Byrne's Eye View

In any popularly elected legislature, there are only two margins of victory that make sense: winning by one vote, and winning all of them. Since the senate has 100 members, it's easiest to convert to and from percentages, so let's consider them: imagine a bill that looks like it's going to pass with 60 votes. It obviously only needs 51. So if there's any way to modify the bill such that it's less appealing to nine senators but more appealing to 51 (or unappealing enough to some senators, such that it loses nine votes, but more appealing to the remaining 51 voters), everyone involved has every incentive to back those changes: the 51 voters-in-favor get to vote for a bill they like more, while the nine new voters-against can now vote against something they don't like. Assuming an evenly-divided legislature, you can classify those bills as "I am a Republican" or "I am a Democrat" bills -- they do everything possible to benefit the narrow majority that supports them, while staying unappealing to the opposition half of neutral voters.

But there are other bills, like the PATRIOT act and various censures and resolutions, which tend to pass with 99 or 100 votes. Those obviously aren't "I'm a Democrat/Republican" bills, but someone's support of them doesn't have any informational content, so I'd consider them "I am an American." bills. Or perhaps "I am not in favor of illegalizing ice cream and puppies." bills. The interesting thing about these bills is that they show exactly the opposite tendency as partisan-identity legislation: if a new resolution has the support of 90 members, every one of those 90 has every reason to lard it with a little more meaningless rhetoric, which puts the remaining 10 in an increasingly dicey situation.

If you imagine bills being proposed at random, it's obvious that they gravitate towards the nearest pole: if they have more than 75 votes in favor, they should reach 100; if they have between 25 and 75, they should find their way to 51 or 49 -- and otherwise, they should end up at 0.[1] In my experience, though, these voting margins are common, but not universal. Here are a few possible explanations:

1. Inefficient negotiations markets: it's just not worth everybody's time to squeeze in a little more pork, so politicians will leave a bill alone as it gets closer to 50/50 support.


2. Too atomic: maybe this happens one session at a time, or one career at a time. "Logrolling" could account for some of this, too: one bill might get a few extra votes in exchange for another bill getting just the right number to pass (obviously, in an efficient market this would have no net effect, since a vote-due-to-logrolling is as much a part of a 51-vote bloc as any other. But it's harder to account for when adjusting the terms of a bill, so it feeds into the first explanation).


3. Politicians are underrated: as painful as it is for a libertarian to say this, maybe politicians are actually voting on principle, and this cynical vote-trading stuff just doesn't appeal to them.

If explanations #1 and #2 are correct, we'd get less pork if we had smaller legislatures, because the marginal utility of changing a bill would be less constant.[2] If explanation #3 is correct, we are lucky, indeed -- as long as they're smart.

[1] Obviously, that doesn't happen. But bills that start out with few prospects of passing, and continue to lose support, just get dropped -- so it's survivorship bias.

[2] Unless one of the constraints is lack of information about other legislators' preferences.

This Washington Post article on illegal timber smuggling in China is notable for what it doesn't say: that since nobody owns the trees that are being cut down, everyone has an incentive to cut them down as quickly as possible.

For a more reasonable perspective, consider Doctor Suess.

A common criticism of anarcho-capitalism is that the private sector can't undertake massive projects -- exploring space, curing disease, doing pure theoretical research, or winning wars -- because the benefits are too dispersed. Imagine an engineering project like a new dam that makes one million people better off by $100 each, but will cost only $20 million. That's a 400% return on investment, but if the dam owners can only convince 10% of their damn customers to pay for it, they're still out. This creates a whole class of investments that are worth making from a total utility standpoint, but can't, in theory, be made without coercing some people into paying their fair share.

That's a good theory, and it would have been valid a generation ago, but I suspect that liquid capital markets have made it obsolete. After all, there are measures of welfare -- GDP per capita, the S&P, life expectancy -- and since people are, on average, better off when these go up, you'd expect the average investor to be biased in favor of shorting them. It'll take a while for people to accept the idea, but any reasonable member of the labor force ought to consistently set aside some money to bet that real wages will drop precipitously, because merely by working, a laborer is betting that those wages will be stable or increasing (and how many people find it worthwhile to come up with the macroeconomic analysis to prove that?).

If we can imagine a world in which people rationally hedge their accidental bets on things they can't control, you can imagine a world in which it's possible to profit from increasing total welfare by betting on total welfare and funding a project with such bets.[1] Take the dam example: rather than charging people for it directly, dam owners could buy an option on some measure of future utility, such that if it increases unexpectedly, they'll get half the increase up to $50 million. Then, they announce their plans, and the indicator estimates that utility will increase by (probability of success) * (increase if success = $100 million) -- so as long as the probability is seen as higher than 40%, the project is funded entirely by speculating on its own outcome (a healthy sort of insider trading that I don't think anyone has thought of banning yet).

There are certainly obstacles (for example, sabotage could be a lot more lucrative, especially if the cost of decreasing the odds of success by X% is less than X% * utility of success. But this idea should partially address the question of how to fund investments with dispersed benefits: if the benefits are measurable, bet on the measurement -- and if they're not, who can argue that there's any benefit at all?

[1] How hard is it to think that people will behave this way about risks? I'm not sure. Some people seem to have Stockholm Syndrome when it comes to risk: they'll bet the a company won't go bankrupt by getting all their income from it, and then they'll double down on that bet by buying stock in the company. On the other hand, these risks are just not fun to think about, so just as people urbanized so they wouldn't have to worry about a bad harvest, I think they'll hedge just because it's boring to worry about unfixable risks.

I thought the deal was off, but apparently Bear Stearns and Citic are almost ready to trade together. This plan involves swapping ownership stakes -- BSC will own part of Citic, and Citic will own part of BSC. Which brings up a question: how do you analyze a situation like that?

To consider a simple version, let's imagine A Corp. and B Corp. Each has 100 shares, and each has $1 of cash. First, let's imagine A owns half of B, and B owns half of A. In this case, you can end up with a foolish recursive increase in value: A's market value is $1, and B's is $1, but since A owns half of B, A is worth $1.50 total. And since B owns half of that, it's worth $1.75 -- and since A owns half of that, A must be worth $1.75/2 + $1.00, or $1.825. A naive investor would keep ratcheting up values until both companies were worth around $2 total: $1 of cash and $1 worth of the other's stock. Of course, that gives a total valuation of $4 on two businesses whose only assets are $2 in cash and a complex ownership structure.

This was apparently one of the problems with Japanese equity markets in the 80's. If A, B, C, and D were connected in a chain of cross-ownership, A's good quarter (in operations) could show up on B's P&L thanks to trading, which would boost C's results, which, in turn, would give D's balance sheet enough heft for it to issue more debt, which it would lend to A, which would...

At some point, it's necessary to just analyze each company operationally first, and then factor in cross-ownership. That's the only way to avoid giving $2 of cash a $4 market value, or making one good quarter for one company siphon cash into four mediocre operations. If BSC and Citic are doing something legitimate, one would expect a simpler structure. As it is, they're knotting up their finances with extra complexity, which, as far as I can tell, only makes it easier for investors to accidentally overprice them. What's their motivation?

Every few days, someone will invite me to join this Facebook group devoted to saving Facebook from being shut down. It's a common feature of populist institutions (like community websites, or political movements) that the easiest way to get popular within the group is to complain about how unpopular one is outside the group. What's annoying about this group is that it's a classic viral phenomenon, because it's a) not going to accomplish anything, and b) not worth having even if it can accomplish something.

a) The group is ostensibly a petition to 'save facebook' from being shut down in the event that Facebook loses the lawsuit filed by ConnectU, a competitor that hired Facebook's CEO to do some coding before Facebook was founded. Apparently the groups hundreds of thousands of members aren't aware that our legal tradition explicitly avoids populist judiciaries -- if law were something decided by the will of the crowd, legislatures would make courts redundant. Courts decide disputes between parties, in this case over a contract, which means the court is deciding whether one party upheld a bargain it made with the other. This is a question of fact, not opinion, so the many, many angry teenagers with opinions on the subject don't really matter.

b) Unless the ConnectU founders are truly vindictive, they aren't going to change a thing. In fact, it's obvious that if they're right about their lawsuit (that they gave Zuckerberg the idea, and he turned ConnectU's idea into Facebook's business), it's obvious that Zuckerberg is a far better programmer and manager than the ConnectU guys. So if they have any sense at all, they'll hand him a substantial chunk of equity, and make him CEO again. It might delay his entry onto the Forbes 400 by a year or two, but he'd come out okay, and they'd be better off. And even if they don't hire him, it takes a pretty vindictive personality to ditch $10 billion or so worth of startup equity just to settle a sophomoric squabble.

But what's most interesting about this phenomenon is the many, many people who seem fanatically devoted to Facebook qua Facebook. I've never thought Facebook was an extraordinary piece of technology, but I've devoted lots of time to extolling reddit over digg, or championing Apple over Windows (an easy fight to win) and Linux (a lot harder, but doable). This despite the fact that reddit is digg plus developmental delays (unlike humans, websites get stupid as they mature) and Apple is FreeBSD plus a nice interface and a lot of hype. Loyal users are one thing -- anyone in the top 1% can get those. But stupidly loyal users, who will argue for hours about user interface minutea, or claim that choice of operating system makes someone ontologically superior, are a rare and valuable asset.

It's just a little embarrassing to be one.

Some other research has shown that, in fact, if you do open up the solution process you can get anywhere from 10X to 100X improvement in problem-solving performance.

-- Karim Lakhani, Harvard Professor

Last-resorts will always look this way. Another way of stating this is that "An unusual solution is more likely to work than a standard solution that has already failed." Richard Feynman had a nice anecdote about this (which I can't seem to find on Google): his incredible toolkit of analytical techniques included a slightly obscure calculus technique (something about inverting and then integrating -- don't ask me for details). When his colleagues had tried all the standard techniques, they'd come to him -- and, pretty frequently, his trick worked. This gave him a great reputation, but only because they didn't know that he would have tried everything they thought of, first.

To really judge an unusual method (like open-sourcing a problem, or Feynman's math trick), you'd have to ask what happens when people try it first. And the obvious result would be: that problem-solving methods tend to be used in about the order of their utility, except in a few non-sound-byteable cases.

I panicked this morning when I saw an earlier version of this article with the headline "Mexico City Evicts 15,000 Traders." Fortunately, these are the traders selling snacks and trinkets, not stocks and bonds. But this part of the article confused me:

The president of one street vendors' union, Alejandra Barrios, criticised the government, saying: "They are not thinking about the fact that these people don't have jobs. What do they think these people will do?"

First of all, worst union representative ever: "You fail to consider the fact that everyone I represent is, basically, a panhandler who produces litter." Second: a union? For street peddlers? Aren't these guys all self-employed? What kind of negotiating power do they have? And for that matter, why would they negotiate wages when they could work on commission? This is probably the purest sales job in the world -- offer something that's almost certainly worthless to a bunch of harried American tourists*, and see how much you can wheedle them into paying for it. What does a union offer these guys?

Terracycle is allegedly the first startup with a negative ecological footprint. I seriously doubt it. I'm sure Craigslist has a lower footprint -- think of all the trees that haven't been chopped down, because it's so cheap to put an ad online instead.

But even Craigslist loses to Ford. Pre-Ford, most roads were knee-deep in horse effluvia. Now the worst we worry about is sever asthma and maybe apocalyptic global warming (which might have happened from horse methane, anyway).

Terracycle is impressive, but they're only the first company to achieve what they have if you use some really bad accounting.

There's now an index tracking African stocks. The most interesting part of this article is that Africa is increasing exports to China.

I can't find any member list for the index, though this offers an exciting -- and almost certainly fictional -- datum:

The continent has consistently had equity markets that have delivered significant dollar returns. These are typically in excess of 100% per annum, as highlighted in our recent report, Africa 1.01 - Unlocking Investment Potential. This stellar performance has, to a large extent, passed unnoticed. But recent, high-profile IPO's, especially in Nigeria have led to a significant increase in investor appetite.

Hillary Clinton's new retirement plan doesn't sound like a national sales tax, but that's what it amounts to. Under the plan, savers can defer taxes on money they earn, but don't spend, until they actually get around to spending it. But another way to rephrase that is that under her plan, spending is taxed and saving is not -- and another way to rephrase that is that there's a sales tax with a pretty low-cost collection mechanism. Who knew Hillary was a Fairtaxer?

The New York Times is once again wringing its hands over lottery-funded education, noting that

In reality, most of the money raised by lotteries is used simply to sustain the games themselves, including marketing, prizes and vendor commissions. And as lotteries compete for a small number of core players and try to persuade occasional customers to play more, nearly every state has increased, or is considering increasing, the size of its prizes — further shrinking the percentage of each dollar going to education and other programs.

But what's really telling is this: “Our people are playing the lottery... We just need to decide which schools we should fund, other states’ or ours.” Even if you can create an artificial monopoly in one place, you can't sustain it everywhere. I suspect that within twenty years, lotteries will be just as profitable as any other gambling business -- that the usual competition from out-of-state arbitrage and higher-payout illegal versions of the same product will push returns on capital down to about where they belong.

And this isn't hard to implement: all you need to do is aggregate the odds for every lottery in the US, and then create a framework for sending money to third parties to buy whichever lottery has the highest return ("highest return", read "lowest loss").

Q: Did you know there was a Boston Stock Exchange?

A: There is. But not for long.

I don't get it. When they were private, stock exchanges competed by buying better technology and offering more services. Now that they're public, they just buy one another out. Is there any other business in which every single member is in play?

Another private equity company forming a vehicle to buy distressed debt. I like how much of the bubble-bust lifecycle they're part of now. It's sustainable, Green investing, in a very meta sense.

Next bubble, they should buy MSNBC on the way up, to keep on hyping -- and then start a joint venture with Sarbanes, Oxley, and Grant's Interest Rate Observer to frown those overpriced deals back down.