One thing that is missing from the debate about economic policy is the critical ingredient of humility. Humility isn’t critical for moral reasons, although humility is a virtue, and we would like our politicians to be more virtuous. Instead, humility is a practical good in the economic-policy debate, because if the effects of economic policies were decisive and predictable, then there would never be a recession or non-trivial unemployment. But there are recessions and widespread unemployment, which means either that politicians’ ability to manage the economy is much more limited than our political rhetoric suggests (more likely) or that incumbents are intentionally enacting bad policies that they know will produce recessions and unemployment (less likely).
Politicians have obvious incentives to pretend that they have more knowledge and power than they do. Nearly as much damage is done by the priesthood of professional economists and journalists, who for their own narrow interests also exaggerate what politics can achieve, be those interests professional or political or some combination of the two (assuming they can be distinguished).
This lack of humility produces headlines and sound bites like this one on Sunday from The Atlantic: “Tax Cuts Don’t Lead to Economic Growth, a New 65-Year Study Finds.” The piece itself, by business editor Derek Thompson, isn’t terrible, and one has to assume that, like most writers, he probably isn’t responsible for his headlines. But the headlines dominate political discourse, which is by its democratic nature shallow.
In fact, correlating tax-rate changes to growth rates is very close to being meaningless. That’s because the relevant comparison isn’t between observed growth rates under various tax regimes but between observed growth rates and the growth rates that we would have observed under different tax regimes. That more meaningful comparison has the academically and journalistically undesirable quality of being unknowable. Social scientists frequently measure the wrong factor because it is measurable, when the right factor is not.
Further, the effects of tax changes probably are not immediate in the vast majority of cases. If our theory is that changes in the tax code change incentives for consumers, workers, investors, and firms, then you have a great number of factors that are going to have effects that become manifest over very different time intervals. If you eliminate the sales tax on computers for one month, then you might expect a spike in month-over-month sales for one month, and that is fairly easy to estimate. If you change capital-gains-tax rates, research-and-development credits, capital-investment-write-off rules, etc., then you have a whole different range of temporal variables, since developing a better artificial hip and building a factory to produce that improved artificial hip are very different enterprises, requiring different time commitments. In an economy as complex as ours, such factors probably are not predictable even in principle.
Which is why even very smart people, such as Atlantic writers, produce maddening paragraphs, such as this one from Mr. Thompson: “Well into the 1950s, the top marginal tax rate was above 90%. Today it’s 35%. But both real GDP and real per capita GDP were growing more than twice as fast in the 1950s as in the 2000s. At the same time, the average tax rate paid by the top tenth of a percent fell from about 50% to 25% in the last 60 years, while their share of income increased from 4.2% in 1945 to 12.3% before the recession.”
All of that is trivially true. The tax code in 2012 is different from the tax code in 1955. Lots of other things are different, too: Japan emerged from the postwar rubble to become a major economic power and then went into gentle decline during the subsequent years, the ruins of Europe were rebuilt, a European monetary union was created and then began coming unglued, Germany was reunited, the Soviet Union was disunited, China began to liberalize its economy, a globalized information economy emerged with India and South Korea winning significant places in it, the Internet became a critical economic reality, the population of the planet more than doubled, worldwide markets were integrated, standardized containerization revolutionized shipping, smallpox was eradicated, life expectancies grew in many parts of the world, U.S. birth rates declined . . . and so on. Telling us that tax rates were X in the 1950s and Y in 2012, while growth was A in the 1950s and B today, tells us something approximating nothing.
It certainly doesn’t tell us “Tax Cuts Don’t Lead to Economic Growth.” Try turning it around: What might the sentence “Tax cuts lead to economic growth” even mean? Maybe: “Tax cuts, independent of all other variables, consistently and predictably lead to economic growth”? I very much doubt that anybody who is not a political speechwriter or talking head would argue such a thing. How about: “In some well-defined circumstances, tax reductions may contribute to higher levels of economic growth than probably would have been observed had higher rates prevailed”? Here we have the opposite problem: Does anybody not believe that? Between the data and the headline falls the Shadow.
After 40,000 years of civilization, we very clever creatures still cannot predict the weather with any reliable degree of detailed accuracy more than about a week out. (But some of us still pray for rain.) Scientists who have spent their lifetimes working on extraordinarily specialized problems routinely are baffled by new and unexpected developments. (But some of us still believe the universe is turtles all the way down.) Our highest-paid stock-pickers routinely are outperformed by darts thrown at a board, by kindergartners, and by monkeys. (But some of us still believe in the sure thing.) On and on it goes: Executives reliably make disastrously bad decisions about their own businesses, and most entrepreneurs fail.
In spite of the massive piles of evidence surrounding them, politicians routinely tell us that if we will merely give them the power to do X, then Y surely will follow. The Obama administration predicted that if the stimulus and other policies were enacted, then unemployment would decline to 5.2 percent. (It isn’t 5.2 percent.) Mitt Romney says that if we enact his agenda, the result will be 4 percent growth. Personally, I think that politicians should be goosed with a Taser every time they use the word “percent” in a future-tense sentence. But to be more charitable, let’s instead conclude that such projections should be viewed skeptically.
Unhappily, many economists desire to play kingmaker and therefore lend the prestige of their discipline to the wishful thinking of politics, where arguments are oversimplified to a point that is indistinguishable from dishonesty. They are aided in this by journalists who provide a bridge from the rigorous world of academic research to the standards-free world of political discourse. The result is something like a fairy tale or just-so story. That voters choose to accept such fanciful promises is another piece of evidence that our politics is not rational but ritual.