Reject!

David Trafimow and Stephen Rice ask "What if Social Scientists had Reviewed Great Scientific Works of the Past?" (HT: @JavierAparicio). Here's the summary of reasons for rejecting a circular earth, Newton's laws of motion, and Borlaug's Green revolution, among others.

I'm reminded of the number of rejects Gordon Tullock received.

I suppose Table 1 will be of value to referees in a hurry to come up with reasons for rejecting submissions.

Riding with the King

A Tweet from a few days ago reminded me of a nice piece of rent-seeking by some of my Public Choice students a few years ago.

They gave me this after semester was over, but they know that everything's a repeated game....

The song titles are pretty cleverly done, as you can see if you zoom in....

Birthdays

Nine years ago, I was walking with Gordon Tullock from Carow Hall over to Buchanan House for a talk. I'd been invited to his 80th birthday party and consequently found out that he and I shared a birthday - the 13th of February. So I said, "Gordon, I hear we have something in common." Gordon replied in a threatening tone, "Well, we'll have to do something about that then, won't we." The rest of the story is here, along with links to other Tullock tales.

This year I've an additional reason to celebrate: I've joined him as a member of the Mont Pelerin Society.

I expect if the news reached Gordon, he'd speculate about whether it would be easier to have me expelled or shot. Or maybe he'd resign in disgust at the declining standards.

Quarantine and the State

Gordon Tullock always chided the anarchists in his public choice classes because of two matters he figured were impossible to handle absent the State. The first: interconnection of roading networks. He didn't think private road owners would make it easy for commuters to move from one owner's roads to the other, or that they'd make it easy for roads to cross each other. The second: quarantine. If a deadly epidemic were to break out, private institutions couldn't use force to implement necessary quarantine.

Turns out that the State can't handle the second one either. Writes MacDoctor of the recent H1N1 flu scare:

The Herald reports that a report by Martin Dawe of the first two weeks of the Swine Flu epidemic is intensely critical of the lack of planning and poor coordination of data resources. This was the MacDoctor’s impression of the entire debacle. I knew of a potentially deadly strain of Mexican flu three days before NZ1 landed with what was to be “ground Zero” of the Swine Flu epidemic. At that time the data looked extremely alarming. Nothing short of a full epidemic response would have been appropriate. Yet people were allowed to walk off that plane and mingle with the population for nearly a day before anyone reacted with any sort of real response. I blogged on how, nearly 3 days after NZ1 landed, a patient who was on that flight wandered into the A&M, in which I was working, with flu-like symptoms, because Healthline had sent them there. That particular lady was not seen by a public health nurse until nearly 24 hours later, when she was promptly given Tamiflu.

So the country with the best chance of anywhere in the world of implementing effective quarantine [we're a very long flight from everywhere except Australia], with public institutions generally regarded as being among the most competent in the world, couldn't handle quarantine. But, at least if they didn't ramp up, that made sense because swine flu turned out to be not all that bad, right? Well...

At the other end of the crisis – when it turned out that it wasn’t really a crisis after all – it took weeks for the public health guys to stop sending me updates and even longer before we were asked not to notify the disease. This is exactly the reverse of what should have happened. When we have data that suggests there may be a dangerous virus on our way, there should be a prompt response to try and limit the spread of the disease. This response should be swift and sharp. It is better to be completely over-the-top about such a thing than dangerously blase. As soon as the data tells us that there is no crisis, or that further intervention is pointless, there should be an equally rapid (or, at least, staged) stand-down. Otherwise massive medical resources are being wasted in a futile effort to stem either a minor disease, or a severe one that is already epidemic.

None of this should really come as much surprise.

Transitional gains trap: Canadian supply management

I'd previously guessed that Canadian dairy quota has aggregate paper value of about $25 billion; I'd suggested that the Canadian government get out of the mess by giving each quota holder a bond equal in present value to the quota and impose a tax on dairy products sufficient to pay off the bond and set to expire with the last bond payment. Moves to an efficient system? Check. Winners compensate the losers? Check. I'm growing more convinced that transitional gains traps are sticky because voters kinda like being stuck in them: they like the idea of small dairy farms supplying their milk and don't much worry about the inefficiencies. The "tax the winners, buy out the losers" solution just seems so easy that it's hard to see what would be stopping it if folks really wanted it. Indeed, when folks are ready for change, compensating the losers barely seems to matter anyway.

William Robson and Colin Busby of the CD Howe Institute in Canada have a new study (National Post story here) out on the costs of supply management and potential for reform. They argue for a regular auction of new quota for the next twenty years until the cap becomes non-binding.

They peg the aggregate value of dairy quota at $21 billion and of all quota over $28 billion. Where I'd just multiplied the number of permits by the cost of a permit, they used a Stats Can Financial Farm Database.

The Robson/Busby plan, combined with tariff-rate quota liberalization, gives a decent path to market liberalization with costs primarily borne by existing quota holders who'd see their quota value reduce to nothing over time. If this were politically feasible, I would support it over my proposed tax and bond solution. Given the number of dairy farmers whose retirement plans hinge on selling off the quota when they hit 65, it might be a brave government that tries it.

I like Robson and Busby's listing of the nonsense that comes with a supply managed system:

Innovations that would otherwise be welcomed for expanding consumer choice – new milk products or substitutes for making ice cream, yogurt and cheese, for example – threaten the cartel system. When domestic milk prices are much higher than their international counterparts, food producers have incentives to import processed goods and substances that are not classified as dairy at the border – such as ice cream with enough sugar content to qualify as a sugar product, not a dairy product. In response, federal authorities have been expanding the list of prohibited products and tightening border controls to keep them out. The adverse impact of supply-management on the competitiveness of Canadian food manufacturers prompted a “Special Class Permit System” in 1995, which allows purchasers of industrial milk for use in dairy products, or purchasers of processed dairy products, to access different prices depending on end use. Food processors, for example, are sometimes allowed to buy cheaper Canadian cheese for use in exported frozen meals. This has created a whole new arena for lobbying and conflict over limited access to cheaper imports.

A nice piece on the whole, but they forgot to cite He Who Must Be Cited on transitional gains traps. Sigh.

In other Canadian supply management news, the battles over Manitoba potatoes continue...

Game theory at the aquarium

Via Schneier, we learn a bit more about cleaner fish. I'd heard of these before: little fish that eat parasites off of bigger fish, with both parties being made better off for the exchange. But it's far more interesting than I'd thought.

The cleaner fishes get to choose cooperate (clean parasites) or defect (eat the client fish's tasty skin mucous) for each client. The client fish often leaves if the cleaner chooses defect.

When working alone, male and female cleaners defect against clients about equally. But when cleaner fish work in pairs (male/female), female defection rates drop considerably. If the client leaves because the female defects, the male punishes her by chasing her aggressively, reducing her likelihood of defecting for the next client. Presumably the (smaller) female isn't as able to punish the male for defection. On the whole, average defection rates are lower for partnered cleaners, and clients seek out cleaners that work in pairs.

Here's the rather nice Nature article, October 2008.

And of course recall that Gordon Tullock invented the field of bioeconomics back in 1971.

Utility enhancing constraints and dairy

In his grad public choice class, Gordon Tullock liked to tell a story from his early career as a State Department official in China, where he was stationed prior to being overrun by the Communists (as best I understand the story).

He said it was common practice for labourers to pull barges up canals using ropes. Another worker would watch over them with a lash, making sure everyone was pulling hard. When he inquired into the practice, he found that the team was paid based on how quickly they would get the barge up the river. None of them, while pulling, could monitor the others to ensure nobody was shirking. And they all knew they'd be tempted to shirk absent a monitor. And so they found it beneficial to assign someone with a whip to watch over the team. The team chose one of their own to stand over them all with a whip, to make sure everyone did his share.

Bernard Hickey today points to more evidence of animal neglect and cruelty in New Zealand's dairy industry. If you go and read the horror stories, be sure to go for the unicorn chaser afterwards.

The Ministry of Agriculture and Fisheries has almost no staff assigned to check into animal welfare complaints. Fonterra, the large dairy cooperative that purchases the vast majority of milk produced in New Zealand, will not get involved in animal welfare cases except where a farmer has been convicted and has not fixed the problem that led to conviction. Of course, since there are almost no animal welfare inspectors, probability of conviction is slim.

Hickey worries that these stories could tarnish New Zealand's dairy image and hurt our position as the largest exporter of the world's traded milk. It's possible. I don't know the probability, but it's possible. Fonterra is residual claimant if New Zealand's dairy industry takes a big hit on this kind of issue. Sometimes, it's worthwhile for a cooperative to designate a whipmaster, and Fonterra seems well placed to take on the job if its shareholders/members want it.

If Fonterra doesn't want to act as whipmaster itself, is there anything that would stop Fonterra from voluntarily paying a levy to the government to fund the hiring of more MAF officers whose sole role would be ensuring animal welfare on Fonterra cooperative farms, then slapping some "animal friendly" stickers on all brands of Fonterra milk? Fonterra's annual turnover is $19.5 billion, with payout to shareholder members of about $9 billion. Would a million or two to equip a team of vets be all that costly in the grand scheme of things? What risk of PR disaster would be sufficient to make the expenditure worthwhile? If there's no impediment to this kind of thing, and Fonterra chooses not to do it, it suggests that Fonterra does not weigh terribly heavily the risks of brand deterioration due to some of its farmer members' rather terrible practices and sees little potential for selling premium "Don't be Evil" product.

Let's hope that Bernard Hickey is successful in making Fonterra deem these kinds of investments worthwhile.

Previous post: animal welfare.

Update: Bernard Hickey's links are broken; try here here here.

Transitional gains traps

I'm wrong about this, but I don't know why I'm wrong. I know that I'm wrong

1. Because it's never been done and
2. Because Tullock says there is no solution

But I don't know why I'm wrong. Maybe you can help me out.

Tullock in 1976 wrote about the Transitional Gains Trap. Suppose that the government puts in place a regulation that confers rents on a few companies. So each of those companies earns an extra $1 million per year, now and forever. The value of the new rental stream has to be capitalized into the price of the fixed asset that draws the rent. And so New York City taxicab medallions, which give their owners the right to run a vehicle as a taxicab, sell for about $750,000. The link is from the homepage of a firm that provides loans to help folks buy taxicab medallions. And in Canada's ridiculous dairy quota management system, the right to milk a cow costs about $25,000. The value of the rent gets capitalized into the asset that's in fixed supply: the permit to run the cab, the right to milk a cow, the land that's eligible for tobacco growing, and so on.

After that capitalization has taken place, the person benefiting from the rental flow is again earning only a normal rate of return on his investment. All of his gain was transitional: the rent-seeker gets a one-off increase in capital value, but no ongoing benefits. Of course, over time, ownership changes; the new owners never enjoyed the transitional gain and earn only a normal rate of return.

Tullock says that, as consequence, reform is well-neigh impossible. While the folks getting the rent are not made better off by it, getting rid of it would impose massive capital losses on them; they'll then lobby up to the expected value of the capital loss to prevent it. And, he says further that there's no way out of it.

The solution seems remarkably simple in principle; since it's not been done, I must be wrong.

For New York Taxis, the City of New York stumps up to buy out all existing medallion holders at a price equal to the average selling price in the quarter prior to folks started talking about a buy-out. They finance this rather large purchase ($750K times about 13,500 licenses = $10 billion) by a bond issue. They then put in place a specific sales tax on taxi rides that leaves the post-change price lower than the prices charged under the medallion system but nevertheless is sufficient to pay off the bond because of reduced deadweight losses and increased numbers of cab rides. The tax expires when the bonds are fully paid off.

The scheme compensates the losers from the change by a tax on the beneficiaries. In the absence of companies that exist solely to facilitate medallion sales, it would be Pareto efficient; instead, it's likely only Kaldor-Hicks. We could imagine some compensation to Medallion Financial Group, though, that would still make the whole thing Pareto.

In the Canadian dairy case, it would be much more complicated because of the way that the Canadian system runs cross-subsidies from "industrial" milk to consumer fluid milk: the tax would have to be on the portions of milk sales that currently earn a premium. Otherwise, it would be similar but would cost a lot more -- best guess, around $25 billion. 978,000 cows * $25,000 per permit.

Think about those numbers. The capitalized value of the rents conferred by the Canadian dairy system and the New York City taxicab system together amount roughly to thirty percent of New Zealand GDP. Ugh.

Ok, so why am I wrong? It looks Pareto to me. What am I missing?

Hoisted from the comments: for a Hayekian game

Commenter Max Marty writes:

I've often wished for a Simcity-ish simulation that operated on actual market principles. I even tried getting a copy of a simulation called "Capitalism" but its just a different flavor of micro-management at the agent level.

Alright, who wants to start a Hayekian inspired game company with me?

I've wondered what a Hayekian Sim City game would look like. If you start ground up with a new city, like Sim City, the game would be trivial: just set an optimal rule structure at the start and make sure there's an insurance market against Godzilla attacks.

A more interesting variant could work something like the following. You're the city comptroller. You've been appointed by the newly-elected mayor to sort out the mess that the city's in. He was elected on a reform platform, but he's worried that if he goes too far, he'll be turfed in the next election and he's more than happy to make you scapegoat if necessary. You get polling data on which issues the public thinks are currently most important and reactions to your latest policy moves. All of the economics works as it should, but implementing optimal policy too quickly could see you turfed from office. So, if voters think education is important, you might still find that a move to eliminate zoning and turn the public schools into charter schools would yield protests and strikes by the teachers' unions (still protected by State-level union rules so you can't just fire the lot of them) and protests by folks who would suffer large losses to property values with the eliminations of restrictive zoning (rich folks basically). What small moves can you make to build support towards optimal policy? How many compromises on other margins are you willing to accept to make small changes on the more important margin?

Or, if folks are upset about the high costs of taxicabs, you might find that eliminating taxicab licensing (restrictions on entry: the New York Medallion system) would dramatically reduce fares but also would yield gridlock as cab owners protested by blocking key intersections and a wave of voter opposition as a poor immigrant who invested his life savings to buy a taxicab medallion, which he planned on passing on to his son, talks about how the reforms would bankrupt him. Can you devise a policy that compensates the losers but remains popular with voters? Tullock says no, but I still think a buy-out of medallion holders financed by a bond issue that's paid off using a tax on taxi fares would be a Pareto move. I'm almost surely wrong though, because there are no Pareto moves left in politics (== $20 on the sidewalk) and because Tullock says there's no system that would work. I just don't know why it wouldn't work.

Add in machinations of the challenger for office and some incentives to implement policies that induce his supporters to leave town: the Curley Effect.

Then add in random swings in voter opinion based on whether the local TV station aired an expose on the free coffee city workers get in their lunchrooms or whether Barry White has just sung a song about how the city should eliminate whacking day.

The game sounds depressing. And frustrating. And is sufficient reason never to run for public office.