U.S. Should Privatize Economic Statistics
Mark Twain famously once remarked that there are lies, damned lies and statistics. Famous economists will tell you that there are also lies, damned lies and government statistics. Or, as one savvy pol put it more than 100 years ago: “Figures don’t lie, but liars will figure.”
From GDP to productivity to unemployment to trade, serious debates quietly rage over just how valid the government’s economics statistics are and what they really mean. These econometrics can have an even bigger impact on the quality of everyday life than a Supreme Court ruling. Alas, the government’s numbers don’t stop at merely distorting the statistics; they actually distort the economy. Letting government econometrics shape economic behavior makes about as much sense as a using a fun-house mirror to shave with a straightedge.
Former Council of Economic Advisors Chairman Michael Boskin argued recently in The Times that one of the government’s most important economic statistics was seriously flawed. The presidential commission he recently headed concluded “that changes in the consumer price index overstate changes in the true cost of living by about 1 percentage point per year,” he wrote back in January. “This may seem like a small amount, but when compounded over time, it matters a lot. Because one-third of the federal budget and parts of the tax code are adjusted automatically by changes in the CPI, the impact on the budget over a dozen years of this over-indexing amounts to $1 trillion.”
*
That’s real money. But wait! It gets worse, Boskin wrote: “The overstatement of inflation obviously also sends confusing, inaccurate signals to private-sector workers, savers, investors and renters. Some private contracts are explicitly tied to the CPI and many more are implicitly.” So Boskin understandably argues that the CPI should be recalibrated.
Other econometricians swiftly asked: “Recalibrated for whom?”
Why, for example, should a CPI designed to measure the cost of living for a working-class family be used to determine the inflation rate for pensioners? Indeed, the Bureau of Labor Statistics--the statistical stewards of the CPI--have been tracking an experimental CPI for senior citizens for over a decade now. Their tentative conclusion? The cost of living for the elderly accelerates about 0.2% faster than today’s CPI. The practical reality is that there is a swirl of genuine statistical, economic and social issues shaping the figure the BLS happens to call the CPI. The point? The government’s inflation metric is a convenient fiction rather than anything approaching an econometric fact. The CPI represents a macroeconomic microcosm of the pernicious and pervasive problem of government statistics.
The distortions occur practically everywhere you find federal econometrics. “I have to tell you, I am amazed at how much the financial markets--particularly the bond markets--can be moved by the release of a statistic that people know will not just be revised--but radically revised--a few weeks later,” marveled George Gould, a Treasury undersecretary in the Reagan administration. “It’s clear that there are many people whose trading strategies seem based on gaming the government’s economic statistics.”
On the surface, the obvious response is, “Who cares?” Who cares if a bunch of traders overreact to an inaccurate unemployment or trade deficit figure? The correct numbers ultimately will out, and the markets will adjust themselves accordingly.
*
But, in truth, this reality begs a more fundamental question: Why should the government even be releasing these statistics? Why, in an era of increasing deregulation and privatization, should the government have a virtual monopoly on key economic statistics? What obligates the government to be the market maker in national econometric statistics?
A hundred years ago, this role made sense. If not the government, who else? But just as the rise of a multichannel cable and satellite universe has caused many to question whether we really need public television today, the richness, diversity and complexity of the U.S.--and global--economy should force us to reevaluate our dependence on Washington’s numbers.
Perhaps it makes more sense for the government to phase itself out as an econometrics engine and encourage a vibrant, competitive, innovative marketplace in economic statistics. The book-to-bill ratio in semiconductors, for example, is a private market statistic that is a global economic bellwether. Should we ask the government to administer it? Or should we look for ways to stimulate the private sector to make the government’s economic measurements less important--and thus less distorting--in the national economy?
The answer should be obvious.
*
This shouldn’t be interpreted as a call to slash and burn the federal statistical agencies the way they were a decade ago. On the contrary, for all their flaws, economic statistics can be useful. But the idea that the government should effectively monopolize trade figures or productivity numbers or employment numbers is as anachronistic as saying that the government should set prices for the interstate trucking business and airplane travel and that telecommunications is a “natural monopoly.”
On the contrary, it makes eminent sense for the government to encourage the privatization and outsourcing of many, if not most, of its econometric assessments--much as it has already done with the consumer confidence index. If the government wants to peg a certain portion of its budget to a CPI, let’s create a competitive marketplace to determine which CPI it should be.
Let’s have AFL-CIO CPIs go up against Business Roundtable CPIs go up against Rockefeller Foundation-sponsored CPIs versus National Bureau of Economic Research CPIs versus the WalMart/American Express CPIs versus the Dow Jones/Visa CPIs versus the AARP CPIs. Maybe we should ask states to do their own CPIs and eliminate the idiocy of national aggregation. Let the best metric win.
Too much choice? Too bad. As hard-working and talented as the BLS economists and econometricians may be, their calculations should create de facto statistical monopolies. The government’s CPI should be pegged to the commercial marketplace’s CPI--not the other way around.
*
Of course, the government may have an important coordinating role to play in the statistical battlegrounds of tomorrow. But, really, we have to ask ourselves if we want our government to play a bigger and more influential role in running the numbers that influence our economic behavior, or if we would prefer that the government had a less distortional impact on our economy through its econometric mis-measures.
We’ve deregulated aviation, trucking, telecommunications and banking and seen tremendous boosts in innovation and productivity. Let’s deregulate our econometric apparat and see what happens.
