Gladwell Chokes
It's hard to beat the thrill of thinking of something interesting to write, and it's crushing to find that it's already been written. This happened to me after I read Malcolm Gladwell's piece in the New Yorker, "The Risk Pool." I found an error in Gladwell's reasoning, only to find a pithier version of my argument posted on his blog. Luckily, he is defending his error, so I have an excuse to pile on.
Briefly, "The Risk Pool" explains the concept of a "dependency ratio" and applies it to two phenomena. A dependency ratio is just the ratio of non-workers to workers. So for instance, in a family with two working parents and three non-working children, the ratio is 3:2.
Gladwell first explains how dependency ratios affect the wealth of nations. A country divides up its resources among its citizens. The average wealth should roughly correspond to the standard of living. The average wealth is just total wealth divided by the total number of consumers. People who work add to both the numerator and denominator. People who don't work merely add to the denominator, so they detract from average wealth.
It's easy to see that, other things being equal, a low dependency ratio will allow a higher standard of living. Whether this makes for a faster-growing economy is unclear to me, but I'll leave that to the experts. Gladwell goes on to apply the concept to the pension plans of American corporations. Here he goes badly astray.
Many declining American industries, Gladwell points out, promised generous pensions and health care to large workforces in the middle of the twentieth century. Now they're having trouble meeting those obligations. Gladwell attributes this to a worsening dependency ratio. For instance, "General Motors today makes more cars and trucks than it did in the early nineteen-sixties, but it does so with about a third of the employees." This is a problem, asserts Gladwell, because it leads to an unfavorable dependency ratio. Each current GM worker has to support several retirees. It was a bad dependency ratio, he argues, that doomed Bethlehem Steel (more later).
The problem with this analysis is obvious. Retirees don't depend on workers to pay their benefits. They depend on GM. GM can pay those benefits so long as its profits are high enough. Gladwell posts just such a critique, sent in by a reader, including this passage:
The number of current workers they have is irrelevant to their ability to pay or not pay for former employees' benefits. If Rick Wagoner could produce umpteen-billion cars single-handedly after having invested in a 100%-automated assembly line he could be wildly profitable and be able to afford all those old retiree costs and have the worst dependency ratio possible.
Gladwell's response (in italics):
This is, I think, a very good example of the "if pigs had wings. . . " line of argumentation. Sure, if Rick Wagoner could produce all of GM's cars by himself, and as a result be wildly profitable, it wouldn't matter how many retirees he had. So what? Rick Wagoner can't do that. He is--like the heads of Bethlehem Steel before him--trapped in a low margin, labor-intensive business, where workforce compensation is a enormous share of the bottom line. McGurky is absolutely right that "GM can't pay their retiree obligations because they have no profits." But why doesn't GM have any profits? Because of the size of their retiree obligations!
Gladwell is very confused. His reader's point was not that robots are about to take over the car industry, but rather that the number of current workers simply doesn't matter (except indirectly). Gladwell defeats the hypothetical by taking it literally!
More importantly, he doesn't explain why dependency ratios have any traction here. After all, if GM could trade its retiree obligations for the obligation to pay a dollar to every American, it would have a much worse dependency ratio, but a much better outlook. Surely the relevant ratio is something similar to the debt-to-earnings ratio, and the nature of the obligations (debt, taxes, pensions) is irrelevant.
Gladwell invokes GM's retiree obligations, but their importance isn't in dispute. The question is whether it matters that GM's workforce has declined significantly over the years. The answer is no: retirees don't care whether their money comes from a company that employs lots of low-productivity workers or a few high-productivity workers. The retirees just want GM to stay solvent and honor its obligations. Any purported tradeoff between profitability and dependency ratios is irrelevant: how are retirees better off if GM hires more workers but becomes less profitable?
Admittedly, laying workers off is less likely to be profitable because those workers will continue to get benefits. That has nothing to do with the existing body of retired workers, though. It's just a feature of GM's current labor agreements.
Gladwell concludes with a frustrating example. A guy named Wilbur Ross bought the restructured Bethlehem Steel. Gladwell writes, "The new Bethlehem Steel had a dependency ratio of 0 to 1. Within about six months, it was profitable. The main problem with the American steel business wasn’t the steel business, Ross showed. It was all the things that had nothing to do with the steel business."
Well, sure, the dependency ratio improved, but only because of the elimination of a huge set of obligations. Give me any company in financial but not economic distress, and I can work wonders by reducing its obligations. This is the magic of Chapter 11. Bethlehem's experience would be identical if its obligations had been to banks, but the dependency ratio concept clearly wouldn't apply there.
At this point I suppose I'm flogging a dead horse, so I'll stop and issue an invitation: Malcolm Gladwell, if you want to defend yourself on the pages of Pur Autre Vie, I will publish what you have to say. Alternatively, I'm happy to write something up for the New Yorker... anything you want... call me.
Briefly, "The Risk Pool" explains the concept of a "dependency ratio" and applies it to two phenomena. A dependency ratio is just the ratio of non-workers to workers. So for instance, in a family with two working parents and three non-working children, the ratio is 3:2.
Gladwell first explains how dependency ratios affect the wealth of nations. A country divides up its resources among its citizens. The average wealth should roughly correspond to the standard of living. The average wealth is just total wealth divided by the total number of consumers. People who work add to both the numerator and denominator. People who don't work merely add to the denominator, so they detract from average wealth.
It's easy to see that, other things being equal, a low dependency ratio will allow a higher standard of living. Whether this makes for a faster-growing economy is unclear to me, but I'll leave that to the experts. Gladwell goes on to apply the concept to the pension plans of American corporations. Here he goes badly astray.
Many declining American industries, Gladwell points out, promised generous pensions and health care to large workforces in the middle of the twentieth century. Now they're having trouble meeting those obligations. Gladwell attributes this to a worsening dependency ratio. For instance, "General Motors today makes more cars and trucks than it did in the early nineteen-sixties, but it does so with about a third of the employees." This is a problem, asserts Gladwell, because it leads to an unfavorable dependency ratio. Each current GM worker has to support several retirees. It was a bad dependency ratio, he argues, that doomed Bethlehem Steel (more later).
The problem with this analysis is obvious. Retirees don't depend on workers to pay their benefits. They depend on GM. GM can pay those benefits so long as its profits are high enough. Gladwell posts just such a critique, sent in by a reader, including this passage:
The number of current workers they have is irrelevant to their ability to pay or not pay for former employees' benefits. If Rick Wagoner could produce umpteen-billion cars single-handedly after having invested in a 100%-automated assembly line he could be wildly profitable and be able to afford all those old retiree costs and have the worst dependency ratio possible.
Gladwell's response (in italics):
This is, I think, a very good example of the "if pigs had wings. . . " line of argumentation. Sure, if Rick Wagoner could produce all of GM's cars by himself, and as a result be wildly profitable, it wouldn't matter how many retirees he had. So what? Rick Wagoner can't do that. He is--like the heads of Bethlehem Steel before him--trapped in a low margin, labor-intensive business, where workforce compensation is a enormous share of the bottom line. McGurky is absolutely right that "GM can't pay their retiree obligations because they have no profits." But why doesn't GM have any profits? Because of the size of their retiree obligations!
Gladwell is very confused. His reader's point was not that robots are about to take over the car industry, but rather that the number of current workers simply doesn't matter (except indirectly). Gladwell defeats the hypothetical by taking it literally!
More importantly, he doesn't explain why dependency ratios have any traction here. After all, if GM could trade its retiree obligations for the obligation to pay a dollar to every American, it would have a much worse dependency ratio, but a much better outlook. Surely the relevant ratio is something similar to the debt-to-earnings ratio, and the nature of the obligations (debt, taxes, pensions) is irrelevant.
Gladwell invokes GM's retiree obligations, but their importance isn't in dispute. The question is whether it matters that GM's workforce has declined significantly over the years. The answer is no: retirees don't care whether their money comes from a company that employs lots of low-productivity workers or a few high-productivity workers. The retirees just want GM to stay solvent and honor its obligations. Any purported tradeoff between profitability and dependency ratios is irrelevant: how are retirees better off if GM hires more workers but becomes less profitable?
Admittedly, laying workers off is less likely to be profitable because those workers will continue to get benefits. That has nothing to do with the existing body of retired workers, though. It's just a feature of GM's current labor agreements.
Gladwell concludes with a frustrating example. A guy named Wilbur Ross bought the restructured Bethlehem Steel. Gladwell writes, "The new Bethlehem Steel had a dependency ratio of 0 to 1. Within about six months, it was profitable. The main problem with the American steel business wasn’t the steel business, Ross showed. It was all the things that had nothing to do with the steel business."
Well, sure, the dependency ratio improved, but only because of the elimination of a huge set of obligations. Give me any company in financial but not economic distress, and I can work wonders by reducing its obligations. This is the magic of Chapter 11. Bethlehem's experience would be identical if its obligations had been to banks, but the dependency ratio concept clearly wouldn't apply there.
At this point I suppose I'm flogging a dead horse, so I'll stop and issue an invitation: Malcolm Gladwell, if you want to defend yourself on the pages of Pur Autre Vie, I will publish what you have to say. Alternatively, I'm happy to write something up for the New Yorker... anything you want... call me.
posted by James | 2:45 PM
2 Comments:
Sounds right
Will
Shit, I guess pigs do have wings. You've earned yourself a guest column on "New York City: 300 Years of Glorious Tradition."
Love,
MG
P.S. "MG" is "GM" backwards, kind of like my analysis.
Post a Comment
<< Home