Taken together, bullet points 2, 3, and 4 are a textbook strawman.
Uhm, point 2 at least is a straight up fact, as real markets typically diverge to varying degrees from perfect markets. I also note you don’t actually dispute point 1, which is the strong statement of a deontological ethical position, and pure deontology remains incompatible with consequentialism, hence the apparent contradiction in ethical systems.
To me, this speaks more to the extent of your motivation to find merit-worthy libertarian writing than to the merit of libertarian ideas.
If I’ve been unimpressed with the arguments of rank-and-file members of a political position, why would I be motivated to look for better writing that may or may not exist? Do you go looking for merit-worthy religious apologetics?
That said, do you know of any libertarian arguments that do not assume either 1) economic freedom as the primary terminal value or 2) assume the efficiency of real-world markets? Both are unwarranted assumptions that seem to underlie many libertarian arguments I’ve seen.
Really? Respectfully, it seems much more plausible, based on the tone of your post, that you’re couching an appeal for your own preferred policy in hypothetical terms than that you’re actually suffering from a failure of imagination.
Well, yes, I prefer policies that are empirically demonstrated to actually work, especially when the cost of trying a system that fails is very high. Why don’t you?
Do you go looking for merit-worthy religious apologetics?
Yes. Diagnosing the faults in Alvin Plantinga’s reasoning is important. Am I to understand you’d prefer a frank exchange of views with Jerry Falwell?
That said, do you know of any libertarian arguments that do not assume either 1) economic freedom as the primary terminal value or 2) assume the efficiency of real-world markets? Both are unwarranted assumptions that seem to underlie many libertarian arguments I’ve seen.
Yes. I included one such argument in the post you just replied to. I quote myself:
One interestding claim [of policy libertarianism]: State actors are (made up of) people who are subject to the same irrational biases and collective stupidity as market actors, and often have perverse incentive structures as well.”
In other words, government decision-makers (i.e. bureaucrats) have just as much trouble integrating new information, violating social norms, and admitting error as consumers or decision-makers for firms, but bureaucrats are also subject to perverse incentives, regulatory capture, etc.
The implied primary terminal value here is welfare-maximization, according to some material standard that I’m assuming we could agree on, given that we’re both here. No specific claim about the efficiency of markets is made. A fortiori, the argument derives some of its strength from the acknowledgment of certain deviations from rational behavior that (once again) we both presumably know about, because we’re both here.
One interesting claim [of policy libertarianism]: State actors are (made up of) people who are subject to the same irrational biases and collective stupidity as market actors, and often have perverse incentive structures as well.”
My main complaint with this argument is that it should be empirically testable. You can implement regulatory scheme X in Area A, and no regulatory scheme in Area B, and see which produces better results. For example, ban all cancer treatments that top doctors agree are useless and dangerous in Area A, keep all treatments legal in Area B, and see which area has higher mortality among cancer patients.
Many libertarians I know have absolutely no interest in doing this, and don’t even like talking about the term “regulatory scheme X” because they prefer to lump all possible regulatory schemes together and judge them on the merit of the first one that comes to mind (this is also a problem with many socialists, for the opposite reason).
I don’t know much about economics, but I do know a bit about public health policy, and the people in charge of that are sometimes very good about using studies to determine whether their government interventions are an overall improvement over the no-intervention case (obvious exception: the FDA, which is very good at running studies, but very bad at running the right studies and doing sane cost-benefit analysis). When these studies show positive results at relatively low cost, a truly consequentialist libertarian ought to admit government regulation has been effective in that case. Instead, they tend to dismiss it as a fluke or start talking about some case where government regulation isn’t effective.
I think the great error in this whole debate is framing it as a conflict between socialists (who supposedly ought to think all government interventions are great) and libertarians (who supposedly ought to think all government interventions are terrible). In reality, some of these will work and some of these won’t. I’d rather people started paying more attention to which were which than become crusaders for bigger or smaller government. I think “Government regulation is bad” (or “is good”) is approximately the same kind of sentence as “Islam is a religion of peace”.
I’m reluctant to jump into a long discussion of the specifics of libertarian public policy—mind killer and all that—but in light of the terrible account of itself libertarianism has given you and SoullessAutomaton, maybe a few nonspecific comments are in order.
There’s such a thing as libertarian public policy research. It happens in think tanks. It gets done by academics (mostly economists), it incorporates peer review, and it usually doesn’t hold with the kind of boorish behavior you’re describing. Many of its hypotheticals are imports from the most inconvenient possible world. Specifically, it acknowledges that market failures exist and that government intervention is sometimes the most effective way to deal with them; that regulation has legitimate uses in service of the public good; and above all that pragmatism and compromise are the only virtues that can survive in the political arena.
Like most public policy it is essentially utilitarian, and its specific claims center around the idea that society is too complex for any central authority to administer efficiently. That’s to say, while there are many good ends the government might achieve through intervention in the economy or the private lives of its citizens, the costs of such intervention—money spent, conventions altered, expectations shifted, power grabbed, responsibility abdicated, and goals co-opted by the political process—are rarely less than the benefits.
You may take issue with any of these claims, but hopefully you can agree that the framework I’m developing here supports more sophisticated answers to the question “As a society, what should we do?” than just chanting “Private GOOD! Public BAD!”
In the specific case of your test of Regulatory Scheme X, the thoughtful libertarian position might go something like this:
Accountability is great. Empirical validation is great. But in this case, your test is a non-starter. No one is going to want this to happen. Drug companies will resist the removal of their products from the marketplace. Doctors will see the legislative call to dispense with certain treatments as a threat to their professional autonomy. Crossover between the AMA and FDA will favor an equilibrium where most experts already support the status quo. Insurance companies will use the opportunity to demand changes elsewhere in their payment structure. Any one of these groups can scuttle the whole project and throw your whole party out of power in the next election by letting it slip to the AARP that you’re planning on taking away something previously covered by Medicare. And even if you manage a legislative or executive miracle, anyone in Area A who wants the banned treatments can just migrate to Area B and get them there.
None of this should be interpreted to rule out the possibility that the test itself could yield invaluable information, saving lives or huge amounts of taxpayer money. But no regulator or legislator has any direct incentive to risk his career and all his political capital for nobody in particular. When talking about cost-benefit analysis, it’s important to remember that government officials implicitly measure costs and benefits to themselves, and that many of the responsibilities government arrogates to itself go unmet as a result.
I appreciate your presentation of the ideas here; it’s more enlightening than most material I’ve seen. That said, I still take issue with some points:
It happens in think tanks. It gets done by academics (mostly economists), it incorporates peer review, and it usually doesn’t hold with the kind of boorish behavior you’re describing.
Peer review by academics is only meaningful if based on a foundation of empirical observation and testable predictions; I think it remains to be demonstrated that macroeconomics has any predictive power whatsoever. Otherwise you end up with something like literary criticism—sophisticated, elaborate arguments unrelated to reality in nearly every way.
and its specific claims center around the idea that society is too complex for any central authority to administer efficiently
This does not require that certain subsections of society may benefit from central administration. This is easily demonstrated by the existence of large, niche-dominating corporations, which tend to be every bit as inefficient and bureaucratic as government. Some problems are such that the benefits of centralization outweigh the costs of bureaucracy.
hopefully you can agree that the framework I’m developing here supports more sophisticated answers to the question “As a society, what should we do?” than just chanting “Private GOOD! Public BAD!”
Agreed completely, but sophisticated answers are still useless without empirical foundations.
But in this case, your test is a non-starter. No one is going to want this to happen.
...which, ironically, leads to the other great flaw of libertarian politics—it proposes that government voluntarily reduce its own power dramatically and promotes increased personal responsibility. This is not a popular idea. People like having an authority, and people in government like being authorities. Large-scale libertarian government stikes me as, unfortunately, every bit as unlikely as Yvain’s government of controlled experimentation.
I think it remains to be demonstrated that macroeconomics has any predictive power whatsoever.
While I am pretty skeptical of the predictive power of a lot of macroeconomics as well, it seems odd to demand empirical research but simultaneously deny that the field in question is amenable to empirical research. A lot of the economics research that is used as support for libertarian positions is based on comparative studies between countries or between jurisdictions within countries. One common thread of research is to attempt to rate countries according to some defined measure of economic freedom and then see if the rating correlates with positive outcomes (GDP per capita being a common choice). There’s all sorts of ways that research can be criticized but to completely rule out such research as admissible evidence would seem to render questions about how to organize society completely beyond the realms of scientific investigation. If studies of this kind are not a valid basis for making decisions then how do you propose ideal enlightened governments should determine policy?
Large-scale libertarian government stikes me as, unfortunately, every bit as unlikely as Yvain’s government of controlled experimentation.
Hence the existence of things like the free state project and seasteading. Libertarians are quite aware of the difficulties of achieving their vision of society through conventional democratic politics. In fact, there’s a recently established blog on that very topic run in part by a less wrong commenter (who is also behind the seasteading idea).
While I am pretty skeptical of the predictive power of a lot of macroeconomics as well, it seems odd to demand empirical research but simultaneously deny that the field in question is amenable to empirical research.
I’m not saying it’s not amenable to empirical research, I just don’t get the impression that any extant research has been fruitful. As I said earlier, I saw serious economists discussing the USA stimulus bill predicting everything from “harmful” to “no effect” to “doesn’t do enough”, and as Robin has observed the chances that any of these predictions will be remembered is close to nil. This is strong evidence that the field as a whole lacks rigor, and that anyone who does know their stuff is being drowned out by the rest.
If studies of this kind are not a valid basis for making decisions then how do you propose ideal enlightened governments should determine policy?
Relying first and foremost on things that are already demonstrated to have worked well is a good start—hence my argument for adopting European-style socialized health care. Also, drop things that have been demonstrated ineffective, like the “war on drugs”.
Beyond that, take action only when necessary, and test new ideas in small areas first when possible. Modern governments are too large and powerful to be making large, expensive mistakes.
Hence the existence of things like the free state project and seasteading.
I’ve read about the seasteading project before, actually, and I think it’s generally a wonderful idea.
This is strong evidence that the field as a whole lacks rigor, and that anyone who does know their stuff is being drowned out by the rest.
I’m not sure what kind of economics you’re thinking of when you say macroeconomics. I have very little confidence in the kind of macroeconomics that tries to relate things like interest rates and savings to money supply using simple formulas, or that tries to give accurate values to ‘multipliers’ for stimulus spending, or that construct mathematical models of the economy and use them to make predictions about future growth from a few aggregated inputs by curve fitting to historical data. I’d agree that most of that is junk.
The kind of macroeconomics I think has some value is that which attempts to gather empirical support for particular policies by comparing outcomes across different countries or jurisdictions, or across different time periods. This kind of research is obviously far from ideal since it is weakly controlled and is often making hard to justify comparisons between different measures. Maybe macroeconomics is not the right term for that kind of research but I’m not sure what else to call it. In terms of gathering empirical evidence for the results of particular policy decisions, that seems about the best we can do at the moment.
Relying first and foremost on things that are already demonstrated to have worked well is a good start—hence my argument for adopting European-style socialized health care.
You do realize how far from an uncontroversial claim that is? I grew up with the NHS in the UK (and I now live in Canada which is also a universal health care system). They are far from perfect systems. Every government for as long as I can remember in the UK has come into power partly on a promise to ‘fix’ the NHS. None have succeeded. I don’t think I’ve heard anyone argue that the healthcare system in the US is fine as it is—there is fairly universal agreement that it is broken. I’m far from persuaded that the best solution is to try and adopt a European model though. There’s plenty of research from libertarian think tanks that provides empirical evidence in favour of health care reforms that reduce government involvement in healthcare rather than increase it. I’m sure there are grounds for questioning some of that research but it is disingenuous to pretend it doesn’t exist.
Beyond that, take action only when necessary, and test new ideas in small areas first when possible. Modern governments are too large and powerful to be making large, expensive mistakes.
And healthcare is a good candidate for the largest and most expensive of them all. Why is healthcare exempt from the principle of testing new ideas on a small scale first?
I’m not sure what kind of economics you’re thinking of when you say macroeconomics.
Primarily the first kind you describe.
In terms of gathering empirical evidence for the results of particular policy decisions, that seems about the best we can do at the moment.
It is, but it is unfortunately limited to examining the results of policies already implemented; anyone justifying novel solutions based on such evidence is likely veering more into the territory of what we agree is junk, and probably making very dubious assumptions about ability to extrapolate trends.
You do realize how far from an uncontroversial claim that is? I grew up with the NHS in the UK (and I now live in Canada which is also a universal health care system). They are far from perfect systems.
Of course, but for all their imperfections they are widely recognized as better than what the USA currently does and win out on almost every objective metric. I also seem to recall that the UK and Canada are often considered some of the worst other than the USA, though at least they spend less than half as much than the US does.
Socialized health care, like democracy, can generally be categorized under “the worst system, except for all the others that have been tried”.
There’s plenty of research from libertarian think tanks that provides empirical evidence in favour of health care reforms that reduce government involvement in healthcare rather than increase it.
On basis of what observations? To my knowledge all other developed nations employ some form of socialized health care.
As it stands now, the USA is the biggest outlier and getting the worst results; an obvious case for applying a little majoritarianism.
And healthcare is a good candidate for the largest and most expensive of them all. Why is healthcare exempt from the principle of testing new ideas on a small scale first?
Because it’s not a new idea. Everyone else has helpfully tried it out for us already and found that it basically works.
On the other hand, aggressively de-regulating and getting government out of health care is, as far as I know, completely untried and untested.
Of course, but for all their imperfections they are widely recognized as better than what the USA currently does and win out on almost every objective metric.
The superiority of the Canadian and British models is not uncontroversial. This policy analysis rebuts many of the arguments for example. It includes numerous objective metrics on which the US does better than Canada or the UK. Debate the claims if you want but don’t pretend the issue is settled.
On basis of what observations? To my knowledge all other developed nations employ some form of socialized health care.
There is a lot of variation between nations. Many nations that have some form of ‘socialized’ health care also have significant amounts of private health care. Many countries have introduced market based reforms within their socialized health care systems in an effort to improve efficiency. Health tourism is increasingly common in Europe.
Because it’s not a new idea. Everyone else has helpfully tried it out for us already and found that it basically works.
‘It’ is not a single idea or system. Ultimately the biggest problem with healthcare reform in the US is that there is very little chance that it will adopt the best practices found in other nations. There are too many powerful special interests for the political process to select policies based on effectiveness. The more government involvement there is in healthcare, the more healthcare will be subject to problems of regulatory capture, special interest lobbying, rent seeking and bureaucracy. De-regulation and reduced government involvement creates incentives for serving patient interests as a primary route to success. Increased regulation and government involvement means that it becomes more and more profitable to focus effort on lobbying, gaming the system and political maneuvering rather than on patient care.
Debate the claims if you want but don’t pretend the issue is settled.
I don’t have the time to respond properly to the linked PDF, but skimming it quickly it doesn’t seem particularly persuasive. Obviously, the argument isn’t “settled” because people still argue about it, but that doesn’t mean both sides have arguments of equal strength.
At any rate, I’m really going to have to drop the discussion at this point because I don’t have the time to go digging up supporting references. If you seriously think that the US system is of comparable quality to European systems our difference of perspective is far too vast to bridge by simple off-the-cuff arguments here.
Thanks for your time, though, this has been enjoyable.
they are widely recognized as better than what the USA currently does and win out on almost every objective metric
The linked document contains a number of objective metrics on which the US does better—waiting times, use of high tech surgical procedures, access to high tech diagnostic equipment, breast and prostate cancer mortality ratios, specialist to patient ratios and patient satisfaction measures. I linked it as evidence to rebut the specific claim that the US is worse on ‘almost every objective metric’.
I don’t have the time to respond properly to the linked PDF, but skimming it quickly it doesn’t seem particularly persuasive.
I’m not asking you to make a detailed rebuttal, I’m just providing evidence of objective measures on which the US does better. I don’t have time for a detailed debate either. You seem to assign an extremely high probability to the belief that healthcare in the US could be significantly improved by adopting a European system though and I’m questioning whether the evidence justifies such high confidence.
Ultimately the only reason this is even a political issue is because of the high levels of government involvement. With less government involvement people could spend their healthcare dollars in the way they thought best without having to persuade anyone else. That’s another reason I oppose high levels of government involvement.
Now the question is, does advocating one regulatory scheme make other regulatory schemes more likely, through some habit-forming mechanism or other? If so, then your version of a libertarian (who thinks the average scheme is bad) should sometimes oppose even good schemes, and your version of a socialist (who thinks the average scheme is good) should sometimes support even bad schemes.
In building bridges between left and right, it’s always a good idea to offer analogies between money and sex, so consider this: utilitarians generally oppose governments telling people whom to mate with and whom not to mate with, even in cases where these people will predictably make decisions that make them and others unhappy, because utilitarians think the good this would do is outweighed by the value of having a bright-line taboo against the government meddling in that sphere. It’s less obvious to me than it is to you that the economy as a whole, or some particular circumscribed aspect of it, isn’t also such a sphere, at least in part.
Since there’s no good that could possibly come from us talking about this other than low-quality thinking and writing practice, I’m putting this sentence here to make myself look like an idiot in case I fail to resist the temptation to post about politics again anytime soon.
Yes. Diagnosing the faults in Alvin Plantinga’s reasoning is important. Am I to understand you’d prefer a frank exchange of views with Jerry Falwell?
I see little value to discussion with either. Given the fundamental problems with theism (primarily a lack of empiricism), I can reasonably expect that “better” theist arguments will only be more elaborate and rigorous argumentation from the same broken axioms. Unless I were preparing for a formal, public debate with a theist it wouldn’t be worth my time.
Yes. I included one such argument in the post you just replied to. I quote myself:
The phrase “policy libertarianism” gets only a couple thousand google hits, many of which are false positives. Eliminating the most common false positive (the phrase “foreign policy, libertarianism”), your own LW comment comes up on the first page of results. The remaining results seem to mostly concern matters of evolutionary vs. revolutionary change as a means of implementing libertarian goals, not arguments for goals, and bear no obvious relation to what you’ve mentioned. If you’re referring to a major school of libertarian thought, I’m assuming there’s another, more popular term for it, but I don’t know how to figure out what it would be, sorry.
In other words, government decision-makers (i.e. bureaucrats) have just as much trouble integrating new information, violating social norms, and admitting error as consumers or decision-makers for firms, but bureaucrats are also subject to perverse incentives, regulatory capture, etc.
This point is not under dispute, but it also does not suffice to prove that governmental action is therefore less effective, especially given imperfect markets, other problematic incentives for smaller agents (e.g., problems of collective action), and empirical evidence showing that governmental programs can sometimes lead to better results than non-governmental programs (e.g., European vs. USA health care).
The implied primary terminal value here is welfare-maximization, according to some material standard that I’m assuming we could agree on, given that we’re both here. No specific claim about the efficiency of markets is made. A fortiori, the argument derives some of its strength from the acknowledgment of certain deviations from rational behavior that (once again) we both presumably know about, because we’re both here.
Some form of welfare-maximization, yes. Various quality-of-life metrics are a (rough) approximation. I’m not sure what else you’re getting at here.
I should say again, it is likely that we agree on the vast majority of actual conclusions. My complaint with libertarianism as a political philosophy is that (like most other political philosophies) it has no apparent, consistent empirical basis and resorts frequently to bottom-line arguments, even though many of their conclusions can be justified rigorously.
I am convinced of this in large part because most libertarians I have encountered have been completely impervious to real-world examples of government programs being more efficient and effective than equivalent non-governmental systems, leading me to conclude that the mainstream of libertarian thought is essentially anti-empirical.
The phrase “policy libertarianism” gets only a couple thousand google hits, many of which are false positives.
It’s a fairly recently coined term. The first use I’m aware of is here. The distinction between policy and structural libertarianism has been picked up quite quickly as many have found it useful.
In that case I remain confused in that it seems to mostly refer to a framework for how to achieve libertarian goals, not for justifying that said goals will successfully confer the advertised benefits (e.g., some type of welfare maximization or general utility).
The post is on a libertarian blog and as such is aimed at an audience who already accept the case for libertarian goals being desirable. It’s making a distinction between libertarians who believe that the best way to achieve their goals is to work within existing democratic systems to promote libertarian policies and those who believe that existing democratic systems are fundamentally inhospitable to libertarian policies and that achieving libertarianism requires addressing structural factors that tend to produce unlibertarian societies.
The ‘policy libertarians’ tend to be the ones focusing on demonstrating empirical support for improved outcomes under libertarian policies. The idea being that it may be possible to get more libertarian policies implemented by appealing to empirical evidence for their efficacy on a case by case basis rather than by trying to convince people that libertarianism is the ‘one true way’. That would seem to be precisely the kind of approach you would seem to prefer.
Uhm, point 2 at least is a straight up fact, as real markets typically diverge to varying degrees from perfect markets. I also note you don’t actually dispute point 1, which is the strong statement of a deontological ethical position, and pure deontology remains incompatible with consequentialism, hence the apparent contradiction in ethical systems.
If I’ve been unimpressed with the arguments of rank-and-file members of a political position, why would I be motivated to look for better writing that may or may not exist? Do you go looking for merit-worthy religious apologetics?
That said, do you know of any libertarian arguments that do not assume either 1) economic freedom as the primary terminal value or 2) assume the efficiency of real-world markets? Both are unwarranted assumptions that seem to underlie many libertarian arguments I’ve seen.
Well, yes, I prefer policies that are empirically demonstrated to actually work, especially when the cost of trying a system that fails is very high. Why don’t you?
Yes. Diagnosing the faults in Alvin Plantinga’s reasoning is important. Am I to understand you’d prefer a frank exchange of views with Jerry Falwell?
Yes. I included one such argument in the post you just replied to. I quote myself:
In other words, government decision-makers (i.e. bureaucrats) have just as much trouble integrating new information, violating social norms, and admitting error as consumers or decision-makers for firms, but bureaucrats are also subject to perverse incentives, regulatory capture, etc.
The implied primary terminal value here is welfare-maximization, according to some material standard that I’m assuming we could agree on, given that we’re both here. No specific claim about the efficiency of markets is made. A fortiori, the argument derives some of its strength from the acknowledgment of certain deviations from rational behavior that (once again) we both presumably know about, because we’re both here.
My main complaint with this argument is that it should be empirically testable. You can implement regulatory scheme X in Area A, and no regulatory scheme in Area B, and see which produces better results. For example, ban all cancer treatments that top doctors agree are useless and dangerous in Area A, keep all treatments legal in Area B, and see which area has higher mortality among cancer patients.
Many libertarians I know have absolutely no interest in doing this, and don’t even like talking about the term “regulatory scheme X” because they prefer to lump all possible regulatory schemes together and judge them on the merit of the first one that comes to mind (this is also a problem with many socialists, for the opposite reason).
I don’t know much about economics, but I do know a bit about public health policy, and the people in charge of that are sometimes very good about using studies to determine whether their government interventions are an overall improvement over the no-intervention case (obvious exception: the FDA, which is very good at running studies, but very bad at running the right studies and doing sane cost-benefit analysis). When these studies show positive results at relatively low cost, a truly consequentialist libertarian ought to admit government regulation has been effective in that case. Instead, they tend to dismiss it as a fluke or start talking about some case where government regulation isn’t effective.
I think the great error in this whole debate is framing it as a conflict between socialists (who supposedly ought to think all government interventions are great) and libertarians (who supposedly ought to think all government interventions are terrible). In reality, some of these will work and some of these won’t. I’d rather people started paying more attention to which were which than become crusaders for bigger or smaller government. I think “Government regulation is bad” (or “is good”) is approximately the same kind of sentence as “Islam is a religion of peace”.
I’m reluctant to jump into a long discussion of the specifics of libertarian public policy—mind killer and all that—but in light of the terrible account of itself libertarianism has given you and SoullessAutomaton, maybe a few nonspecific comments are in order.
There’s such a thing as libertarian public policy research. It happens in think tanks. It gets done by academics (mostly economists), it incorporates peer review, and it usually doesn’t hold with the kind of boorish behavior you’re describing. Many of its hypotheticals are imports from the most inconvenient possible world. Specifically, it acknowledges that market failures exist and that government intervention is sometimes the most effective way to deal with them; that regulation has legitimate uses in service of the public good; and above all that pragmatism and compromise are the only virtues that can survive in the political arena.
Like most public policy it is essentially utilitarian, and its specific claims center around the idea that society is too complex for any central authority to administer efficiently. That’s to say, while there are many good ends the government might achieve through intervention in the economy or the private lives of its citizens, the costs of such intervention—money spent, conventions altered, expectations shifted, power grabbed, responsibility abdicated, and goals co-opted by the political process—are rarely less than the benefits.
You may take issue with any of these claims, but hopefully you can agree that the framework I’m developing here supports more sophisticated answers to the question “As a society, what should we do?” than just chanting “Private GOOD! Public BAD!”
In the specific case of your test of Regulatory Scheme X, the thoughtful libertarian position might go something like this:
Accountability is great. Empirical validation is great. But in this case, your test is a non-starter. No one is going to want this to happen. Drug companies will resist the removal of their products from the marketplace. Doctors will see the legislative call to dispense with certain treatments as a threat to their professional autonomy. Crossover between the AMA and FDA will favor an equilibrium where most experts already support the status quo. Insurance companies will use the opportunity to demand changes elsewhere in their payment structure. Any one of these groups can scuttle the whole project and throw your whole party out of power in the next election by letting it slip to the AARP that you’re planning on taking away something previously covered by Medicare. And even if you manage a legislative or executive miracle, anyone in Area A who wants the banned treatments can just migrate to Area B and get them there.
None of this should be interpreted to rule out the possibility that the test itself could yield invaluable information, saving lives or huge amounts of taxpayer money. But no regulator or legislator has any direct incentive to risk his career and all his political capital for nobody in particular. When talking about cost-benefit analysis, it’s important to remember that government officials implicitly measure costs and benefits to themselves, and that many of the responsibilities government arrogates to itself go unmet as a result.
I appreciate your presentation of the ideas here; it’s more enlightening than most material I’ve seen. That said, I still take issue with some points:
Peer review by academics is only meaningful if based on a foundation of empirical observation and testable predictions; I think it remains to be demonstrated that macroeconomics has any predictive power whatsoever. Otherwise you end up with something like literary criticism—sophisticated, elaborate arguments unrelated to reality in nearly every way.
This does not require that certain subsections of society may benefit from central administration. This is easily demonstrated by the existence of large, niche-dominating corporations, which tend to be every bit as inefficient and bureaucratic as government. Some problems are such that the benefits of centralization outweigh the costs of bureaucracy.
Agreed completely, but sophisticated answers are still useless without empirical foundations.
...which, ironically, leads to the other great flaw of libertarian politics—it proposes that government voluntarily reduce its own power dramatically and promotes increased personal responsibility. This is not a popular idea. People like having an authority, and people in government like being authorities. Large-scale libertarian government stikes me as, unfortunately, every bit as unlikely as Yvain’s government of controlled experimentation.
While I am pretty skeptical of the predictive power of a lot of macroeconomics as well, it seems odd to demand empirical research but simultaneously deny that the field in question is amenable to empirical research. A lot of the economics research that is used as support for libertarian positions is based on comparative studies between countries or between jurisdictions within countries. One common thread of research is to attempt to rate countries according to some defined measure of economic freedom and then see if the rating correlates with positive outcomes (GDP per capita being a common choice). There’s all sorts of ways that research can be criticized but to completely rule out such research as admissible evidence would seem to render questions about how to organize society completely beyond the realms of scientific investigation. If studies of this kind are not a valid basis for making decisions then how do you propose ideal enlightened governments should determine policy?
Hence the existence of things like the free state project and seasteading. Libertarians are quite aware of the difficulties of achieving their vision of society through conventional democratic politics. In fact, there’s a recently established blog on that very topic run in part by a less wrong commenter (who is also behind the seasteading idea).
I’m not saying it’s not amenable to empirical research, I just don’t get the impression that any extant research has been fruitful. As I said earlier, I saw serious economists discussing the USA stimulus bill predicting everything from “harmful” to “no effect” to “doesn’t do enough”, and as Robin has observed the chances that any of these predictions will be remembered is close to nil. This is strong evidence that the field as a whole lacks rigor, and that anyone who does know their stuff is being drowned out by the rest.
Relying first and foremost on things that are already demonstrated to have worked well is a good start—hence my argument for adopting European-style socialized health care. Also, drop things that have been demonstrated ineffective, like the “war on drugs”.
Beyond that, take action only when necessary, and test new ideas in small areas first when possible. Modern governments are too large and powerful to be making large, expensive mistakes.
I’ve read about the seasteading project before, actually, and I think it’s generally a wonderful idea.
I’m not sure what kind of economics you’re thinking of when you say macroeconomics. I have very little confidence in the kind of macroeconomics that tries to relate things like interest rates and savings to money supply using simple formulas, or that tries to give accurate values to ‘multipliers’ for stimulus spending, or that construct mathematical models of the economy and use them to make predictions about future growth from a few aggregated inputs by curve fitting to historical data. I’d agree that most of that is junk.
The kind of macroeconomics I think has some value is that which attempts to gather empirical support for particular policies by comparing outcomes across different countries or jurisdictions, or across different time periods. This kind of research is obviously far from ideal since it is weakly controlled and is often making hard to justify comparisons between different measures. Maybe macroeconomics is not the right term for that kind of research but I’m not sure what else to call it. In terms of gathering empirical evidence for the results of particular policy decisions, that seems about the best we can do at the moment.
You do realize how far from an uncontroversial claim that is? I grew up with the NHS in the UK (and I now live in Canada which is also a universal health care system). They are far from perfect systems. Every government for as long as I can remember in the UK has come into power partly on a promise to ‘fix’ the NHS. None have succeeded. I don’t think I’ve heard anyone argue that the healthcare system in the US is fine as it is—there is fairly universal agreement that it is broken. I’m far from persuaded that the best solution is to try and adopt a European model though. There’s plenty of research from libertarian think tanks that provides empirical evidence in favour of health care reforms that reduce government involvement in healthcare rather than increase it. I’m sure there are grounds for questioning some of that research but it is disingenuous to pretend it doesn’t exist.
And healthcare is a good candidate for the largest and most expensive of them all. Why is healthcare exempt from the principle of testing new ideas on a small scale first?
Primarily the first kind you describe.
It is, but it is unfortunately limited to examining the results of policies already implemented; anyone justifying novel solutions based on such evidence is likely veering more into the territory of what we agree is junk, and probably making very dubious assumptions about ability to extrapolate trends.
Of course, but for all their imperfections they are widely recognized as better than what the USA currently does and win out on almost every objective metric. I also seem to recall that the UK and Canada are often considered some of the worst other than the USA, though at least they spend less than half as much than the US does.
Socialized health care, like democracy, can generally be categorized under “the worst system, except for all the others that have been tried”.
On basis of what observations? To my knowledge all other developed nations employ some form of socialized health care.
As it stands now, the USA is the biggest outlier and getting the worst results; an obvious case for applying a little majoritarianism.
Because it’s not a new idea. Everyone else has helpfully tried it out for us already and found that it basically works.
On the other hand, aggressively de-regulating and getting government out of health care is, as far as I know, completely untried and untested.
The superiority of the Canadian and British models is not uncontroversial. This policy analysis rebuts many of the arguments for example. It includes numerous objective metrics on which the US does better than Canada or the UK. Debate the claims if you want but don’t pretend the issue is settled.
There is a lot of variation between nations. Many nations that have some form of ‘socialized’ health care also have significant amounts of private health care. Many countries have introduced market based reforms within their socialized health care systems in an effort to improve efficiency. Health tourism is increasingly common in Europe.
‘It’ is not a single idea or system. Ultimately the biggest problem with healthcare reform in the US is that there is very little chance that it will adopt the best practices found in other nations. There are too many powerful special interests for the political process to select policies based on effectiveness. The more government involvement there is in healthcare, the more healthcare will be subject to problems of regulatory capture, special interest lobbying, rent seeking and bureaucracy. De-regulation and reduced government involvement creates incentives for serving patient interests as a primary route to success. Increased regulation and government involvement means that it becomes more and more profitable to focus effort on lobbying, gaming the system and political maneuvering rather than on patient care.
I don’t have the time to respond properly to the linked PDF, but skimming it quickly it doesn’t seem particularly persuasive. Obviously, the argument isn’t “settled” because people still argue about it, but that doesn’t mean both sides have arguments of equal strength.
At any rate, I’m really going to have to drop the discussion at this point because I don’t have the time to go digging up supporting references. If you seriously think that the US system is of comparable quality to European systems our difference of perspective is far too vast to bridge by simple off-the-cuff arguments here.
Thanks for your time, though, this has been enjoyable.
The linked document contains a number of objective metrics on which the US does better—waiting times, use of high tech surgical procedures, access to high tech diagnostic equipment, breast and prostate cancer mortality ratios, specialist to patient ratios and patient satisfaction measures. I linked it as evidence to rebut the specific claim that the US is worse on ‘almost every objective metric’.
I’m not asking you to make a detailed rebuttal, I’m just providing evidence of objective measures on which the US does better. I don’t have time for a detailed debate either. You seem to assign an extremely high probability to the belief that healthcare in the US could be significantly improved by adopting a European system though and I’m questioning whether the evidence justifies such high confidence.
Ultimately the only reason this is even a political issue is because of the high levels of government involvement. With less government involvement people could spend their healthcare dollars in the way they thought best without having to persuade anyone else. That’s another reason I oppose high levels of government involvement.
This is by far the most useful discussion of the subject I have ever had. I’m starting to think this rationality stuff might actually work out.
Now the question is, does advocating one regulatory scheme make other regulatory schemes more likely, through some habit-forming mechanism or other? If so, then your version of a libertarian (who thinks the average scheme is bad) should sometimes oppose even good schemes, and your version of a socialist (who thinks the average scheme is good) should sometimes support even bad schemes.
In building bridges between left and right, it’s always a good idea to offer analogies between money and sex, so consider this: utilitarians generally oppose governments telling people whom to mate with and whom not to mate with, even in cases where these people will predictably make decisions that make them and others unhappy, because utilitarians think the good this would do is outweighed by the value of having a bright-line taboo against the government meddling in that sphere. It’s less obvious to me than it is to you that the economy as a whole, or some particular circumscribed aspect of it, isn’t also such a sphere, at least in part.
Since there’s no good that could possibly come from us talking about this other than low-quality thinking and writing practice, I’m putting this sentence here to make myself look like an idiot in case I fail to resist the temptation to post about politics again anytime soon.
I see little value to discussion with either. Given the fundamental problems with theism (primarily a lack of empiricism), I can reasonably expect that “better” theist arguments will only be more elaborate and rigorous argumentation from the same broken axioms. Unless I were preparing for a formal, public debate with a theist it wouldn’t be worth my time.
The phrase “policy libertarianism” gets only a couple thousand google hits, many of which are false positives. Eliminating the most common false positive (the phrase “foreign policy, libertarianism”), your own LW comment comes up on the first page of results. The remaining results seem to mostly concern matters of evolutionary vs. revolutionary change as a means of implementing libertarian goals, not arguments for goals, and bear no obvious relation to what you’ve mentioned. If you’re referring to a major school of libertarian thought, I’m assuming there’s another, more popular term for it, but I don’t know how to figure out what it would be, sorry.
This point is not under dispute, but it also does not suffice to prove that governmental action is therefore less effective, especially given imperfect markets, other problematic incentives for smaller agents (e.g., problems of collective action), and empirical evidence showing that governmental programs can sometimes lead to better results than non-governmental programs (e.g., European vs. USA health care).
Some form of welfare-maximization, yes. Various quality-of-life metrics are a (rough) approximation. I’m not sure what else you’re getting at here.
I should say again, it is likely that we agree on the vast majority of actual conclusions. My complaint with libertarianism as a political philosophy is that (like most other political philosophies) it has no apparent, consistent empirical basis and resorts frequently to bottom-line arguments, even though many of their conclusions can be justified rigorously.
I am convinced of this in large part because most libertarians I have encountered have been completely impervious to real-world examples of government programs being more efficient and effective than equivalent non-governmental systems, leading me to conclude that the mainstream of libertarian thought is essentially anti-empirical.
It’s a fairly recently coined term. The first use I’m aware of is here. The distinction between policy and structural libertarianism has been picked up quite quickly as many have found it useful.
In that case I remain confused in that it seems to mostly refer to a framework for how to achieve libertarian goals, not for justifying that said goals will successfully confer the advertised benefits (e.g., some type of welfare maximization or general utility).
The post is on a libertarian blog and as such is aimed at an audience who already accept the case for libertarian goals being desirable. It’s making a distinction between libertarians who believe that the best way to achieve their goals is to work within existing democratic systems to promote libertarian policies and those who believe that existing democratic systems are fundamentally inhospitable to libertarian policies and that achieving libertarianism requires addressing structural factors that tend to produce unlibertarian societies.
The ‘policy libertarians’ tend to be the ones focusing on demonstrating empirical support for improved outcomes under libertarian policies. The idea being that it may be possible to get more libertarian policies implemented by appealing to empirical evidence for their efficacy on a case by case basis rather than by trying to convince people that libertarianism is the ‘one true way’. That would seem to be precisely the kind of approach you would seem to prefer.
That clarifies the relevance, thank you.