Today is the ninety-ninth anniversary of the legendary fire at the Triangle Shirtwaist Factory, which reigned for ninety years as the worst workplace disaster in New York history. A hundred and forty six workers died that day, most of them young women. Escape routes were cut off by doors that were kept locked to prevent employee pilfering. The main exit from the factory floor was designed so that only one person at a time could pass through; departing workers had their handbags inspected by a night watchman. “It comes down to dollars and cents against human life, no matter how you look at it”, in the words of then-Fire Chief Edward Croker.

Well, yes, of course it comes down, at least in part, to dollars and cents against human life (where “dollars and cents” are, of course, stand-ins for “a whole lot of other things we care about”). The interesting question is whether the terms of trade were favorable. In other words: If the workers, in advance of the fire, had been fully informed of all the risks and all the potential consequences, would they have wanted those doors locked or open? Or more generally: When the New York state legislature responded to the fire with over two dozen new occupational health and safety laws, were they compounding the disaster?

I propose to take the question seriously. Would the workers have preferred to have working fire exits?

It helps to know that the labor market in the garment industry was highly competitive on both the supply and demand sides. There were hundreds of garment factories in lower New York, some of them in the same building as the Triangle factory. (The fire was confined to Triangle’s three floors.) They drew their workers from the teeming tenements of the Lower East Side, both as direct employees and through independent contractors. In a competitive labor market, workers are paid their marginal product. (For this there is ample theory and evidence, which no economist disputes.) This means that if one additional seamstress can add six dollars a week to the company’s revenue, then all seamstresses of that skill level are paid six dollars a week. (Six dollars a week is a historically accurate wage rate.)

Now let’s suppose that in the presence of open doors, the typical employee pilfers two blouses a week, with a wholesale value of 60 cents apiece. (Sixty cents is a guestimate based on the retail price of a little over a dollar, which I found in an old Sears catalogue. Two pilfered blouses a week is a number that I just made up. If you don’t like that number, feel free to adjust my calculations.) The employee’s marginal product falls by $1.20, so competitive pressures force the wage to fall by $1.20 as well. That’s a 20% wage cut (though it’s partly offset by the pleasure of having a very full clothes closet and an infinite supply of rags).

Now our question becomes: Would a worker, given the choice and fully recognizing the risk of fire, have taken a 20% wage cut in order to keep the exit doors unlocked?

Well—would you take a 20% wage cut in order to keep the exit doors at your workplace unlocked? (I know, I know, they’re already unlocked—-but pretend they’re not). I wouldn’t. Of course this proves little, because you and I live in a very different world than a 1911 garment worker. But the differences cut both ways. On the one hand, the garment worker occupied a wooden building filled with fabric and tissue paper. That makes the exit door more valuable. On the other hand, the garment worker was a lot closer to starvation than we are. That makes the wage cut harder to swallow.

My guess is that the second effect is bigger, but I could be wrong. It might not be too hard to resolve the matter with a little effort, a little arithmetic, and a little bit of data on how the demand for safety varies with income. In any event, I’m extremely skeptical that our desperately poor garment worker would have chosen the wage cut.

What other evidence speaks to the preferences of the workers?

In one direction, there is no pre-fire record of workers offering to take wage cuts in exchange for better fire safety, or of firms anticipating that they could cut their wage bills by putting in better fire doors. There was plenty of labor unrest, but so far as I am aware, it was almost all about wages and hours, not safety. The most straightforward reading is that workers preferred higher wages to more safety, but an alternative reading is that workers were blissfully unaware of the extent of the fire risk.

But in the other direction, there is an extensive post-fire record of workers applauding the new safety regulations, despite the fact that the regulations must have depressed wages. The most straightforward reading is that workers preferred more safety to higher wages, but an alternative reading is that workers were blissfully unaware of how the new laws would affect their wages. Another alternative reading is that pilfering was never the major problem I’ve been envisioning, so the effect on wages was minimal. (It’s hard to measure the effect directly because wages change for many other reasons as well.)

Bottom line: I can’t be sure (and I’ve pointed out several reasons I might be wrong), but I’m guessing that no 1911 garment worker would have wanted to work in a factory with unlocked exit doors. If I’m right, they got the mix of risk and income they’d have chosen. The fire was tragic but the market worked.

Share/Save

The column oozes nonsense throughout, but it will be instructive to hone in on one exceptionally silly paragraph. Here is Hansen trying to explain why cap-and-trade is inferior to a carbon tax:

Consider the perverse effect cap and trade has on altruistic actions. Say you decide to buy a small, high-efficiency car. That reduces your emissions, but not your country’s. Instead, it allows somebody else to buy a bigger S.U.V.—because the total emissions are set by the cap.

First, this is true. But second, it has nothing to do with cap-and-trade. The same observation applies to a carbon tax. Carbon taxes work by driving up the price of gasoline and other pollutants so people will use less of them. Consider the perverse effect that has on altruistic actions. Say you decide to buy a small, high-efficiency car. You buy less gasoline, driving down its price, and encouraging someone else somewhere to buy a bigger S.U.V.

True, you have only a tiny tiny effect on the price of gas. But how big an effect do you need to trigger one more S.U.V. sale in a country of 300 million people?

Here, now, is the whole point: We don’t have to guess at the answer to that. A little bit of economics reveals that to an excellent first approximation, the carbon tax and the cap-and-trade program must have identical effects.

Here’s why: If we reduce our gasoline consumption by, say, 30%, then the price of gasoline must rise by just enough so that consumers are willing to cut their purchases by exactly 30%. If the price rises any less than that, consumers will want more gas than they can get, and they’ll bid the price up further. So the price consumers pay for gasoline is completely determined by the fall in consumption—and it doesn’t matter what triggers that fall. Cut gas consumption 30% via cap-and-trade and the price of gas will rise. Cut gas consumption 30% via a gas tax and the post-tax price of gas will rise by exactly the same amount. This one observation forces the programs to be equivalent in almost every relevant way.

Mr. Hansen prefers the carbon tax, because the revenue can be distributed to the public. (He calls this a fee-and-dividend system.) But by the same token, with cap-and-trade, the revenue from selling permits can be distributed to the public. And it’s not hard to prove (using arguments similar to the preceding paragraph) that the revenue from selling permits must exactly equal the revenue from raising taxes. So again the programs are entirely equivalent.

Mr. Hansen worries that with cap-and-trade, energy producers will demand free permits—essentially claiming the permit revenue for themselves. But energy producers that powerful could just as easily lay claim to tax revenue as to permit revenue. So again, the choice of policy makes no difference. He worries that politically powerful producers can get themselves grandfathered out of cap-and-trade. Yes, just as easily as they can get themselves grandfathered out of a carbon tax.

If you want to believe there’s a meaningful difference between a carbon tax and cap-and-trade, you’ve got to look elsewhere than the effects on pollution, incentives, consumer prices, profits, or government revenue. In other words, you’ve got to look at decidely secondary effects. Here’s a menu of choices:

• A carbon tax can be too small (allowing too much pollution) or too big (retarding too much economic activity). Similarly, a cap-and-trade program can be too small or too big. For somewhat technical reasons, I believe that it’s easier to calculate the “just right” carbon tax than to calculate the “just right” number of cap-and-trade permits. If I’m right, this is an argument for a carbon tax.
• A carbon tax is pay-as-you-go for producers. Cap-and-trade permits, depending on how the system is designed, can be a big upfront expense. This handicaps small companies with limited cash reserves and limited borrowing opportunities. This is an argument for a carbon tax.
• If you favor either of these programs in the first place, you presumably want it to last awhile. Cap-and-trade is politically self-perpetuating, because companies in possession of valuable permits will lobby to maintain the program that gives those permits their value. The carbon tax, by contrast, will come under constant assault from anti-tax lobbyists and therefore might be more difficult to maintain. This is an argument for cap-and-trade.

Me, I lean toward a carbon tax, just like James Hansen. But the differences are largely inconsequential, and almost entirely divorced from what Mr. Hansen thinks they are.

Share/Save