Estimating the “Liberty Bloc” in Each State, Part 3

In the third and final installment of this series (part 1 here, part 2 here), I investigate the truth of that hackneyed Margaret Mead quotation, “Never doubt that a small group of thoughtful, committed citizens can change the world. Indeed, it is the only thing that ever has.”

Is that really true when it comes to libertarians? To recap, in parts 1 and 2 of this series, I investigated whether votes for Ron Paul, per capita donations to Ron Paul, Libertarian presidential votes in 1996-2004, and Libertarian presidential votes in 2008 all correlate together at the state level. If they do consistently track together across the states, that fact implies that there is some underlying factor explaining all of them. It turns out that they do correlate together rather strongly, and I interpret the extracted common factor as the size of the liberty bloc in each state. So the question is – does this liberty bloc have any real influence on politics, or does its minority status doom it to irrelevance?

To test the political influence of libertarians, I model state respect for individual freedom as a function of libertarian constituency, liberal constituency, political institutions, and some demographic controls. In short, I’m trying to find out whether states with more libertarians are freer. The standard and most plausible way to interpret a correlation between state ideology and policy is causal: libertarians influence the political process in their states. It is also possible that libertarians tend to move to states that are freer to begin with, but most of us are not that footloose. Many of us end up stuck in places like New York. (Cough.)

The dependent variable in this regression model is state-level freedom, including both economic and personal freedom, as measured by the “Ruger-Sorens Index” (RSI) in our study “Freedom in the 50 States.” However, I’m going to use the latest and greatest data that haven’t been published yet (next version of the study coming out in January). This is the regression equation:

“Unionization” is the percentage of workers covered by collective bargaining contracts in 1977 (I chose an early year because freedom can have reciprocal effects on unionization), “PctBlack” is the percentage of the state population that is black (the reason for including this variable is to capture well-known “racial threat” dynamics, whereby whites in states that have larger black populations are more racist), and “LegProf” is legislative professionalism, a technical term for how similar a state legislature is to Congress in terms of salary, staff, and session length. I expect states with more union members and racists to be less free, and states with well-paid legislators, large legislative staffs, and long sessions to be less free. Both the “liberal” and “libertarian” variables (defined in part 2 of this series) have been rescaled from 0 to 1.

Here are the results:

Regression with robust standard errors                 Number of obs =      50
F(  6,    43) =   23.16
Prob > F      =  0.0000
R-squared     =  0.7344
Root MSE      =  .11918

|               Robust
freedom     |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
libertarian |   .1731908   .0760284     2.28   0.028     .0198649    .3265167
liberal     |   .5440847   .2185759     2.49   0.017     .1032845    .9848849
liberalsq   |  -.8819435   .1994393    -4.42   0.000    -1.284151   -.4797358
union77     |   -.010695   .0028045    -3.81   0.000    -.0163508   -.0050391
estbkpct    |  -.0055999   .0019713    -2.84   0.007    -.0095754   -.0016244
legprof     |    -.37702   .1805852    -2.09   0.043    -.7412048   -.0128351
_cons       |   .9625205   .0842816    11.42   0.000     .7925504    1.132491

All my hypotheses are confirmed, and most interestingly, we see that states with more libertarians are freer. The effect of liberalism on freedom is less clear from a quick look at the table, so a plot will be more effective. Here is how liberalism affects freedom, when all other variables are set to their means:

So a little bit of liberalism might help, but a lot really hurts. Incidentally, I also tried including urbanization rate and percentage religious/Christian/evangelical, but none of those variables had any effect on freedom. The common perception among libertarians that more urbanized areas are less free is simply not true, once you control for things like unionization and liberal ideology.

To conclude, then, libertarians do make a difference, though not as much as liberals and conservatives make. If Idaho had only as many libertarians as Illinois, it would no longer be the fifth-freest state in the country; instead, it would be only about as free as Florida, Iowa, or North Dakota, slotting in at #11. If it had only as many libertarians as the least libertarian state, Mississippi, Idaho would be very close to Utah, around #20.


8 thoughts on “Estimating the “Liberty Bloc” in Each State, Part 3

  1. It seems like there are a variety of other causations possible. As you allude, people may chose where to live based on how free it is. It may also be that state freedom affects people’s political beliefs. I also wonder whether non-libertarians in states with more libertarians are more libertarian (liberaltarians or such). The latter would be important because it would mean that moving new libertarians to the state would have a less than expected effect.

    I wonder if NH can serve as an intervention study of the effect of 900 libertarians moving to a state?

  2. Patri,

    Thanks for stopping by. You raise some important points. Responses interspersed below…

    As you allude, people may chose where to live based on how free it is.

    True, although I think this accounts for only a small part of the correlation at most. Other political science work that analyzes the correlation between citizen and policy ideology assumes that the causation runs from the former to the latter. Also, preliminary analyses that I’ve done on migration seems to show that everyone prefers to move to places with more freedom, regardless of ideology. For migration endogeneity to be a strong component of the observed correlation, it would have to be the case that left-liberals and populists really prefer to move to unfree places – and I’m not sure that they do. (Most of those federal workers in DC seem to want to live in northern VA rather than Maryland.)

    I also wonder whether non-libertarians in states with more libertarians are more libertarian (liberaltarians or such).

    I tend to agree. The ideological distribution in each state probably is a bell curve (or “bell plane” with two-dimensional ideology). Thus, while only the hardest-core libertarians actually vote LP consistently, give to Ron Paul, etc., states with more of those really hard-core libertarians probably also have more soft-core libertarians, more libertarian-leaning conservatives and liberals, and fewer, less radical populists.

    The latter would be important because it would mean that moving new libertarians to the state would have a less than expected effect.

    Depends. If the relatively hard-core, self-aware, libertarian activists have no effect beyond their own numbers, then clearly the migration wouldn’t affect overall state culture much – they wouldn’t be shifting the bell plane so much as creating a secondary mode at one extreme. But if they are savvy and active enough to move the body politic in a libertarian direction, then they would have an effect.

    I wonder if NH can serve as an intervention study of the effect of 900 libertarians moving to a state?

    Indeed… The difficulty is how to ascribe changes in the state to that migration. We need valid “before” and “after” indicators. Fiscal data is one possibility. On a related point, I’d like to run some hypotheticals, assuming new libertarian activists in NH have the same effect that libertarians might elsewhere (in terms of shifting the overall ideological distribution), and seeing how – on the basis of the regression equation – we would expect freedom to change with different levels of in-migration.

  3. As a socially conservative Republican in NH, it doesn’t surprise me that increasing liberalism has a nonlinear effect upon freedom. While I oppose same-sex marriage, abortion, and illegal drug use (incuding marijuana), liberals certainly approve of redefining marriage to include same-sex couples, and allowing the killing of one’s own pre-born child. On the other hand, while I detest smoking, I was outraged when liberals pushed through the smoking ban in restaurants. Anti-discrimination laws deprive us of the freedom to make our own decisions. Property rights are not respected by liberals. Basically I view any law that proscribes an individual’s (or group’s, for that matter) behavior or action as something that limits freedom. It seems to me to be Democrats who are especially fond of passing laws that tell us what we cannot do. I’m sure that Republicans to some extent are guilty of that as well, but not so much in my opinion.

    By the way, does the beta_sub_2 coefficient look more significant if the liberal squared term is removed from the equation? Presumably the residuals look more randomly distributed with the liberal^2 term than without.

  4. Abortion isn’t included in the Freedom Index, but drug, gambling, and prostitution laws are. Same-sex marriage doesn’t count either, but same-sex civil unions and domestic partnerships are positive for freedom in the Index. The most conservative states do tend to score badly on these elements of freedom, although some liberal states like Illinois do as well.

    Without the square, “liberal” becomes negative and significant – so yes, the residuals are less heteroskedastic when the square is included.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s