In spite of their static, mechanistic models losing the barest claims to relevance in an environment where ready availability of risk capital has fueled unrivaled levels of dynamic, productivity-boosting innovation, denizens of old-school, hair-shirt economics are sticking to their guns. Unless the Fed resumes its rate-hiking campaign soon, they insist, it’s only a matter of time before demand pressures finally drive inflation significantly higher. Typical of these reactionary fulminations, the Aug. 18 New York Times featured a silent scream by columnist Floyd Norris, aghast at the full calendar of new stock offerings and warning  “the slowdown is vanishing.” Writes Norris: “To fight inflation, the Fed will have to slow the economy. And the lesson of 2000 may be that to do that it will have to apply far more pain than it already has.”

The “lesson” of recent years, of course, is precisely the opposite. If the speed-limit conception of the inflation process had any pertinence, surely there would be some evidence to confirm it after three years averaging better-than 4% growth. But not only has robust growth not heralded rising inflation, aside from temporary distortions in the indexes caused by crude oil. Mounting evidence strongly suggests that zero inflation is now a practical reality.  What is currently registering as inflation in the official data is in fact a statistical illusion, to a significant degree reflecting the statisticians’ inability to accurately measure those New-Economy forces that have been key to the growth pick-up in recent years.  

Output per Hour

Ferreting out the zero inflation reality requires rationalization of the government’s calculations of the rise in output per hour worked – i.e., productivity growth. Under the conventions of national income accounting, inflation is measured as the difference between nominal and real output – as captured by the GDP price deflator. The reliability of these estimates, however, depends critically on the accuracy of productivity measurements generated by the Bureau of Labor Statistics. To the extent that mismeasured productivity growth underestimates gains in real output, it augments the proportion of total production classified as “nominal,” thus overstating inflation. And despite the impressive productivity gains recorded in recent years, the data leaves little doubt that this performance has been substantially underestimated.

The BLS estimates that total non-farm business productivity has grown at an average rate of slightly less than 2.5% on a four-quarter moving average basis over the past three years. While this year’s second quarter posted a one-time year-on-year surge to 5%, the longer-term average likely approximates the bureau’s estimate of current trend productivity growth. At the same time, the BLS shows output per hour in manufacturing, which reached 7% y-o-y growth in the second quarter, growing at a 5% trend rate over the past three years. In other words, trend productivity growth in manufacturing,  according to BEA estimates, has been running at a pace approximately double that of overall non-farm productivity. But here’s the rub: service-sector workers account for about 60% of non-farm output. The government doesn’t release a separate estimate for service sector productivity. But if the BLS estimates that total growth in non-farm output per hour has been less than half that of manufacturing, it must follow that the service sector has shown virtually no productivity acceleration during this period. Under the BLS estimating framework, service sector productivity remains virtually imperceptible, growing at a rate of about 0.5% per year.

Is it plausible to think that the service sector has missed out on the technological revolution that has driven the robust expansion over the past several years? Hardly.  Had service productivity actually been lagging so badly, corporate profits would have been slumping rather than soaring, real wages would at best have remained stagnant, and job losses would have been widespread. Squaring the circle, we find that the GDP price deflator has risen at an average rate of 1.6% over the past three years. If service sector productivity was estimated as being no lower than 2.1% -- less than half the trend rate for manufacturing over this period -- reported inflation would be zero. In all probability, actual service-sector productivity growth is running at a rate no lower than 3-3-3.5%.  If the data were more accurately reflecting real-world conditions, the economy would be seen as experiencing not modest levels of core inflation, but moderate deflation, at a rate of about 1-2%.

Economists have long known that the service sector presents particular measurement challenges, resulting in overstated inflation estimates. In a 1991 report, researchers at the New York Fed found that “errors in measuring service sector inflation arise frequently because of the difficulty of identifying standard output units” for many types of services. “As a consequence, quality and productivity improvements are often lumped into price changes, more often than not exaggerating price increases.” Measurement error frequently arises from the fact that output in a large segment of the service sector is inferred from the number of hours worked, which means that productivity growth – for statistical purposes – is impossible. The problem may have become even more acute in recent years due to the difficulty of measuring the value of “intangible” products, such as software copyrights and biotech patents that have been spawned by the surge in entrepreneurial innovation. “Creativity is poorly measured in the U.S. economy,” says Philadelphia Fed economist Leonard Nakamura in July/August issue of the bank’s Business Review.  “Retail innovations and the proliferation of new products that result from creative activity have made it more difficult to measure the inflation rate. Indeed, our official statistics almost certainly overstate inflation.” 

Evidence that all traces of inflation have been wrung from the economy is entirely consistent with various market price indicators that have shown dollar purchasing power strongly rising over the past several years. Since late 1996, for example, the dollar’s appreciation in terms of both gold and foreign exchange has been nearly identical, at about 28%. Simple extrapolation from past trends, of course, offers no ironclad basis for  forecasting the future. But the recent behavior of the most sensitive, forward-looking monetary signals show no deviation from the strong-dollar path of recent history. Dollar/gold is now trading in a range some $10 below its one-year moving average price of $285 per ounce, and the Fed’s dollar exchange-rate index trades near 14-year highs. Also, the Treasury yield curve remains inverted across the entire maturity spectrum, with inversion of the 30-year/2-year spread averaging about 50 basis points since late January. Market commentators are in near-unanimous agreement that the inversion is primarily a reflection of the Treasury’s long-term debt pay-down program. That rationale fails to explain, however, how the 5-year note can currently be trading at a yield nearly 20 basis points below the 2-year issue when the Treasury has no plans for buying back maturities in the 5-year range. The bottom line is credit market participants would not accept lower yields on longer-term assets were inflation risk driving the discounting process.   

Non-existent inflation should prove to be a boon to dollar-denominated portfolio assets, provided it is sustained. On a historic real-return basis, even current long-term yields around 5.7% present outstanding value in a zero-inflation environment. At present, the 6.5% funds rate poses the primary short-term obstacle to significant additional gains. The betting here, however, is that notwithstanding the current tone of debate, the Fed’s next move is more likely to be a cut than a hike. A potentially significant challenge to the Phillips Curve/NAIRU orthodoxy appears to be underway within the system. At some point, perhaps before year-end, recognition that the overnight rate is out of line with inflation realities could well foster expectations for Fed easing by early next year, setting the stage for powerful rallies in both fixed-income and equities markets.