?

Log in

No account? Create an account

Border experiments - Open Knowledge — LiveJournal

Sep. 7th, 2006

01:47 am - Border experiments

Previous Entry Share Next Entry

Immigration is an issue that deeply divides libertarians. Although I suspect the majority of libertarians are open borders advocates, a significant minority favor strictly limited immigration. They fear that a horde of stupid, socialist immigrants will, among other things, overwhelm our water supply, clog our roads, steal our stuff, and vote themselves ever increasing amounts of welfare.

Pro-immigrant libertarians, on the other hand, emphasize the costs of preventing people from fleeing socialist hellholes, and finding work where they can best fulfill their highest potential. They also oppose immigration restrictions as unethical infringements on the right of association and the right to private property.

How could these differences be resolved, in principle? Some of the differences result from fundamaental differences in values. For example, someone who highly values woodlands is not going to be happy with Hong Kong population densities. Such valuations are aesthetic judgments that are unreconciliable via evidence and rationality. The best that can be achieved is to set up society in such a way that there are variety of environments, each suited to different tastes.

However, many differences regarding economic growth, crime, and effects on politics could theoretically be resolved. If economists were gods, for example, they could establish multiple universes, each differing initially only in the U.S. border policy. By comparing the outcomes in each of these alternate universes, the effects of different border policies could be measured and compared.

Unfortunately, economists are not gods.

However, what if we ran different border experiments at a smaller scale?

If the anti-immigrantion advocates are right, we should see benefits to limiting immigration at smaller scales. After all, if it's bad for a an illegal Mexican to enter my country, then isn't it also bad to allow a illegal Mexican to enter my state? Or my hometown?

So my suggestion is this: set up a dozen or so 10 x 10 mile plots in the Texas desert along the Mexican border. Each plot would be subject to the same laws as the rest of Texas, with one exception -- each plot would have a different border law. So one plot would allow anyone in the world to move and buy land there. Another plot would allow no one except existing Americans to buy land. Another plot would only allow those who passed an IQ test to enter. And so on.

Exit from each plot would be governed by existing U.S. law. If you're an U.S. citizen or green card holder you can pass freely. Otherwise, you would be treated as an immigrant from your country of origin, just as you would at any other border crossing.

To prevent illegals from crossing into the rest of the U.S., each plot would be surrounded on all four sides by a secure wall. This itself would be a valuable test -- if the U.S. border control can't contain a few 10 x 10 mile plots, they'll certainly be unable to prevent immigrants from crossing a continent wide barrier.

The health of the each of these plots could be measured. What's the lifespan? GDP? Disease rates? Population density? Crime rates? Partisans, at the outset, could wager on the future of each plot.

Presumably, over time, it would become clearer which border policy resulted in a healthier, wealthier society.

Moreover, if a particular border policy proved successful, the size of the plot could be expanded gradually, perhaps growing large enough to encompass the entire country. Thus, better border policies could be implemented, without the risk and disruption from switching abruptly to a new and unproven border policy.

Comments:

[User Picture]
From:zzzing
Date:September 7th, 2006 10:26 am (UTC)
(Link)
if it happened in reality, the plot with the most lenient immigration laws might be totally overwhelmed with immigrants because they might see it as an opportunity to eventually enter the US proper via some kind of loophole. or the plot might self regulate because the population would get too dense and people wouldn't go there no matter what, or it might be beneficial to that plot to have a hugely dense population of immigrants. i guess we need to see what could actually happen. i guess i just answered my own question.

this is why we need supercomputers that can calculate human behavior and patterns in society like the weather. get the scientists working on that.
(Reply) (Thread)
[User Picture]
From:atek128
Date:September 7th, 2006 11:42 am (UTC)
(Link)
if the lenient plot also had free trade and lenient government it would likely become another singapore, hong kong, lichenstein, or dubai. At first it would suck, thousands of refugees fleeing their with only the clothes on their backs, much like hong kong when Mao was winning. However, with no welfare or other handouts to attract vagabonds, in all likelyhood the vast majority of people would be coming to work. What happens when you combine lots of people in a small area with capitalism... division of labor, market transactions, and eventually prosperity...

I'd only add as a caveat, that the province should be allowed to import goods through the US without paying US tariffs or observing US quotas.

atek3

(Reply) (Thread)
[User Picture]
From:candid
Date:September 7th, 2006 01:10 pm (UTC)

putting on my economist's hat

(Link)
I have a hard time seeing how you would keep this a controlled experiment. If I were MeCHA (or the Mexican government, or Al Qaeda), I would devote any and all resources necessary to making sure the "most lenient" plots prospered. And if I were a Buchananite, I would devote resources to making sure those same plots failed.

In the end, my suspicion is that the "outcome" would depend more on the proxy struggle between pro- and anti- immigration forces, and less on the merits / demerits of different immigration policies.
(Reply) (Thread)
[User Picture]
From:crasch
Date:September 7th, 2006 04:28 pm (UTC)

Re: putting on my economist's hat

(Link)
What specific machinations do you have in mind?

It's difficult to see how you could construct any experiment of this nature that would be resistant from all machination. The question is whether we could learn something from it anyway.

Most of the U.S. population supports strict immigration controls. Assuming that our current national border policies are the result of the battle between the pro and anti-immigrant forces, then the anti-immigrant forces have historically been ascendant.

Therefore, it seems to me that the experiment would be biased at the outset against the lenient plots. If the lenient plots were economically and socially healthy and robust despite those machinations, then I think that is useful information.
(Reply) (Parent) (Thread)
[User Picture]
From:candid
Date:September 7th, 2006 05:51 pm (UTC)

Re: putting on my economist's hat

(Link)
What specific machinations do you have in mind?

Some obvious ones are

* Manipulating the crime rate.
* Subsidizing / sabotaging the economy.
* Subsidizing / sabotaging public services (parks, garbage collection, etc...).

I also think that the plots with restrictive border law would remain empty, since I'm not sure who would want to move right to the border.
(Reply) (Parent) (Thread)
[User Picture]
From:crasch
Date:September 7th, 2006 06:10 pm (UTC)

Re: putting on my economist's hat

(Link)
Well, without doing the experiment, it's difficult to know the magnitude of those effects. However, here's how I would address the concerns:

What do you mean by manipulating the crime rate? If you mean, that the statistics would be falsified by partisans, that could be corrected by random sampling by independent observers.

As for illegitimate subsidies, what counts as an illegitimate subsidy? This effect could be neutralized by providing subsidies to the other plots of an equal amount.

As for sabotage, that could be prosecuted as with any other crime.

We could also set up plots in other desolate parts of the country. North Dakota, for example. If all the lenient plots become prosperous, little "Hong Kongs", and all the restrictive plots remained desolate wasteland, regardless of region, it would difficult to argue that it was all due to subsidies/sabotage from partisans.

I also think that the plots with restrictive border law would remain empty, since I'm not sure who would want to move right to the border.

What do you predict would happen with the lenient plots?
(Reply) (Parent) (Thread)
[User Picture]
From:candid
Date:September 7th, 2006 07:01 pm (UTC)

Re: putting on my economist's hat

(Link)
What do you mean by manipulating the crime rate?

Well, if I wanted the crime rate to go down in small town X, I could use my own money to hire some sort of private security force to supplement the police. If I wanted the crime rate to go up, I'm sure I could hire criminals to make it happen.

As for illegitimate subsidies, what counts as an illegitimate subsidy?

I never said the word "illegitimate". Imagine that I am doing a "random" experiment to see whether Textbook A or Textbook B does a better job teaching calculus. If, say, the publisher of Textbook A hires a bunch of tutors to help the class using that book, don't you think that might invalidate the experiment?

As for sabotage, that could be prosecuted as with any other crime.

I don't know that sabotaging a social experiment is a crime.

What do you predict would happen with the lenient plots?

Honestly, I think they would become shitholes, and that their main use would be as staging grounds for people to sneak into the rest of the country.
(Reply) (Parent) (Thread)
[User Picture]
From:crasch
Date:September 7th, 2006 07:45 pm (UTC)

Re: putting on my economist's hat

(Link)
Well, if I wanted the crime rate to go down in small town X, I could use my own money to hire some sort of private security force to supplement the police. If I wanted the crime rate to go up, I'm sure I could hire criminals to make it happen.

I'm not sure how the hiring of private security forces invalidates the experiment. Open borders advocates argue that society can adapt to the influx of immigrants, in part by hiring more security forces. If the property owners and civic leaders within the free movement zone hired additional private security to control the crime rate, that is what open borders advocates would predict. And they could afford it because of the expected increase in economic growth.

As for hiring criminals to boost the crime rates, how many people of means would spend their money that way? If it came out that MinuteMen or some such were doing it, it would hurt their cause in the public perception. Likewise, if the Mexican government wants to increase border freedom, it would be stupid of them to try to invalidate the experiment.

As for illegitimate subsidies, what counts as an illegitimate subsidy?

I never said the word "illegitimate". Imagine that I am doing a "random" experiment to see whether Textbook A or Textbook B does a better job teaching calculus. If, say, the publisher of Textbook A hires a bunch of tutors to help the class using that book, don't you think that might invalidate the experiment?


One of the arguments of the open borders advocates is that it would attract foreign investment, which more restrictive regimes do not attract. However, if you define all outside foreign investment as a "subsidy" then, a priori, you've eliminated one of the advantages of open borders.

I'm willing to entertain the notion that Soviet/Cuba style subsidies should be compensated for, but I'm not willing to say that all foreign investment is such. Presumably, you don't think that all foreign investment is a subsidy, so I'd like to know more about how you define it.

For example, if someone buys a plot of land, and sets up a church, is that a subsidy?
What if the church offers free daycare?


I don't know that sabotaging a social experiment is a crime.

Well, then what do you mean by sabotage? When you said sabotage, I had in mind Mecha activists pouring sand in the MinuteMen plot's garbage trucks.
(Reply) (Parent) (Thread)
[User Picture]
From:candid
Date:September 7th, 2006 08:47 pm (UTC)

Re: putting on my economist's hat

(Link)
Likewise, if the Mexican government wants to increase border freedom, it would be stupid of them to try to invalidate the experiment.

For many, many years will I remember this day, on which you assumed rational, foresighted behavior of government!
(Reply) (Parent) (Thread)
[User Picture]
From:crasch
Date:September 7th, 2006 09:01 pm (UTC)

Re: putting on my economist's hat

(Link)
Heh.

Oh, I think that they may well do something stupid.


However, if we're going to credit the Mexican government with enough rationality to try to skew the experiment by hiring criminals to boost the crime rate, then it seems to me that we should also credit them with enough rationality to recognize the political repercussions if they were caught.
(Reply) (Parent) (Thread)
[User Picture]
From:crasch
Date:September 7th, 2006 04:35 pm (UTC)

Re: putting on my economist's hat

(Link)
In the end, my suspicion is that the "outcome" would depend more on the proxy struggle between pro- and anti- immigration forces, and less on the merits / demerits of different immigration policies.

Also, assuming that this happens, it would be useful to know in what ways partisans tried to skew the results. Subsequent experiments could then be designed to mitigate the effects of partisan manipulation.
(Reply) (Parent) (Thread)
[User Picture]
From:candid
Date:September 7th, 2006 05:51 pm (UTC)

Re: putting on my economist's hat

(Link)
Man, you have an aspergers-y solution for everything! ;)
(Reply) (Parent) (Thread)
[User Picture]
From:visgoth
Date:September 7th, 2006 01:57 pm (UTC)
(Link)
http://www.fallacyfiles.org/division.html
(Reply) (Thread)
[User Picture]
From:crasch
Date:September 7th, 2006 04:16 pm (UTC)
(Link)
What part of my argument do you think is an example of the division fallacy?
(Reply) (Parent) (Thread)
[User Picture]
From:visgoth
Date:September 7th, 2006 04:29 pm (UTC)
(Link)
That a 10x10 mile section of Texas could possibly come even vaguely close to accurately modeling the economy and society of the United States as a whole.
(Reply) (Parent) (Thread)
[User Picture]
From:crasch
Date:September 7th, 2006 04:47 pm (UTC)
(Link)
That may be true. However, scientists use imperfect models all the time. A mouse is not perfect model of a man, and a Texas farm is not a perfect model of a country. But we've learned a great deal about ourselves by studying mice. The question is whether these plots can tell us something interesting about immigration. Nativists argue that the immigrants harm our country by causing crime, burdening social services, and destroying the environment. If that's true, then those effects should be as visible at the town level as at the national level, since national statistics are simply the aggregation of statistics from the local level.

Therefore, to argue that I've committed the division fallacy, you have to show that the ill-effects of immigration at the local level aren't measurable, but somehow become measurable at the national level.
(Reply) (Parent) (Thread)
[User Picture]
From:visgoth
Date:September 7th, 2006 06:03 pm (UTC)
(Link)
I'm not saying it couldn't tell us anything interesting. I'm saying that the results would tell us nothing about how similar actions would impact a very different environment on a vastly larger scale.

In your proposed model, people who are skilled at growing pecans, being allowed free entry might be a great boon to the society, so an abundance of people with such skill and an open border policy may work fabulously. But those same pecan growers would have a hell of a difficult time making use of that skill in most of the US, which makes that result meaningless.

Your proposal assumes that what works on one particular plot of land in Texas can translate to the entire US. Having seen the difference between land near El Paso and land near Brownsville, I submit that it is a trivial task to find two such plots directly on the Texas/Mexico border that do not even closely model one another.

If you proposed your experiment be done in numerous locations across the country I concede that it may well show something useful and valid.

And I believe that when I (as a reviewer) say that an undefined portion of Texas is not a reliable model of the entirety of the United States (which seems self-evident) it falls to you (as the experimenter) to show that it is, or to revise your experiment to address that shortcoming.
(Reply) (Parent) (Thread)
[User Picture]
From:crasch
Date:September 7th, 2006 06:22 pm (UTC)
(Link)
If you proposed your experiment be done in numerous locations across the country I concede that it may well show something useful and valid.

Great! This is blog post of an idea, that, as far as I know, has never been published anywhere else. Therefore, obviously, all of the details have not been worked out. For example, I have not picked out specific plots of land in Texas. However, do you at least concede that comparable plots of land could be found, in principle?

I also think it would be good idea to try in different parts of the country.

As for whether the results would say anything about what we should do about the national borders, you may be right. However, one of the virtues of this idea is that we don't have to translate it into national border policy all at once. If it turns out to be successfull, we could gradually expand the region governed by the optimal border law (whatever it might be) until the region is large enough that you would agree that it said something useful about what our national border policy would be.

(Reply) (Parent) (Thread)
[User Picture]
From:visgoth
Date:September 7th, 2006 07:10 pm (UTC)
(Link)
However, do you at least concede that comparable plots of land could be found, in principle?

Depends on what you mean. If you mean "sufficiently comparable to one another," (in a given state), then probably yes. If you mean "representative of the nation, in Texas, on the Mexico border," then no, I do not. But I believe (in principle) that if one were to set up your proposed experiment in Texas, New York, Michigan, Ohio, California, and so on (as random examples, not as hard and fast "these should be the locations") a sufficiently representative "average" of the nation could be achieved.

Perhaps I was too brief in my two earliest replies. I certainly didn't mean to say "useless idea," merely to point out what jumped out at me as a significant weakness in it.

While I see the potential usefulness of your proposed approach, I think it is a less than optimal solution. I think just making incremental changes to policy and watching the effect would work better. I'll write a bit more extensively on it in my journal.
(Reply) (Parent) (Thread)
[User Picture]
From:ernunnos
Date:September 7th, 2006 03:34 pm (UTC)
(Link)
We already have localized experiments. We call them "California", "Arizona", "Texas"... And within those, smaller experiments like "South Phoenix". Within and between those experiment zones, people cast their vote regarding the success of those experiments by moving. This manifests economically - at least during non-bubble periods - in property values. And we can also observe the political impact by noting any consistent differences between the types of representation each zone ends up with.

The tests are being done.

Of particular interest to me is one where right across the street from where I used to live, Guadalupe. It's a 1 square mile town, right between Phoenix and Tempe. It's so small, and the borders so stark, that it even looks like a deliberately created experiment. And if you look into its history, it kind of is. And I even voted on that one.

(Reply) (Thread)
[User Picture]
From:crasch
Date:September 7th, 2006 04:15 pm (UTC)
(Link)
While the places you cite may be natural experiments of a sort, they are not a test of what a region with open borders would look like. None of those places have truly open borders, nor are they controlled for climate, soil fertility, proximity to ports/ocean, geography, history, etc.

The plots would all start off as very similar climate, location, legal regime (aside from border control), and would have no or little pre-existing population.

If you're right about the ill-effects of open immigration, it should be quickly apparent in the lenient plots. And you could make a killing betting against people like me.

At a minimum, we could see how well your wall would work.
(Reply) (Parent) (Thread)
[User Picture]
From:ernunnos
Date:September 7th, 2006 09:35 pm (UTC)
(Link)
While the places you cite may be natural experiments of a sort, they are not a test of what a region with open borders would look like.
Sure they are. Immigration law is largely unenforced in those regions. They aren't completely controlled, but no social experiment ever is. That's why we have multivariate analysis.

I'm already betting against people like you when I didn't buy a house in Guadalupe. And so are you.

(Reply) (Parent) (Thread)
[User Picture]
From:crasch
Date:September 7th, 2006 04:19 pm (UTC)
(Link)
Note that one of the plots can have whatever border regime the nativists want too, which could be even more restrictive than the existing border policy. If it turns out that is the optimal policy, it would bolster your own argument.
(Reply) (Parent) (Thread)