= 27d567bc2d48d9f40e9ccc20ea370b15)
JoshuaZ comments on Open Thread: June 2010 - Less Wrong
You are viewing a comment permalink. View the original post to see all comments and the full post content.
= 783df68a0f980790206b9ea87794c5b6)
Loading…= b715d02e85f7b3b4ae65ca3d3e450868)
Subscribe to RSS Feed
Powered by Reddit
= f037147d6e6c911a85753b9abdedda8d)
You are viewing a comment permalink. View the original post to see all comments and the full post content.
Comments (651)
Simply responding with a Randian quote doesn't show that government doesn't work. Moreover, there are some things where government has worked well. At the most basic level, one needs governments to protect property rights, without which markets can't function. Similarly, various forms of pooled goods are useful (you are welcome to try to have roads run by private industry and see how well that works) But even beyond that, government policies are helpful for dealing with negative externalities. In particular, some forms of harm are by nature spread out and not connected strongly to any single source. The classic example is pollution. Since pollution is spread out, the transaction cost is prohibitively high for any given individual to try to reduce pollution levels they are subject to. But a government, using regulation and careful taxation, can do this efficiently. In some situations, this can even be done in conjunction with market forces (such as cap and trade systems). In the US, this was very successful in efficiently handling levels of sulfur dioxide. See this paper. Governments are often slow and inefficient. But to claim that well-thought out policies never exist? That's simply at odds with reality.
Even from a libertarian point of view, pollution is something that causes harm, like murder or theft. The governments job is to enforce laws that mitigate sources of harm and, when possible, correct harms against individuals. A person or corporation who puts out some amount of pollution should be forced to pay for any clean up or harm that they make.
If you drive a car, you emmitted some fraction of the pollution that caused temperatures to go up, caused smog induced illness and some other miscellaneous harms that cost some amount of money. If that amount of money was 40 billion dollars, and you contributed 1 billionth towards the harm, you sshould pay 40 dollars.
This should be even less controversial than imprisoning murderers
Sadly it isn't. I consider(ed) myself libertarian, and then found that most self-identified ones reject that reasoning entirely. Pity.
I was also unpleasantly suprised to find that there was a group of people griping about programs that would make it easier to identify cars that weren't liability-insured or pollution-tested, and this was called a "libertarian" position.
ETA: And libertarian-leaning academics don't seem to "get" why paying polluters to go away isn't a solution, and don't even understand what problem is supposed to be solved, even when hypothetically placed in such a situation! (See the exchange between me and Hanson in the link.)
ETA2: I edited an EDF graphic to make this cute picture about the pollution issue and Coasean reasoning. ETA3: Full blog post with original graphic
It's not so much that it doesn't solve the problem as things just don't work that way. For starters, current energy distribution methods are local monopolies, so they are strongly regulated on price because the competition mechanism doesn't work as it should. The idea that a customers might "choose" cleaner energy doesn't always work.
Second, some logging companies tried that. They had an outside company, come in, do an inspection, and certify the ecological viability of their practices. There were a fair number of people who actually were willing to pay a little more. The problem is, another set of companies came by, inspected and approved themselves (with a different label that they invented) , and customers weren't able to tell the difference. That's a problem.
Also, to a great extent, electricity is fungible. Suppose you have both windmills and coal-fired plants connected to the same electrical grid, and they both generate equal amounts of power. Now suppose I tell the electric company that I only want to buy power from the windmills, so instead of getting half wind power and half coal power, I get 100% wind power (on paper). However, the electric company doesn't actually have to change the way it produces electricity in order to do this. All they have to do slightly increase the percentage of coal power that they deliver to everyone else (on paper). So all that changes is numbers on paper, and there's exactly as much coal power being generated as before.
Your noise pollution example is a potentially problematic one for libertarians but the obvious answer that occurs to me is the one I would expect many thoughtful libertarians to make. You are assuming a libertarian world with largely unchanged amounts of public space which is a problematic combination. The space outside your window has no reason to be public space. You would see a lot more 'gated community' type arrangements in a more libertarian society. People with low noise tolerance could choose to live in communities where the 'public' space was owned by a municipal service provider with strict rules about noise pollution. Anyone not adhering to these rules could be ejected from the property.
Many common problems with imagined libertarian societies dissolve when you allow for much greater private ownership of currently public land than currently exists.
What's the difference between a government and a "municipal service provider"?
It's easier to move out? You are not born under a landlord. You do not swear fealty to the flag of the landlord. Nobody thinks the landlord should be able to draft you for civil service. The landlord cannot put you in jail for failing to pay rent. There's a long, long list of other differences where the landlord as government analogy breaks down. I'm surprised anyone still brings it up.
EDIT: Ha. You changed it. In reality, not necessarily that much, although it's nice to have extra governmental agency that you can choose to pay or not, and that is accountable to the government in a transparent way. Asking the government to regulate itself is almost as dumb as asking a logging company to regulate itself.
Well, If you expect a landlord to perform the functions of a government, by, say, regulating noise levels for the benefit of tenants, then doesn't the analogy hold in this particular case? If regulation is bad, does it matter if it's regulation by landlord or regulation by city council?
It does matter if one has guns (or SWAT teams) and the other relies on non-violent persuasion.
::does some Googling::
If a landlord tries to have you evicted, and you refuse to leave when a court rules that you must do so, local law enforcement is allowed to physically remove you from the property. That doesn't sound non-violent to me.
This is a fair point. I would note, however, that eviction typically requires repeated notification, and opportunities for you modify your behavior before encountering violence.
Contrast with how your local sheriff can bust down your door in the middle of the night, shoot your dogs, destroy your property, and arrest you merely for suspecting you of possessing marijuana. And then be praised for it even if you are innocent.
Municipal services are generally provided by a local government but this is largely an artifact of the way modern democracies are organized. Private arrangements are fairly rare in the modern world but cruise ships, private resorts, corporate campuses and on a smaller scale large managed apartment buildings provide examples of decoupling the idea of provision of municipal services and government.
What if you had a dozen different companies that provided services like that. They would have a monopoly in different areas, however, the local governments would still be able to choose which one they wanted, and at any time they were displeased they could switch. Actually, this is a good idea!
You can probably go further than that. Municipal services can be unbundled and can operate without a geographical monopoly. This is already widely done for cable and telecoms in the US and UK and for electricity and gas in the UK. Some countries do it for water and sanitation services. There are examples worldwide of it being done for transportation, refuse collection, health and education. Arguments that such services are a 'natural monopoly' are usually promoted most strongly by those who wish to operate that monopoly with government protection.
Allow me to rephrase.
If the "municipal service provider" has the power to enforce its edicts on noise level (because it has the power to exile those who violate them), then doesn't that mean that it has exactly the same power over noise that a government would - and the same potential to misuse that power?
I tend to think that the right of exit is the ultimate and fundamental check on such abuses of power. This is why I favour decentralization / federalization / devolution as improvements to the status quo of increasing centralization of political power. I think that on more or less every level of government we would benefit from decentralization of power. City-wide bylaws on noise pollution are too coarse-grained for example. An entertainment district or an area popular with students should have different standards than a residential area with many working families. Zoning rules are an attempt to make such allowances but I think private solutions are likely to work better. I'd at least like to see them tried so we can start to see what works.
So the issue is that of scale, then?
And the right of exit is conditional on there being somewhere to go. Finding such a place can sometimes be difficult.
It worked out pretty well for the US.
Hitler was kind to animals. Even accepting your dubious claim it is not enough to show that government sometimes achieves positive outcomes (and don't forget to ask what criteria are being used to determine 'positive'). The relevant question is whether government intervention produces an overall net benefit. Generally it seems you can make the strongest case for this in small, relatively homogeneous countries. These results do not necessarily scale.
There are an awful lot of hidden assumptions in this statement.
Can in theory and ever actually do in practice are worlds apart. Negative externalities are one of the stronger economic arguments for government intervention but actual examples of government regulation rarely approximate the theoretical regulatory framework proposed by economists. This is largely because the behaviour of governments is determined primarily by public choice theory and not by the benevolent, enlightened pursuit of economic rationality.
I agree with most of what you said. That's one of the reasons I gave the historical example of SO2. The claim being made by the person I was responding to was not a remark about net gain but the claim that regarding "Good quality government policy" that "There is no more evidence for that than there is for God" and then backing it up with an argument from irrelevant authority. So giving examples to show that's not the case accomplishes the basic goal.
There's a pretty good precedent for this happening in the form of the railway system in early America. I think I'd classify it as a market failure as private roads and railways have a way of becoming local monopolies and having an enormous advantage when it comes to rent-seeking behavior.
It's not that it's impossible, I just don't think it's a very good idea.
One of the hidden assumptions I was thinking of is the assumption that government built roads have been a net benefit for America. The highway system has been a large implicit subsidy for all kinds of business models and lifestyle choices that are not obviously optimal. America's dependence on oil and outsize energy demands are in large part a function of the incentives created by huge government expenditure on highways. Suburban sprawl, McMansions, retail parks and long commutes are all unintended consequences of the implicit subsidies inherent in large scale government road construction.
American culture and society would probably look quite different without a history of government road construction. It's not obvious to me that it would not look better by many measures.
Yes but you'd be stuck with complex and inefficent series of toll roads. It might work, but I doubt. Not efficiently anyways.
Not necessarily. If you've ever been to Disney World, it's not like that. And hell, government roads in the states and Japan often dissolve into a complex and inefficient series of toll roads, at least in some areas.
I'm much more worried about uncompetitive practices, like powerful local monopolies and rent seeking behavior.
Disney world owns the land, they can do whatever they want. But here in order to make efficient roads, we have to use eminent domain. A private company wouldn't be able to do that. In order to have a governmentless society, you have to a) create a nearly impossible to maintain system of total anarchy like exists in parts of Afghanistan today or b) create a very corrupt and broken society ruled by private corporations, which is essentially a government anyways.
The Kelo case allows government to use its eminent domain powers on the behalf of private companies. Why couldn't a private road builder borrow this government power?
You actually support the Kelo case? To me that's like a Glenn Beck conspiracy theory come to life.
Yup. Mind killed. I'm out, guys. Was fun while it lasted.
Why do you assume I support the Court's decision? All I did was state that under current United States law, Houshalter's objection was possible to overcome.
The government does use private contracters in many cases for different projects. It might work on roads, I'm not sure if they already use it, but its still alot differnet from asking a private corporation to decide when and where to build roads.
They do. And private corporations or councils already decide where to build the roads for some things, it's just that all of those things only work if they're already connected to other infrastructure, which, in the US, means public federal, state and locally built roads.
Well, I think you aren't really imaginative enough in your view of anarchy, but... I'm not an anarchist and I'm not going to defend anarchy.
I disagree with the idea that efficient roads require imminent domain. It's not even hard to prove. All I have to do is give one example of a business that was made without imminent domain. The railroad system, which I brought up before.
I still mostly think a nation of private roads is a bad idea, since it's hard to imagine a way or scenario in which they wouldn't be a local monopoly.
Actually, in the U.S. at least, railroads did get lots of land grants, right-of-way rights, and similar subsidies from the government. So yeah.
Which is part of the reason I think it's a bad idea. The railroads constantly petitioned for those rights, that money and essentially leached off the American people. That's what rent seeking means.
Are railroads that good an example? Some railroads and subways were built using eminent domain although I don't know how much. And many of the large railroads built in the US in the second half of the 20th century went through land that did not have any private ownership but was given to the railroads by the government.
Railroads are a good example of a bad idea. The reason I picked them is that they were terrible, if I was going to pick innovative and creative real estate purchases by private industry, I'd be talking about McDonalds or Starbucks.
That's interesting. I wouldn't expect there to be many examples of working privatized roads and their effects on a nationwide scale, but if there were, I'd love to see more about them, or even a good paper based on a hypothetical.
I think you're stuck in the mindset of 'if it wasn't for our government provided roads where would we drive our cars?'. Such a world would probably have fewer private cars and be arranged in such a way that many ordinary people could get by perfectly well without a car, as is the case in many European and Japanese cities.
This article might help you understand some of the hidden assumptions many Americans operate under. Note: this guy has some rather wacky ideas but his articles on 'traditional cities' are pretty interesting.
I strongly agree with you that the US federal government has spent too much on road subsidies over the years and should decrease its current spending.
That said, not everywhere is Juneau, Alaska; not all sites connected to government roads are a "Suburban Hell," and not all inhabitants of the suburbs would prefer to live in a "Traditional City." Roads are useful for accommodating a highly mobile, atomistic society that exploits new resources and adopts new local trade routes every 20 years or so. Cars and parking lots are useful for separating people who have recently immigrated from all different places and who really don't like each other and don't want to have much to do with each other. Interstate highways were built for evacuation and civil defense as well as for actual transport. Finally, regardless of whether you prefer roads or trains, some level of government subsidy and/or coordination is probably needed to get the most efficient transportation system possible.
In any case, this thread started out as a discussion of Traditional vs. Bayesian rationality, did it not? Improving government policy was merely the example chosen to illustrate a point. It seems unsportsmanlike to shoot that point down on the grounds that virtually all government does more harm than good. Even if such a claim were true, one might still want to know how to generate government policies that do relatively less harm, given a set of political constraints that temporarily prevent enacting a strong version of (anarcho)libertarianism.
The failure of government is not a problem of not knowing which government policies would do relatively less harm. The primary problem of government is that there is little incentive to implement such policies. Trying to improve government by working to figure out better policies is like trying to avoid being eaten by a lion by making a sound logical argument for the ethics of vegetarianism. The lion has no more interest in the finer points of ethics than a politician does in the effects of policy on anything other than his own self-interest.
Your link provides very little evidence for your claim. At the national level, to say that a program costs $1 million per year is unimpressive. Suppose, for the sake of argument, that the multiplier effect for mohair production is quite low, say, 0.5. I suspect that is it rather higher than that, since multiple people will go and card and weave and spin the damn fibers and then sell them to each other at art fairs, but let's say it's 0.5. That means you're wasting $500,000 a year. In the context of a $5 trillion annual budget, you're looking at 1 part per 10 million, or an 0.00001% increase in efficiency. Why should one of our 545 elected representatives, or even one of their 20,000 staffers, make this a priority to eliminate? The amazing thing is that the subsidy was eliminated at all, not that it crept back in. All systems have some degree of parasitism, 'rent', or waste. This is not exactly low-hanging fruit we're talking about here.
More generally, I have worked for a few different politicians, and so far as I could tell, most of them mostly cared about figuring out better policies subject to maintaining a high probability of being re-elected. None of them appeared to have the slightest interest in directly profiting from their work as public servants, nor in exploiting their positions for fame, sex, etc. Those are just the cases that make the news. In my opinion, based on a moderate level of personal experience, the assumption that politicians are primarily motivated by self-interest at the margin in equilibrium is simply false.
What did you take my claim to be? The example in the link is intended to illustrate the fact that the problem of politics is not one of figuring out better policy. It is an example of a policy that is universally agreed to be bad and yet has persisted for over 60 years, despite a brief period in which it was temporarily stamped out. The magnitude of the subsidy in this case may be small but there are many thousands of such bad policies, some of much greater individual magnitude, and they add up. The example is intentionally a small and un-controversial example since it is intended to illustrate that if even minor bad policies like this are hard to kill then vastly larger ones are unlikely to be eliminated without structural reform.
Giving this appearance is fairly important to succeeding as a politician so this is not indicative of much. I find it more relevant to judge by actual actions and results produced rather than by words or carefully cultivated appearances.
As a well known politician once noted, you can fool some of the people all of the time.
Some governments cause much less damage than others, so I think there's something to study.
I mentioned elsewhere that governments of relatively small states with relatively homogeneous populations seem to do better than average. Scaling these relative successes up appears problematic.
That site you linked to has an article comparing Toledo, Ohio to Toledo, Spain. Its kind of unfair because Toledo Ohio is a relativley small city and is dying economically. I was kind of offended because I live really close to there, but he does make a point.
Toledo, Spain: Pop 80,810, Unemployment 10% (estimated from Wikipedia figures). Toledo, Ohio: Pop 316,851 (city), Unemployment 13%.
Wow. That's really very eye-opening. And as someone who has spent time in old cities outside the US and doesn't even drive, I'm a bit shocked about how much of an assumption I seem to be operating with about what a city should look like.
Japanese cities still have massive infrastructure and public transportation subsidies. It's not OMG how can we not have cars?; it's OMG how can we actually have transportation in a non governmental way that actually operates in a healthy market?
City scale transportation infrastructure doesn't require large amounts of governmental involvement. Traditional European cities evolved for much of their history with minimal government involvement. City level infrastructure would be well within the capabilities of private enterprise in a world with more private ownership of public space. Large privately constructed resorts (think Disneyland) illustrate the feasibility of the concept although they are not necessarily great adverts for its desirability.