With tax hikes being anathema to many Republicans, and Democrats similarly loath to cut entitlements, some in Congress and the White House are embracing a way to do both without admitting to either. The idea is to change the index used to measure inflation, employing what is known as the chain-weighted or “chained” consumer price index to determine cost of living entitlement adjustments and changes in tax brackets. It turns out this would slow the pace of growth in government programs while increasing taxes, albeit in a manner subtle enough that it may be hoped to go unnoticed by the bases of both parties.
To appreciate the elegance of this bit of statistical legerdemain, one has to understand the difference between the currently-used unadjusted CPI and the chained CPI. The traditional index is often criticized by economists for not taking into account changes in consumer behavior, such as switching purchases from one type of computer to another as technology and/or prices change, or from steak to chicken as the price of beef rises relative to poultry. Historically, measured inflation would have been suppressed by such substitution (by about 0.38% less per year by some estimates), and therefore so would cost-of-living adjustments to entitlement payouts and the CPI-driven ratcheting up of tax-brackets. The prevailing guess seems to be that a shift to the chained CPI would save the government in excess of $200 billion over the next ten years.
Now while swapping chicken for beef may reduce the rate of increase of the index, it does not reflect how happy consumers are about having been forced by prices and budget constraints to make such substitutions – what economists refer to as the change in utility of consumption. All else being equal, so long as consumers can find something to else to eat as beef prices rise – chicken, Spam, Soylent Green (Gen Y readers: look it up) – the government gets to count no increase in inflation nor any diminution in the quality of life*. Thus, under the chained CPI, the cost of living can be re-defined as the cost of surviving, begging the age-old question: “this is living?”
This indifference to reductions in lifestyle means the usefulness of the chained CPI goes beyond allowing a less than courageous Administration and Congress to effectively cut entitlements and raise taxes without admitting to doing so. As I’ve noted before, inflation represents a real default on the debt of the currency-debasing sovereign or sovereigns (see “Europe Learns To Default The American Way, Restoring Transatlantic Balance Of Irresponsibility”). Should a sovereign’s central bank (e.g. the U.S. Federal Reserve) monetize the debt by purchasing government bonds and issuing currency to do so, then defining inflation downward insulates the budget from the upward spending and tax-bracket adjustments that would otherwise occur. This nominal fiscal sterilization of monetizing the debt may be accompanied by some degree of political cover, at least until the electorate gets tired of eating canned tuna instead of salmon filet, or the markets decide the jig is up and start selling off the bonds and/or currency in question. Given the growth path of the U.S. national debt and a Fed balance sheet already bloated with Treasuries purchased under “quantitative easing,” the “soft” default of monetization becomes that much more tempting when the attending inflation is made statistically stealthier.
To be sure, we should give the chain-weighted CPI credit for adjustments it might be getting right, such as changes in the purchase of goods due to technological advancement. However, here too the devil is in the details, for while not including new technology goods fast enough might bias the unadjusted CPI up by neglecting rapid early price declines, some tech purchases reflect the coercive nature of needing to keep up with the state of the art in order to function in society, as with a new television standard or operating system. Many innovation-driven purchases should be regarded not as the ongoing acquisition of ever cheaper, higher quality gadgets but as shocks; that is, sporadic and, to varying degrees, unanticipated additional forced expenditures increasing the real cost of living. After all, while the price of new computers may reflect a decline in the cost of a byte of memory and digital television look better than analog, one cannot function in society with 64K of RAM and an old-standard TV any more than one could now get by using the stone tools of a neolithic hunter-gatherer. The issue is not just improvements in quality but the social context in which they occur, i.e., the cost of living in modern society. Further, a shift in, say, video display buying from 60” flat-screens to cheap tablets might be driven by constrained budgets as much as innovation, and consequently reflect diminished lifestyle. Quality improvement can be in the eye of the beholder, and their subjective nature may give policymakers leeway to define away inflation even as it becomes more acute.
Just to keep government and central banks honest, economists may have to pay more attention to “pain-weighted” adjustments to CPI – the “pained” CPI, if you will, measuring the cost of getting back to the higher standard of living the chained CPI’s calculation assumes we don’t miss.
As the saying goes, there are three kinds of falsehoods: lies, damned lies, and statistics. That aphorism’s origin is disputed but generally dated to the nineteenth century. The Bureau of Labor Statistics, which calculates the CPI, was founded in 1884.
*N.B. for my economist friends: of course a formerly constant utility under a budget constraint (e.g., achieved under minimized expenditure a la Hicksian demand) may be made unobtainable if the price vector changes sufficiently, excluding the prior indifference curve from the feasible consumption set.