Wednesday, September 24, 2008

Is Bill Fitzsimmons confused?

Today in the Harvard Crimson, an intriguing story about the future of the SAT in elite college admissions (emphasis mine):
The SAT Reasoning Test may one day be optional for Harvard applicants, according to Dean of Admissions and Financial Aid William R. Fitzsimmons ’67.

The best predictor of college success is not the SAT, but rather tests that examine knowledge of a standardized curriculum, such as SAT subject tests, said Fitzsimmons, who over the past year led a commission of leading admissions officials that is recommending that colleges rely less on the SAT.

Fitzsimmons said that in the future Harvard may give students the option of taking five or more SAT Subject Tests in lieu of the SAT Reasoning Test or its frequent alternative, the ACT.
Over the past couple decades, most validity studies produced the same result: the SAT II subject tests were more predictive than the ordinary SAT. This led to all manner of philosophizing about the performance of standardized tests, as many commentators concluded that "curriculum-based" subject tests were empirically superior.

There was only one problem: the superior performance of the SAT "subject tests" was misleading, as it reflected the high predictivity of the SAT II Writing test, which (astonishingly enough) provided a more useful subscore than the SAT itself. One of the most extensive empirical studies, which observed cumulative GPAs over four years in the UC system, recently reestablished this conclusion. Now that the SAT II Writing is a part of the ordinary SAT, it's likely that the superior performance of SAT IIs has disappeared.

In any case, extolling the performance of SAT IIs as vindication of a curriculum-based approach to standardized testing was always wrong: the SAT II writing was a generic test on grammar and writing skill unconnected to any particular "curriculum," and it alone provided the SAT II's supposed edge in predictivity. Unless Bill Fitzsimmons has access to very different internal data at Harvard, it appears that he may simply be confused...

Borrowed prosperity

Megan McArdle is making sense:
A substantial minority of libertarians, and fellow-travelers on the right, view this crisis as just desserts. America has borrowed prosperity from the future, and now the bill has come due...

But I don't see how it is possible, over the long term, to have a bubble that consists of "borrowing from our future". We can't get money from the future--their financial system isn't integrated with ours. We can commit some people in the future to give others money, and in that sense our current political happiness with various old-age entitlements might be said to be borrowed from the future. But we cannot, ourselves, spend the actual cash that rightfully belongs to them.

Well, actually we can. We can borrow from foreigners. But America borrows in her own currency, and at pretty low interest rates. Our current account deficit is certainly worrisome, but not "the economy needs to contract by 1/3 to fix it" worrisome. Moreover, we still do most of our borrowing from good old US citizens.

Credit markets don't expand our wealth by borrowing from the future's precious prosperity reserves. They increase production by vastly increasing the efficiency of individual savings, funneling them into long term investments that require larger sums than most individuals can amass.
Yves Smith... not so much:
The sorry fact is the US has consumed at an unsustainable level. We need to reduce consumption and increase savings (and reducing debt is a form of savings). Reduced consumption means a fall in GDP. In Britain, which is going through its own credit crunch, the officialdom has said that the public will experience a fall in living standards. Why are we unwilling to accept the inevitable?
Smith doesn't present any theory of economic growth to accompany her confident assertion that we must experience a "fall in GDP," perhaps because there is no compelling theory to offer. The assumption that "reduced consumption means a fall in GDP" is embarrassingly undermotivated, especially when this reduced consumption comes as a result of increased savings. In fact, one key conclusion of almost every basic growth model is that a higher savings rate increases steady-state GDP.

As McArdle points out, it's impossible to have a bubble that "borrows from the future." The capacity of our economy is determined by capital, labor, and technology, none of which can be transported from the future to the present. (Yes, depletion of natural resources is an exception, but that's not the argument that Smith and most of her fellow-believers are making.) Once we acknowledge this simple fact, we're left with borrowing from foreigners. This is sizable -- our trade deficit is almost 6% of GDP -- but if our economy grows at 2 or 3 percent a year, we're adding the equivalent of our entire current account deficit in productive capacity to the economy every few years. This is hardly consistent with Smith's notion that if foreigners stop giving America a free ride, we're in for a sustained period of lower living standards.

In any case, it's certainly not clear that our current account deficit is inflating GDP, and confident statements on the matter reveal an obliviousness to what GDP actually is. Our consumption is higher because we can run deficits, although as I noted above the benefit isn't as overwhelming as one might think. But GDP doesn't measure consumption -- it measures what we produce. If the Voodoo God of Economics decided tomorrow to force the United States to have a neutral current account balance, there would be two opposing effects on GDP, whose comparative magnitudes are not at all clear:
  1. Americans, now that they could no longer achieve their former levels of consumption by running a trade deficit, would increase their labor effort to produce more at home, thus increasing GDP.
  2. With lower capital inflows (the natural concomittants of a trade deficit) coming in from abroad, and Americans eager to use their current production to maintain consumption, total capital investment in the US would fall, causing a long-term decline in the GDP growth path.
What's the lesson here? A little intuition goes a long way. Yves Smith has a Harvard MBA and 25 years of experience in investment banking, but despite her apparent erudition in financial instruments, she doesn't have the basic mental model of concepts like GDP that is necessary to make informed statements about the future trajectory of the economy. Her seemingly mature declaration that we have to "accept the inevitable," which involves a "decline in GDP," is actually an allegorical vision of borrowed prosperity that doesn't match the facts.


This just came into my inbox (emphasis mine):
We currently maintain a usenet news server in the department. After monitoring the logs for several months it appears this service is basically unused. Even though usenet was created here in the department (see for more information) we would like to discontinue its use. Before we do so, we would like to double check that no one relies on this service, or expects to use the service in the future. If you are using or plan to use the service, please contact us. If we receive no replies, the last day of operation will be October 1st.


Lab Staff

Thursday, September 18, 2008


In the world of liberal economic policy, there are two basic approaches. Some policymakers proceed in a staid, conventional way: they want to subsidize health care, make the tax system more progressive, and generally offer a series of piecemeal interventions that benefit lower-income Americans. Others offer a more comprehensive vision, involving fundamental changes in the economy that will bring shared prosperity for all. They attack inequality and stagnant growth at their roots, "transforming America's economy through clean energy, innovation, and opportunity." These people manage to seem both more exciting and more mature than the former group, stepping beyond the tired mainstays of liberal politics to bring us an integrated approach to the challenges of the 21st century.

Only problem is that they're a bunch of hacks.

Now, that's a little harsh, especially when the Center for American Progress (which I pick as a representative example of a much broader phenomenon in political advocacy) and I are in agreement on many basic issues of domestic policy. But there's no avoiding the fact that supposedly "comprehensive" approaches to the economy tend to be exercises in windy rhetoric, far inferior to the boring incrementalist mix of progressive taxation, public education, and health care funding that keeps us less aspirational types busy.

Let's begin with CAP's opening essay: "Capturing the Energy Opportunity: Creating a Low-Carbon Economy." This happens to be a goal that I completely support. But beyond making the basic point that climate change is damaging and that we should cut emissions to avert it, CAP's document lacks any semblance of intelligent analysis. Consider this randomly selected paragraph:
Our traditional understanding of energy security has been largely limited to assuring adequate supplies of energy to fuel our economy. That will remain a necessary concern, of course, but not a sufficient one. Going forward our leaders will have to act on an understanding of energy security that turns not just on the supply but on the carbon content of the energy we use. Otherwise, we will consign ourselves long-term to the mercy of international markets and an increasingly variable climate.
Yes, we should start paying more attention to the carbon content of fuels, and this does represent a change from the traditional direction of energy policy. But take a look at the final sentence: "otherwise, we will consign ourselves long-term to the mercy of international markets..." How is this an "otherwise"? The point they're presumably trying to make here is that in the past, we've focused mainly on the security of energy supplies -- i.e., not consigning ourselves long-term to the mercy of international markets -- and that now we will also have to focus on the carbon content of fuels. "The mercy of international markets" is a supply consideration, not a carbon one. They claim that part of our current problem is an single-minded focus on supply, and then say that unless we change away from this focus on supply, we will be at the mercy of supply problems

It all makes very little sense. Perhaps they're trying to say that carbon-intensive fuels always have supply problems, and that cutting on carbon emissions is ultimately the best way to limit our exposure to this scarcity, but then they're wrong: coal is the most carbon-intensive fuel around, but it's in fairly abundant and secure supply.

You might think that this is nitpicking, but I think it's a very revealing example of the style that characterizes their entire report -- generic bits of policy-talk stuffed together with no underlying understanding of how the economy actually works. Although many of their proposals are good, even the solid ideas are buried by a gush of vapid "transformation" rhetoric. To wit:
What has been missing to date is the political will in Washington to seize the energy moment, put in place a series of tough, mandatory rules of the road, back them up with targeted government investments, and begin the work of transforming our economy. The old way of addressing environmental issues apart from the main workings of the economy—as “externalities” or “amenities” in the language of economics—no longer applies. We are confronted now with an issue that is paramount to the preservation of our environment and the sustainability of our eco-systems as well as critical to our national security and central to our hope for a new era of economic growth and prosperity.
Huh? How does the concept of addressing carbon as an externality "no longer apply"? What does this even mean?

The essay on international trade, "Virtuous Circle: Strengthening Broad-Based Global Progress in Living Standards," is even worse. It loves to talk about "cooperative" approaches to policy, US "leadership" in achieving "strategic objectives," and so on. But whenever it descends from its seemingly infinite reserve of generic policy jargon and attempts (even for a moment) to deal with actual facts, it reveals a shocking lack of understanding:
Globalization has helped lift an impressive 300 million (mainly Chinese) people out of extreme poverty over the past two decades, but it has also exacerbated inequality among and within most countries, quite significantly in many cases. The persistence of structural economic biases that promote export production over domestic consumption in emerging economies contributes to this trend or, at a minimum, represents a missed opportunity for reducing it and making economic growth in them and the world economy as a whole more resilient.
Actually, world inequality has been declining, primarily due to the export-driven Chinese growth that the authors are so ready to dismiss. Yes, the Chinese population could be doing a lot better if the government stopped trying to prop up the dollar and let its people enjoy the full products of their labor, but it's certainly not clear that this bias toward export promotion is bad for economic growth, like the authors claim. In fact, the entire point of artificially undervaluing a currency to promote exports is that exports are a critical sector for growth in developing economies, and that the compounded gains of rapid economic growth will overwhelm the welfare loss from lesser consumption in the short run. I'm not sure whether this logic is valid or not -- while economists traditionally have been skeptical, Dani Rodrik is presenting a decent case -- but it's simply obtuse of the authors not to mention (or recognize) it at all.

The paper goes on in similar fashion: a serious-sounding list of worries about globalization that in reality reflects little more than paranoia about "unfair" trade practices and an astonishingly crude understanding of the theory behind international trade. It's disappointing, but it shouldn't be surprising, given that none of the authors of the major CAP economic policy documents have any actual economics training. (click through to the biographies)

And this brings us to one of the major reasons I'm excited about an Obama administration. Although he sometimes descends into shady, CAP-like rhetoric about globalization and "green jobs" on the stump, he's put legitimately qualified people like Jason Furman and Austan Goolsbee in charge of formulating his campaign's economic policy. I'm sick of hacks with law degrees posing as experts on the economy (as I hope I've demonstrated, the results aren't pretty) and I can't wait for a return to boring, economically-minded liberal incrementalism.

A long-winded discussion of redistribution

Many of you have probably seen this excellent chart displaying the effects of the Obama and McCain tax plans according to the shares of population in each income category (from chartjunk):

Even though Obama's plan raises almost $1.3 billion more than McCain's, it results in drastically lower taxes for the vast majority of the population. Look at the -5.5% in the bottom income category: a 5.5% decline in taxes paid as a share of income. That's huge.

Say that we're trying to measure the impact on human welfare from these tax plans. One reasonable function for individual utility is "u = log c": the benefit to each person from a level of consumption c is the logarithm of that consumption. Under this function, no matter how rich you currently are, an increase of X% in your ability to consume will benefit you the same amount: someone making $100,000 a year will benefit about the same from an additional $50,000 as someone making $20,000 will benefit from $10,000. 

This already has dramatic implications. Under this metric, for instance, total welfare would skyrocket if we took $500,000 from everyone making $1 million and distributed it to 50 people making $20,000 apiece. After all, an extra dollar for someone who consumes $1 million every year is 50 times less valuable than it is for someone making $20,000. Calculating the relative merits of Obama's plan and McCain's plan becomes easy: just compare the area of the red and blue bars in the graph above, which are already weighted by population. (Obama's plan wins.) But even logarithmic utility seems to assign a little too much importance to income gains among the rich. Is a doubling in a millionaire's wealth really as valuable as a doubling in a poverty-line family's wealth?

It turns out that "u = log c" is part of a family of functions that economists call "constant relative risk aversion" (CRRA) utility functions. For a number of theoretical and empirical reasons, economists like to use these functions to measure individual utility. The one important parameter in these functions is called the "risk aversion" parameter, and it measures how the benefit from additional consumption decreases as someone consumes more and more. A risk aversion parameter of 1 gives us the "u=log c" equation, where the usefulness of an extra dollar declines in inverse proportion to wealth. Based on the data, most economists consider this to be a lower bound for the risk aversion parameter: it's probably at least a little higher, closer to 2. This makes a big difference, since at a risk aversion parameter of 2 the importance of a marginal dollar is in inverse proportion to the square of consumption. Suddenly, money for someone with $20,000 a year is 2500 times as important as money for someone with $1 million. (As you might guess, a risk aversion parameter of 3 means that the importance of each dollar declines in proportion to the cube of total comsumption, and so on...) 

Obviously, there is a limit to redistribution: at some point, we hit the peak of the Laffer Curve, and taxes begin to discourage work so heavily that raising them actually loses us revenue. Somewhere before this peak, we maximize the social utility that can be achieved by changing the tax structure, as confiscatory marginal tax rates discourage so much value creation that they arguably lower welfare across the board. But we're anywhere near either of these points -- certainly we're not at the peak of the Laffer Curve, and in my admittedly incomplete mental model of the economy, our marginal rates are far, far below the level at which they start damaging the lower and middle-income families that should predominate in any social welfare calculation.

As far as I'm concerned, the social welfare implications of even mild redistribution, like that in Obama's plan, are so vast that they swamp almost every other consideration. Maybe I sound like a utilitarian caricature, summing up "CRRA utility functions" and making public policy into a cold numbers game. I use these methods, however, because they're good reflections of reality: they embody eminently reasonable assumptions about the relative value of wealth, and such assumptions are ultimately necessary for us to decide the appropriate amount of redistribution in society. Lowering taxes on a fmaily that already makes $10 million simply isn't an important social priority; to me, it's only important as an incentive for work and wealth creation (which benefits people further down the ladder), and at the margin we have ample policy space for higher marginal rates and a more redistributive tax scheme in general.

Granted, some people have different philosophical priors about the proper role of government. They think that government should stay out of the "social welfare" business entirely and concern itself only with the maintenance of basic public institutions. But for those of us who don't hew exclusively to this line -- I, for one, lie at some indeterminate position between libertarianism to utilitarianism -- there is no reason to oppose a more progressive tax policy, especially the mild and reasonable kind espoused by Obama.

Unemployment is the problem

In his Sunday commentary in the Washington Post, "Quit Doling Out That Bad Economy Line," Donald Luskin used the last quarter's reported GDP growth, which came at an annualized rate of 3.3%, to claim that the economy is actually in great shape. Most people found this self-evidently absurd. But even putting aside the much lower GDP growth rates in previous quarters (conveniently ignored by Luskin) and the likelihood that the most recent result will be revised downward as more data comes in, his approach is fundamentally misplaced. The most serious consequence of recessions isn't a slight drop in GDP; it's the unemployment that should concern us.

Back in the 1980s, economist Robert Lucas attempted to mathematically estimate the cost of fluctuations in economic output. His idea was simple: what price should we pay in exchange for a permanent smoothing of the business cycle? He wanted to estimate what the cost of growth that isn't stable -- maybe it's -1% one year and 5% the next, instead of a consistent 2% -- really is.

His result was shocking. Using reasonable values for all the relevant economic parameters, including risk aversion, he claimed that we should only be willing to pay the equivalent of .1% of lifetime consumption, even for a complete smoothing of economic growth. It's hard to imagine any theoretical result less consistent with the immense weight that voters and politicians place on preventing recessions.

But there was a catch: Lucas's model used the concept of a "representative agent," where losses in production are felt by a single entity that "represents" consumers more generally. This isn't necessarily bad -- often economists need representative agents to make their models tractable -- but in this case, it's a critical decision. Temporary jitters in the growth rate hardly make any difference to the welfare of a single agent. If your wages fall 1% this year and increase 5% next year, you might be a little annoyed, but your life won't drastically differ from the parallel universe where you receive constant raises of 2%. 

The real economy, however, consists of many different consumers, some of whom feel the effects of a recession much more intensely than others.  A 100% drop in income for one person, obviously, is a lot worse than a 1% drop for 100 people. If we average out the effects of a recession using a "representative agent," we'll barely begin to capture the pain that it causes, making Lucas's 0.1% figure an extreme underestimate.

Lucas's calculation, however, does have some value: it shows us what the real problem with economic fluctuations is. In a world without unemployment, where gains and losses are shared among all consumers, we have little reason to care about the perfect stability of the economy. But in a world where some families experience sudden, severe changes in income, we have a very serious problem to address.

This is why Luskin's fixation on short-run GDP is so wrong. Our current downturn is painful because it's throwing millions more people out of work. That's much more important than transitory fluctuations in GDP.

Wednesday, September 10, 2008

Why the core CPI is so important

Over the past few months, I've seen a profusion of commentators angrily insisting that the "core CPI," a widely-followed measure of inflation that excludes food and energy prices, is some kind of dark conspiracy against the public. It's hard for me to think of any complaint that reveals a deeper level of economic ignorance.

The Federal Reserve follows consumer prices because it has a mandate to prevent runaway inflation in the economy. If it detects embedded inflation, it has one basic tool at its disposal: massive tightening of monetary policy through higher interest rates. Unfortunately, this is a highly contractionary measure, and in the past it's resulted in recessions like the one in the early 1980s, where unemployment peaked at an astonishing 10.8%, the highest level in the post-war era. Since combating inflation through tight monetary policy is so painful, it's critical that we only do so when it's absolutely necessary -- when there are inflationary expectations embedded in the economy.

Now we come to the purpose of the core CPI. Food and energy prices are much, much more volatile than other prices in the economy, and a sudden jump in the price of oil can lead to quarterly inflation that appears alarmingly high. If the Federal Reserve sees that the CPI is rising too rapidly, it's critical to know whether this increase comes in all sectors of the economy, in which case it signals embedded inflation that must be confronted with monetary policy, or whether it's mainly a result of sudden jumps in commodity prices like oil. A failure to distinguish between these two kinds of inflation can lead to unnecessary recession: the Fed mistakes transitory commodity-price inflation for deeper inflationary sickness, and induces massive unemployment through interest rate hikes in a misguided attempt to stem the problem. Needless to say, this benefits nobody.

Since the main consequence of failing to isolate core CPI is unnecessary recession and unemployment, it's laughable to think of core CPI as an elitist device designed to ignore the plight of the masses. Yet this is precisely the line taken by economic ignoramuses who bash core CPI, like the embarrassingly dim Alan Abelson:
And, of course, fresh fuel for the bulls came the very next morning as that same Coué chorus -- or should we more properly call them "core-ists" -- seized on release of the consumer-price index as confirmation that all's well on the inflation front. What better proof, they exulted, than the miserly 0.1% uptick in the CPI -- once you take out pesky items like food and energy. As for the 0.7% rise in the CPI when food and energy are included, "pshaw," they scoffed, that's just the "headline number"!

For the life of us, we must confess, we don't quite understand why "headline" is a pejorative. Do those contemptuous of us benighted souls who award it more serious notice laugh and sneer when a banner headline in their daily proclaims, "War Is Declared" or "World's Tallest Building Collapses" or "Mets Win"?
The worst part is that core CPI isn't used by the government for anything other than evaluating monetary policy. As the Bureau of Labor Statistics notes, all the prominent legislative uses of the CPI -- like cost-of-living adjustments to Social Security -- rely on a full statistic that includes food and energy. Eliminating core CPI would have literally no consequence other than the unnecessary misery and unemployment caused by attempts to eliminate inflationary expectations that don't actually exist.

At least there's an upside to all this nonsense. Writers who bash the core CPI are offering you a very clear and valuable signal: they don't have the foggiest understanding of economic policy. Stay away from them.

Adventures in basic logic

Rereading EPI's anti-Wal-Mart manifesto this morning, I noticed an... interesting argument:
There is another point to be raised here about the implicit horse-race between prices and wages that underlines the Wal-Mart debate. Essentially, the defenders of Wal-Mart argue that the price-depressing effects of Wal-Mart outrun the wage-depressing effect, leading to rising purchasing power for American workers. However, the prices that are reduced through Wal-Mart's expansion constitute an ever-shrinking share of American families' expenditures.
Yes, we've seen a decline in the importance of the disposable goods purchased at Wal-Mart to the average family's budget. As EPI notes, this implies that the lower prices at Wal-Mart and similar big-box retailers are becoming less important to family prosperity. But the decline in the importance of retail goods relative to the economy at large also implies that the retail sector is becoming less prominent in the labor market, and therefore that the supposed wage-depressing effects of Wal-Mart are also becoming less important. This doesn't alter the trade-off (if one exists) between lower prices and higher wages in any clear direction. It merely means that our choice -- whatever it is -- will have smaller consequences. That's hardly justification for implementing the wrong policy!

How did the authors, so eager to note that Wal-Mart's lower prices are becoming less important to families, fail to realize that this also means that the chain's negative effects are diminishing in significance, and that two effects, taken together, leave the basic debate unchanged? I'll leave that for the reader to decide...

Tuesday, September 09, 2008

Extreme inequality

In a recent post about inequality, Matt Zeitlin points out the critical difference between "90-10" inequality, where the share of income going to earners around the 90th percentile shoots up, and inequality within the top decile, where income is further skewed toward the richest of the rich. While the two are often wrapped together into a more generic discussion about "inequality," they are arguably separate phenomena with very different causes and policy implications. Matt explains further, referencing a piece by Robert Gordon and Ian Dew-Becker:
One type of inequality is “90-10″ inequality. This is the standard, skills based technological change explanation of inequality. Basically, managers got more money and benefited the most from productivity gains, which were unchained from median wages in the late 1970s. There’s also the unbounded increase in the college wage premium, which is no longer cyclical. This is the 90th percentile, and above, running away from everyone else - or, everyone else staying behind, while the 90th percentile’s income inequality shoots up.

But that’s not the entire inequality story, the other is inequality within the top decile. This is the inequality that liberals and progressives find so objectionable. This inequality that isn’t driven so much by technological change, lagging minimum wage, or decreased unionization, but instead by the “superstar effect” (JK Rowling can sell more books than any other author in history) or the institutional failures behind skyrocketing CEO pay. And the only way to address that inequality is massively increasing marginal income rates.

In an Obama administration, we’ll probably see a bunch of policies that address 90-10 inequality, along with modest tax increases on the rich. But I don’t think we’ll see a decrease in the type of inequality that people really don’t like. Even if median wages go up, they’ll still be far outpaced by wages at the top, which have gone up by 121 percent at the 99th percentile, 236 percent at the 99.9th percentile and 617 percent at the 99.99th percentile. We’ll be living in interesting times, to say the least.
I completely agree. I'm afraid, frankly, that much of the political rhetoric about extreme inequality at the top has little policy substance. Unless you have some way to change corporate governance and bring down CEO pay -- easier said than done, especially when there's no consensus that CEO pay is truly inflated relative to the fundamentals -- your only avenue for redistributing the incomes of the super-rich is taxation. Sometimes this is a simple matter, as it is with closing the loophole used by hedge fund managers, but in general it's much more complicated. Past a certain point, higher marginal tax rates on the rich become excruciatingly inefficient as a revenue source, and by discouraging work and entrepreneurship they begin to do more harm than good.

So we've lost our manufacturing base?

After hearing another protectionist complain that the United States has "completely lost" its manufacturing base, I decided to make a list of the ten countries with the highest manufacturing total value added. These statistics are from the World Bank's World Development Indicators, and give 2005 values expressed in constant 2000 dollars:
  1. United States: $1,680,400,056,320
  2. Japan: $1,100,973,867,008
  3. China: $645,806,030,848
  4. Germany: $420,691,902,464
  5. United Kingdom: $221,911,023,616
  6. France: $200,727,838,720
  7. Italy: $188,876,914,688
  8. South Korea: $184,509,972,480
  9. Canada: $127,123,423,232
  10. Brazil: $110,720,647,168
Listening to Lou Dobbs every night, you'd never know that the United States still leads the world in manufacturing output -- and by a healthy margin! No one disputes that the US economy has become substantially less manufacturing intensive over the last several decades, or that it is in relative terms much less dependent on manufacturing that other world economies. In my view, this is because US firms have found it optimal to specialize in high-skill design and idea creation rather than more labor-intensive manufacturing. But our economy is still so enormous and productive that it has the highest manufacturing output in the world! It's an interesting tidbit to keep in mind, especially for those people under the impression that because we don't manufacture sneakers anymore, we no longer have a manufacturing base at all.


Apologies that I haven't posted for so long -- I was sick over the weekend and had neither the energy nor the free time to write anything meaningful.

With that said, I'd like to highlight a tiny segment from one of Ezra Klein's posts today. It's a throwaway line, not at all central to his larger point (which is actually quite good), but it's a good example of the unfortunate, lingering presence of "competitiveness" rhetoric in popular economic discourse today:
But the next major blow -- and we're talking larger than 3,000 dead and our sense of security shattered -- is fairly likely to be a pandemic or global economic dislocation stemming from global warming or a sustained reduction in American living standards as we find ourselves incapable of competing as the global labor force doubles. (italics added)
It is true in principle that a decline in a country's "competitiveness" can cause a significant reduction in living standards. For instance, if a country has an extremely specialized economy, where it uses export revenue from its expertise in some particular sector to import the majority of essential consumer goods, then the rise of other countries that compete more cheaply or efficiently in the same sector can lead to significant economic dislocation. Note, however, that this story only makes sense if imports are essential to the country's consumption. If a country supports that vast majority of its consumption through domestic production, changes in the international economy will only make marginal differences to domestic well-being.

As it turns out, this is the case for the United States: according to the most recent figures, imports account for only 18.4% of US GDP. If, by some great cataclysm, America's terms of trade -- the relative prices of its exports to its imports -- decreased by 50%, making imports effectively twice as expensive, the worst that could happen is a decline in US purchasing power by this amount, 18.4% of GDP. This is a substantial loss, but it could be erased with only a decade of GDP growth. And of course, in the real world, the losses would be much lower, because consumers in the United States would substitute away from the suddenly much more expensive imports toward domestically produced goods. Thus even in an absurdly inflated hypothetical scenario, it's hard to see a loss of "competitiveness" doing serious damage to Americans' standard of living.

In one of his best 90s-era pieces, Competitiveness -- A Dangerous Obsession, Paul Krugman eviscerated competitiveness rhetoric and revealed it to be a product of economic ignorance:
After all, the rhetoric of competitiveness -- the view that, in the words of President Clinton, each nation is "like a big corporation competing in the global marketplace" -- has become pervasive among opinion leaders throughout the world. People who believe themselves to be sophisticated about the subject take it for granted that the economic problem facing any modern nation is essentially one of competing on world markets -- that the United States and Japan are competitors in the same sense that Coca-Cola competes with Pepsi -- and are unaware that anyone might seriously question that proposition. Every few months a new best-seller warns the American public of the dire consequences of losing the "race" for the 21st century. A whole industry of councils on competitiveness, "geo-economists" and managed trade theorists has sprung up in Washington. Many of these people, having diagnosed America's economic problems in much the same terms as Delors did Europe's, are now in the highest reaches of the Clinton administration formulating economic and trade policy for the United States...

The idea that a country's economic fortunes are largely determined by its success on world markets is a hypothesis, not a necessary truth; and as a practical, empirical matter, that hypothesis is flatly wrong. That is, it is simply not the case that the world's leading nations are to any important degree in economic competition with each other, or that any of their major economic problems can be attributed to failures to compete on world markets. The growing obsession in most advanced nations with international competitiveness should be seen, not as a well-founded concern, but as a view held in the face of overwhelming contrary evidence. And yet it is clearly a view that people very much want to hold -- a desire to believe that is reflected in a remarkable tendency of those who preach the doctrine of competitiveness to support their case with careless, flawed arithmetic.
Krugman's case has both basic logic and empirical evidence behind it. He proceeds to list some basic calculations, which demonstrate that growth in major industrial countries has been almost entirely determined by domestic considerations, and sums up his case accordingly. I hate to quote such a large portion of the article, but it's so good that I can't resist:
In each case, the growth rate of living standards essentially equals the growth rate of domestic productivity -- not productivity relative to competitors, but simply domestic productivity. Even though world trade is larger than ever before, national living standards are overwhelmingly determined by domestic factors rather than by some competition for world markets.

How can this be in our interdependent world? Part of the answer is that the world is not as interdependent as you might think: countries are nothing at all like corporations. Even today, U.S. exports are only 10 percent of the value-added in the economy (which is equal to gnp). That is, the United States is still almost 90 percent an economy that produces goods and services for its own use. By contrast, even the largest corporation sells hardly any of its output to its own workers; the "exports" of General Motors -- its sales to people who do not work there -- are virtually all of its sales, which are more than 2.5 times the corporation's value-added.

Moreover, countries do not compete with each other the way corporations do. Coke and Pepsi are almost purely rivals: only a negligible fraction of Coca-Cola's sales go to Pepsi workers, only a negligible fraction of the goods Coca-Cola workers buy are Pepsi products. So if Pepsi is successful, it tends to be at Coke's expense. But the major industrial countries, while they sell products that compete with each other, are also each other's main export markets and each other's main suppliers of useful imports. If the European economy does well, it need not be at U.S. expense; indeed, if anything a successful European economy is likely to help the U.S. economy by providing it with larger markets and selling it goods of superior quality at lower prices.

International trade, then, is not a zero-sum game. When productivity rises in Japan, the main result is a rise in Japanese real wages; American or European wages are in principle at least as likely to rise as to fall, and in practice seem to be virtually unaffected.

It would be possible to belabor the point, but the moral is clear: while competitive problems could arise in principle, as a practical, empirical matter the major nations of the world are not to any significant degree in economic competition with each other. Of course, there is always a rivalry for status and power -- countries that grow faster will see their political rank rise. So it is always interesting to compare countries. But asserting that Japanese growth diminishes U.S. status is very different from saying that it reduces the U.S. standard of living -- and it is the latter that the rhetoric of competitiveness asserts.
One might hope that Krugman's decade-long crusade against the purveyors of "Pop Internationalism" would have made a significant dent in the rhetoric of popular economic discussion. Alas, hardly any difference is perceptible, and policy proposals are almost invariably dotted with references to our "competitiveness" in a "changing global economy." Krugman, meanwhile, has gone on to fame as a generalist political commentator, and his brilliant earlier work as an economic popularizer sits mostly forgotten. Read and learn.

Thursday, September 04, 2008


Noting Palin's promise to be an advocate for special needs children, Mark Schmitt makes a great point:
This is my least favorite trait in modern conservatism -- the carving out of a sympathetic exception for the single family need or health problem that you have personal experience with. When a man did it -- Senator Gordon Smith of Oregon -- I called it "Miss America Conservatism," in the sense that each Republican has his or her little platform issue -- in Smith's case, mental health funding, because of his son's suicide -- that shows their soft side, and then they go back to the demeaning pageant of cutting taxes and slashing Medicaid. The lesson in having a child with special needs is not "we need more attention for kids with special needs," it should be, "life hands out lots of difficult circumstances and lots of families need different kinds of help, so we're all in it together."
This has always annoyed me as well. You're certainly entitled to oppose government efforts to help people with special problems or needs, but you can't turn around and advocate an expansion of government into the few areas where you happen to have a searing personal experience.

Wednesday, September 03, 2008

Why the One Percent Doctrine is wrong

After I wrote my post on McCain and the risk of nuclear war this afternoon, a commenter pointed out that my perspective on "1% risks" appears to resemble the "One Percent Doctrine" advocated by Dick Cheney and made famous in Ron Suskind's book. According to Cheney's philosophy:
"If there was even a 1 percent chance of terrorists getting a weapon of mass destruction -- and there has been a small probability of such an occurrence for some time -- the United States must now act as if it were a certainty."
This is a ludicrous idea. If we think that some event has a 1% chance of happening, we should simply treat it as if it has a 1% chance of happening. Proper evaluation of risk demands our best estimates of probability and damage, and artificially inflating our estimate from 1% to 100% accomplishes nothing.

I think I understand the intuition behind the One Percent Doctrine: the idea is that some events are so disastrous that they should dominate our decision-making even when they are only expected to occur with low probability. As I mentioned in my post, this is certainly true of all-out nuclear war, which entails the destruction of human life on Earth as we know it. Even here, however, I don't advocate overstating our probability estimates; I simply think that a 1% probability of nuclear winter is intolerable on its own terms.

Meanwhile, I don't think that "terrorists getting a weapon of mass destruction" is nearly enough of a catastrophe that a 1% chance should always shape our policy decisions. Although there's a natural tendency to shrug off the difference between large disasters and overwhelmingly large disasters, this distinction is actually critical. While a few weapons of mass destruction in the hands of terrorists might result in millions of deaths and some cities destroyed, an exchange of thousands of nuclear warheads would cause billions of deaths and the destruction of almost the entire world. To be sure, both would be cataclysmic, but the latter is at least a thousand times worse than the former.

In fact, there is a good case to be made that a 1% probability of a million deaths in a terrorist attack should be no more significant than many of the other considerations involved in foreign policy. After all, if there is a 1% chance of a million people dying, then the "average" number of deaths is 10,000. This is well within an order of magnitude of the number of troop deaths in the Iraq and Afghanistan campaigns, and it's far below the total number of fatalities believed to be a result of the Iraq war. Granted, our analysis of unlikely catastrophes should be more sophisticated than a simple probability-weighted average of deaths, but I think that such a calculation does provide a good starting point for estimating magnitude. The risk of terrorist attacks isn't grave enough that we should roam the world in quixotic pursuit of whatever "threats" appear on the horizon.

Give only to one charity

When you've long held a position that runs against conventional wisdom, and when no one else seems to notice the basic insight that led you there, discovering that same idea was put forward in a well-known online publication is quite the thrill. Such was the case when I first encountered Steven Landsburg and this wonderful little 1997 article arguing that we should only give to one charity:
CARE is a noble organization that fights starvation. It would like your support. The American Cancer Society is a noble organization that fights disease. It would like your support, too. Here's my advice: If you're feeling very charitable, give generously—but don't give to both of them.

Giving to either agency is a choice attached to a clear moral judgment. When you give $100 to CARE, you assert that CARE is worthier than the cancer society. Having made that judgment, you are morally bound to apply it to your next $100 donation. Giving $100 to the cancer society tomorrow means admitting that you were wrong to give $100 to CARE today.

You might protest that you diversify because you don't know enough to make a firm judgment about where your money will do the most good. But that argument won't fly. Your contribution to CARE says that in your best (though possibly flawed) judgment, and in view of the (admittedly incomplete) information at your disposal, CARE is worthier than the cancer society. If that's your best judgment when you shell out your first $100, it should be your best judgment when you shell out your second $100.
Exactly! In the end, you will always have to use some estimate of the amount that a charity will benefit from each additional dollar. If you think that charity X will (on average) save one life for every $100,000 that it receives, and charity Y will (on average) save one life for every $200,000, you should give your contributions exclusively to charity X. Maybe you're wrong about charity X being superior, but the mere presence of uncertainty shouldn't affect your decision. If your best guess is that X is better than Y, then to the best of your knowledge, diverting money from X to Y will cost lives.

Many people complain that it's wise to "diversify," which demonstrates that they either (1) don't understand the economic rationale behind diversification, or (2) admit that they're giving for their own enjoyment, not for the benefit of the recipients. As Landsburg explains:
When it comes to managing your personal portfolio, economists will tell you to diversify. When it comes to handling the rest of your life, we give you exactly the same advice. It's a bad idea to spend all your leisure time playing golf; you'll probably be happier if you occasionally watch movies or go sailing or talk to your children.

So why is charity different? Here's the reason: An investment in Microsoft can make a serious dent in the problem of adding some high-tech stocks to your portfolio; now it's time to move on to other investment goals. Two hours on the golf course makes a serious dent in the problem of getting some exercise; maybe it's time to see what else in life is worthy of attention. But no matter how much you give to CARE, you will never make a serious dent in the problem of starving children. The problem is just too big; behind every starving child is another equally deserving child...

Here's a thought experiment for charitable diversifiers. Suppose you plan to give $100 to CARE today and $100 to the American Cancer Society tomorrow. Suppose I mention that I plan to give $100 to CARE today myself. Do you say, "Oh, then I can skip my CARE contribution and go directly on to the American Cancer Society?" I bet not.

But if my $100 contribution to CARE does not stop you from making CARE your first priority, then why should your $100 contribution to CARE (today) stop you from making CARE your first priority tomorrow? Apparently you believe that your $100 is somehow more effective or more important than my $100. That's either a delusion of grandeur or an elevation of your own desire for satisfaction above the recipients' need for food.
Unless you are so fabulously wealthy (or pick such obscure charities) that giving all your money to a single organization would saturate its finances and cause its per-dollar effectiveness to decline, there is simply no way that you can justify diversifying your contributions as a way of helping those in need. For those who are quantitatively inclined, Landsburg notes that this logic even has a simple mathematical basis.

You might say that some amount of egotism motivating charity is inevitable. I'd agree. But I also think that we should work to stop our vanity from deciding which urgent causes deserve our attention. If we're self-aware enough to realize that personal satisfaction is our main reason for giving, then we should also try to make sure that our personal satisfaction lines up with actual benefits. As I wrote long ago* on this topic:
There's still cause for optimism. If we are really interested in feeling "good" about our actions, then we need to bring perceived good closer to actual good. How? To put it bluntly, we need more information stuffed down our throats. If the nation is headed for economic disaster because we elected careless leaders, let us see it. If people are starving because we aren't donating enough food, let us see it. If they are dying because we aren't spending enough for their treatment, let us see it. Maybe, if we're confronted with hard reality on a daily basis, our desire to be good will lead us to genuinely good actions.

*I actually wrote about this in my application essay for the University of Chicago. After procrastinating until the last possible night before the online application was due, I was in desperate need of an essay topic, and ultimately settled upon how our patterns of charity demonstrate an essential selfishness, using much the same argument that Landsburg makes in his article. (I also made a clumsy analogy to vitamin pills that was way too strange to discuss here.) Ever since, I've been curious what the people reading my application thought of it. I suppose that couldn't have been too bad, since they did give me a full scholarship -- maybe there was an economist on the committee!

So true

Bryon York is uncharacteristically dead-on at the Corner:

"I don't usually engage in these scenarios, but I'll do it here. If the Obamas had a 17 year-old daughter who was unmarried and pregnant by a tough-talking black kid, my guess is if that they all appeared onstage at a Democratic convention and the delegates were cheering wildly, a number of conservatives might be discussing the issue of dysfunctional black families."

Risk and McCain

One of the key points in my recent discussion of climate change is the importance of unlikely but extreme disaster scenarios. Unfortunately, voters usually aren't responsive to the magnitude of such risks, even if an objective analysis demonstrates that they are critical to proper policymaking. This raises an interesting question: what other risks are unaddressed in the political mainstream?

One obvious answer comes to mind: nuclear war. Of course, we can't say with certainty that there is a nontrivial chance of nuclear war. The political and structural factors leading to nuclear conflict are far too complicated to model mathematically, and the very fact that we are around to discuss the question indicates that we have no empirical evidence to construct a probability distribution. We're left with vague subjective considerations, ill-suited for rigorous risk analysis.

But any eventuality that involves the destruction of the world as we know it demands our urgent attention, even if we can't estimate its likelihood with any precision. A principled refusal to consider risks so opaque that they lie "outside of a probability distribution" (as Jim Manzi puts it in another context) may cause us to ignore the most dangerous possibility of our time.

Consider the implications of a 1% chance of nuclear war between the United States and Russia within the next decade. If you take the position (as I do) that the near total destruction of humanity is thousands of times worse than practically any other policy-relevant scenario at hand, even a 1% chance of nuclear confrontation should be enough to dominate your decision-making. If there is any way to eliminate that 1% possibility, or even to clip it a little to 0.9%, you should pursue it at the expense of almost all ordinary political considerations.

Do I know that the probability is 1%, or even within an order of magnitude of 1%? Of course not -- again, there is no rigorous way to estimate the likelihood of an event as theoretically and empirically hard to quantify as nuclear war. But retreating to my intuition for a moment, I think that in many political environments, 1% is a very plausible estimate, perhaps even a lower bound. A foreign policy that aggressively confronts Russia within its sphere of influence, reviving the tension that existed during the Cold War, creates a very real chance of all-out conflict.

How? The possibilities are almost too numerous to list. Maybe a leader on either side will prove unpredictable and irrational, moving a situation beyond the point of no return. Maybe America's effort to install missile defense systems on Russia's periphery will make a paranoid Russian government suspect a first-strike attempt, and jump to retaliate after some faulty sensor readings. Maybe a deranged jihadist organization, seeking to destroy both America and Russia, will take advantage of existing tension by detonating a nuclear weapon and disguising it as an attack from the opposing nuclear superpower. Maybe an atmosphere of suspicion, coupled with thousands of massively destructive weapons on hair-trigger alert, will lead to a escalation that we never could have predicted.

A 1% chance of nuclear war, or anything close to it, is unconscionable. And while it is difficult to tell exactly which policies best prevent nuclear disaster, it is easy to determine what we should not do. We should not elect a president who advocates gratuitously insulting Russia by expelling it from the G8, alludes to religion while declaring his solidarity with a nation at war with Russia, and has established a pattern of making major decisions in an impulsive and poorly considered way.

To me, this is the most convincing case against McCain of them all...