The Real Problem With Reinhart-Rogoff
By Laurence Kotlikoff
Carmen Reinhart and Kenneth Rogoff's paper "Growth in a Time of Debt," has come under scrutiny due to a coding error discovered by Thomas Herndon, a Ph.D student at the University of Massachusetts, Amherst. The coding error appears, however, to be inconsequential to the Reinhart-Rogoff finding that periods of low growth coincide with periods of high official government debt. Nor is does it have anything to do with the results reported in Reinhart-Rogoff's outstanding and justifiably famous book, "This Time Is Different."
Research mistakes happen, and this was an honest one made by two of our country's premier economists whose integrity is beyond question.
Unfortunately, there is a much deeper and uncorrectable problem with the debt figures used in Reinhart-Rogoff's paper. This problem is not that some of the data on official government debt were inadvertently omitted from the analysis. The problem is that, in fact, none of Reinhart-Rogoff's official debt series provide a meaningful, as in theoretically well-defined, measure of a country's indebtedness.
The trouble with official data
To the contrary, economic theory makes clear that what's counted as official debt reflects nothing about a country's underlying economic liabilities and everything about what the country does and doesn't call "official borrowing."
It’s easy to get queasy about using official debt numbers to capture true government obligations. Just find a retiree receiving periodic payments from Uncle Sam, some of which are called interest and principal on U.S. Treasury bonds and some of which are called Social Security benefits. Next ask why Reinhart-Rogoff include in their debt measure only the present value of the future bond payments and ignore completely the present value of the Social Security payments?
The answer is there is no economic answer. Lawyers will say that Treasury bills and bonds are backed by the “full faith and credit of the United States.” But these fancy words don’t preclude formal default, let alone informal default via government-caused inflation.
Social Security benefits, on the other hand, are inflation-projected and supported by the 40-million strong American Association of Retired Persons. So each dollar of Social Security’s current $60 trillion debt arguably counts more than a dollar of the public’s $12 trillion holding of U.S. official debt in terms of its likely implied burden on today’s and tomorrow’s taxpayers. Yet, Reinhart and Rogoff include not one penny of Social Security debt in forming their historical time series of Uncle Sam’s obligations. Nor do they include the present values of Medicare and Medicaid’s massive future benefit commitments as well as those of more modest transfer payments.
And what about obligations to future discretionary spending, like maintaining our military, repaving federal highways, or gassing up Air Force One? Why do these expenditures commitments get zero weight in the Reinhart-Rogoff analysis, while expenditure commitments to interest and principal on Treasury bills and bonds, half of which are owned by foreigners, get full weight? Moreover, in their focus on gross official debt, Reinhart-Rogoff ignore all the assets available to cover the government’s bills, particularly the projected time path of all future tax receipts.
Unfortunately, economic theory doesn’t say whether to call a dollar received by the government “taxes” or “borrowing.” Consequently, what gets measured as “official debt” is purely an artifact of what is and isn’t labeled “borrowing.” This is easiest to see in the case of Social Security, where our contributions could be called “borrowing” and part of our future benefits could be called “interest plus principal” on this borrowing.
It's about labeling
In recent decades, Chile, Argentina, Kazakhstan, Russia, and many other countries engaged in precisely this relabeling in “reforming” their pensions. Rather than have workers pay social security “taxes,” it had them contribute to pension funds. It then borrowed the contributions from the funds to pay current beneficiaries. Presto, the contributions were no longer “taxes,” but “borrowing.”
Labeling choices aren’t innocent. At the time of its reform, Chile was reporting a surplus, which the Chilean Navy wanted to tap to buy a used U.S. aircraft carrier. By enacting the “reform,” Labor Minister Jose Pinera was able to tell the admirals, “Sorry, the surplus is gone.” When Argentina, Hungary, and Russia recently reversed course and partially or fully nationalized their pension systems, they were able to relabel some or all workers’ contributions as “taxes” and, thereby, report a smaller deficit.
In the U.S., Europe, and Japan, politicians have spent six decades running massive Ponzi schemes using the word “taxes” rather than “borrowing” to keep them off the books. This secured oldsters’ votes and kept youngsters on board with promises of future benefits. Today, however, there are too few youngsters to pay the oldsters and people are starting to see “pay as you go” for what it is – take as you go.
It’s easy to go back historically and reclassify, for each of Reinhart-Rogoff’s countries, all or part of each country’s annual “tax payments” as “borrowing by the government” and reclassify subsequent annual benefit payments as, in part, “repayment of principal plus interest” on that borrowing. Each relabeling will produce a different time series for government debt and, consequently, a different set of results.
At its core, then, the Reinhart-Rogoff study relates a real economic quantity – the growth rate of GDP – to a non-economic and wholly arbitrary linguistic construct. With the right set of words, one can produce whatever time-paths of country-specific official debts one wants and, thus, whatever correlation between “official” debt and GDP one seeks.
Mind the fiscal gap
The pity here is that we have prominent, well-trained economists, like Reinhart-Rogoff and Paul Krugman, arguing for and against austerity based on “official” debt figures that have absolutely no clothes.
What then should we measure when it comes to the big economic questions associated with the word “debt,” namely are the liabilities being left for future generations more than they can handle and did their accumulation promote consumption over saving, limit investment by limiting saving, and limit growth by limiting investment?
The fiscal gap – the present value difference over the infinite horizon – between all projected spending, including servicing official debt, and all projected taxes is economics label-free measure of the unpaid bills being left to today’s and tomorrow’s children. It’s $222 trillion, based on the Congressional Budget Office’s latest projections. This true measure of our children’s fiscal abyss, which, by the way, grew $11 trillion over the last year, is almost 20 times the size of official debt held by the public!
The relentless postwar expansion in the fiscal gap fueled a truly amazing consumption spree by oldsters that drove our national saving rate from 14 percent in 1950 to 1 percent last year. The ratio of the average oldster’s consumption to the average youngster’s consumption is now more than twice what it was back then. Domestic saving is the main determinate of domestic investment, so it’s no surprise that take as you go has also wiped out most of domestic investment. And less domestic investment has meant slower economic growth. In sum, Reinhart and Rogoff are right. They just aren’t using the right numbers to show they’re right.
Laurence Kotlikoff is an economist at Boston University and co-author of The Clash of Generations.