john_s” is the standardized value of Johnson-Weld campaign donations per person, and “rand_s” is the standardized value of Rand Paul contributions per person. D.C. is the “state” with the most libertarians due to its huge campaign contributions to these two candidates, even though the Libertarian ticket did worse than average there in the actual election. Wyoming comes next. It looks pretty libertarian across the board but was especially supportive of Rand. Alaska is third; it was a great state for the Johnson-Weld ticket. New Hampshire came fourth and was above average on all three measures.

I also looked into the possibility that Johnson-Weld did worse in swing states because of tactical voting, but I could find no evidence for this hypothesis. States where the polls were close did not show lower third-party support, surprisingly.

These results are pretty similar to those from previous years, except that I hadn’t looked at D.C. before. It’s not surprising that D.C. would score so highly on campaign contributions, because lots of people there are really interested in politics. Maybe I should have used campaign contributions to each candidate as a share of all campaign contributions, rather than per capita, but I’m unpersuaded that this would be the right way to go. D.C. really may have lots of libertarians – who are nevertheless swamped by those of other ideologies. Wyoming does a lot better than in previous years, and Montana significantly worse. Alaska has always been near the top in these numbers. New Hampshire hasn’t separated itself from the pack despite the Free State Project. It was the best state other than D.C., New Mexico, and Massachusetts for Johnson-Weld campaign contributions, but it was only moderately above average for Johnson-Weld votes and Rand Paul campaign contributions.

Did the emergence of the state reduce the rate of human death from warfare? Steven Pinker’s outstanding book, The Better Angels of Our Nature: Why Violence Has Declined, surveys many reasons why you are less likely to die from violence today than your ancestors were. Part of his explanation is that warfare was constant in stateless, anarchic societies, but the emergence of the state, beginning about six thousand years ago, helped reduce this problem. He has nice things to say about Thomas Hobbes’ thesis in Leviathan, that a powerful government is necessary to rescue people from their natural state of constant warfare.

In my most recent Learn Liberty blog post, I question this finding of Pinker’s. I argue that the evidence he presents for the claim does not suffice to prove it, because there are other factors that could explain declining rates of war death. Moreover, even if the state reduced war death somewhat, we can’t necessarily infer from that fact alone that the state increased human welfare. From the post:

[T]here is an important conceptual problem for the claim that the rise of the state improved human welfare by reducing violent deaths.

After all, early states arose almost exclusively out of conquest, as Pinker concedes. They started as roving bands of armed robbers, who eventually found that converting robbery into regularized taxation would destroy less wealth and generate more revenue over the long run. Autonomous peoples do not go into “subject” status willingly.

More at the link.

My latest Learn Liberty blog post is on the topic above and can be found here. Excerpt:

Kant’s moral philosophy justifies extremely strong individual rights against coercion. The only justification for coercion in his philosophy seems to be defense of self or others. His ideal government therefore seems to be extremely limited and to allow for the free play of citizens’ imaginations, enterprise, and experiments in living.

Now that the 2016 election results are available by town for New Hampshire, I thought I would take a look at where libertarian candidates tended to do well or poorly, and how that pattern compared with conservative versus progressive support by town.

To measure libertarian voting by town, I used different variables in different years.

For 2008, I used:

  • Ron Paul vote share in the Republican presidential primary,
  • Libertarian Party vote share in the general presidential election (both Bob Barr and George Phillies had separate Libertarian candidacies in New Hampshire that year),
  • Libertarian Party vote share in the general gubernatorial election.

For 2012, I used:

  • Ron Paul vote share in the Republican presidential primary,
  • Libertarian Party and Ron Paul write-in vote share in the general presidential election,
  • Andrew Hemingway vote share in the Republican gubernatorial primary of 2014. (So this is sort of a 2012-4 measure really.)

For 2016, I used:

  • Rand Paul vote share in the Republican presidential primary (he had dropped out of the race, but over 1% of voters voted for him anyway),
  • Frank Edelblut vote share in the Republican gubernatorial primary,
  • Libertarian Party vote share in the general presidential election,
  • Libertarian Party and Aaron Day vote share in the general U.S. Senate election.

I also calculated conservatism vs. progressivism by town for 2012 and 2016.

For 2012 conservative voting, I used:

  • Republican Party general presidential election vote share,
  • Republican Party general gubernatorial election vote share (2012),
  • Republican Party general gubernatorial election vote share (2014).

For 2016 conservative voting, I used:

  • Republican Party general presidential election vote share,
  • Republican Party and Aaron Day vote share in the general U.S. Senate election,
  • Republican Party general gubernatorial election vote share.

I also looked at how these variables correlated with each other. The strongest correlations I found outside the Republican candidates’ correlations with each other were between LP + Day Senate 2016 vote share and Ron Paul 2012 primary vote share (r=0.51), LP + Day Senate 2016 vote share and Trump vote share (r=0.37), Rand 2016 and Ron 2012 primary vote share (r=0.32), LP + Day Senate 2016 vote share and Edelblut ’16 vote share (r=0.30), Johnson ’12 + Ron Paul write-ins and Ron Paul 2012 primary vote share (r=0.51), Trump vote share and Ron Paul 2012 primary vote share (r=0.45), Paul ’12 and Paul ’08 vote share (r=0.49), Paul ’08 and Hemingway ’14 vote share (r=0.43), and Paul ’12 and Hemingway ’14 vote share (r=0.34). Interestingly, the Johnson-Weld ticket really didn’t correlate with anything else at all, suggesting that most of their voters were simply anti-Clinton and anti-Trump, not libertarian leaners. Most of those voters probably won’t stick around for future Libertarian candidacies unless they don’t have a lot of options.

These correlations also imply that a lot of Ron Paul’s 2012 primary vote came from disaffected, non-conservative, potentially populist or nationalist Republicans. This is consistent with what I reported here on Pileus years ago about how Ron Paul in 2012 added a bunch of anti-establishment, moderate to liberal independents and Republicans to his libertarian base. It also suggests, perhaps, some reason for optimism about the Trump phenomenon. A lot of his voters are simply alienated and not that strongly ideological. If they could vote for both Ron Paul in 2012 and Donald Trump in 2016, these people must be ideologically flexible, even if consistently anti-establishment.

So where can you find libertarians, progressives, and conservatives in New Hampshire? These maps tell the tale. (Unincorporated townships and towns with fewer than 100 votes are not mapped.)

nhlib16 nhlib12 nhlib08 nhcon16 nhcon12

Please note that each variable is recentered each year, so that it is impossible to compare towns across years in any absolute sense (“this town is becoming more libertarian” is an inference you absolutely cannot make from these data), though you could make comparisons over time, relative to the average town in New Hampshire (“this town used to be quite a bit more libertarian than other towns, but now it’s only average” is something you could say based on these data). The recentering has to be done because different candidates run and are included in the calculations in different years.

Now then, where are the libertarians? The results aren’t greatly different from those I’ve reported before, with New Hampshire’s most libertarian towns generally lying in the Appalachian mountains and foothills of the western part of the state. However, in 2016 there is a notable change, with the libertarian center of gravity shifting southward to Cheshire and western Hillsborough counties. This surprised me a bit at first, but it makes sense once you consider that conservatarian Republican gubernatorial candidate Frank Edelblut is from western Hillsborough County and absolutely dominated the vote in his hometown and nearby towns, coming close to knocking off then-presumptive nominee Chris Sununu (now governor elect) statewide.

Republicans, meanwhile, are much stronger in the southeastern third of the state than elsewhere in 2012, but in 2016 they had more even support throughout the state, making clear inroads into the relatively deprived North Country. This is consistent with the general shift of the party to the nationalist radical right with accompanying loss of support among upscale groups and growth among the white working class.

Scatter plots make these phenomena clearer (click to expand).

The last scatter plot, in particular, shows that between 2012 and 2016 Republicans lost, relatively to the average town, in upscale towns like Hanover, Bedford, New Castle, Amherst, Hollis, and New London (most of these are strongly Republican towns, but Hanover is the most Democratic town in the state). Meanwhile, Republicans gained in the small towns of the North Country, places like Landaff, Berlin, Northumberland, and Millsfield. Libertarians (not necessarily Libertarian Party, but libertarianish candidates) look to have gained in western Hillsborough County, towns like Greenville, Mason, and New Ipswich (all heavily GOP). Again, this pattern really reflects the strength of Republican Frank Edelblut in his primary.

There’s been some debate about whether independent conservatarian candidate Aaron Day (former chairman of the Free State Project Board of Directors) cost Republican Kelly Ayotte her U.S. Senate seat at this past election. Skeptics point to the fact that Day and Libertarian Brian Chabot between them about equaled Gary Johnson’s vote percentage in the presidential race (Johnson-Weld got 4.2% in New Hampshire, Chabot and Day between them about 4.1%). They say Day merely siphoned off Libertarian voters from Chabot, not Republican voters from Ayotte.

They’re probably wrong. Here’s why.

First, let’s note that Clinton and Trump were the two least popular major-party presidential candidates in the history of polling. In particular, Trump was (and is) by far the least popular presidential candidate (and now, president-elect) that we’ve ever seen. Under those conditions, you expect an outsized third-party vote share in the presidential race. By contrast, Maggie Hassan and Kelly Ayotte, the two U.S. Senate candidates, were reasonably popular – and about equally so. So we should expect a small third-party vote share in the U.S. Senate race. The fact that the third-party vote share in the Senate race equalled that in the presidential race therefore suggests something else is going on – perhaps a particularly strong independent candidacy. Libertarian Chabot got 1.7% of the vote, while independent Day got 2.4% of the vote – so if one of them was the particularly strong candidate, it was Day.

To try to see whether Day was siphoning off Republican or Libertarian votes, I looked at town-by-town results for all 239 New Hampshire jurisdictions with voters in this race. I then looked at how the number of Libertarian and Republican presidential votes by town correlated with Day’s support. The results are in the figure below.

Aaron Day Vote Sources

The coefficient estimate on percentage of the vote for Trump-Pence in a town is 0.05 and is highly statistically significant, as you can see by the tiny confidence interval on the estimate. This estimate means that for every 20 additional Trump-Pence voters in a town, one additional voter cast a vote for Day in the Senate race.

The coefficient estimate on percentage of the vote for Johnson-Weld in a town is 0.08 but not statistically significant. This estimate implies that for every 12 additional Johnson-Weld voters in a town, one additional voter cast a vote for day in the Senate race — but again, we can’t be sure this is really any different from zero.

Day’s vote share was a little more than one-twentieth of the Trump-Pence ticket’s in New Hampshire. In other words, we can be reasonably confident that almost all of Day’s electoral support came from Republicans, not Libertarians.

To verify that Day specifically was the spoiler, not Brian Chabot, we can do the same exercise for Brian Chabot’s vote shares by town. The figure below shows those results.

Brian Chabot Vote Sources

Look at the difference! Chabot pulled overwhelmingly from Johnson-Weld voters. For every four additional Johnson-Weld voters in a town, Chabot got an additional one vote. Meanwhile, Chabot did siphon a few Republicans. For every 80 additional Trump-Pence votes in a town, Chabot got about one vote. Still, these results suggest that over half of Chabot’s support came from Libertarians, not Republicans.

Now, Kelly Ayotte lost by 1,000 votes, just over 0.1 percentage points. My estimates suggest that without Aaron Day in the race, Kelly Ayotte would have won her race by about 15,000 votes, similar to Republican Chris Sununu’s margin of victory in the governor’s race.

Is it possible that without Aaron Day in the race, those 16,000 or so disaffected Republicans would still have voted for some other third-party candidate or a write-in, or just that race blank? Sure, it’s possible. We can’t rule out that possibility because we can’t do an experiment in which we randomly assign some towns’ ballots to have Aaron Day on them and some not to have him on them. But at minimum these results strongly suggest that Kelly Ayotte disaffected a decisive share of Republican voters who went for Trump and Sununu but not for her.

What if we can’t make government smaller?” the Niskanen Center’s Will Wilkinson asks. He says that the evidence, particularly Wagner’s Law, shows that government spending is impervious to political assault, and libertarians should make their peace with big government. Instead, libertarians should focus on reforming regulations to foster competition and the market process.

I have a different read of the evidence from Will’s. At the Learn Liberty blog, I write,

Governments do have a tendency to grow. However, the U.S. has cut government consumption significantly in the past and could do so again. The drivers of welfare spending are the aging of the population and rising health care costs, not political support for new programs.

I support those claims with a series of charts. Check it out!

I’ve started a new blogging gig at Learn Liberty, a project of the Institute for Humane Studies. I’ll be putting links to these posts here. My posts there will have the benefit of an editor, which is probably something I need.

The first is on partisan rationalization and why epistocracy may not save us after all – a suitable topic after this election, no? Here’s the lede:

Highly informed voters are also highly biased. That’s a serious problem for democracy, but also for any other system of political decision-making in big groups.

Two new books, Against Democracy by Georgetown moral philosopher Jason Brennan and Democracy for Realists: Why Elections Do Not Produce Responsive Government by Princeton University political scientist Christopher Achen and Vanderbilt University political scientist Larry Bartels, deal intellectual hammer blows to the political system so many of us take for granted: democracy.

%d bloggers like this: