Global: The Great Productivity Fade
Stephen Roach (New York)
America’s productivity miracle has arguably been the single most important macro development in the global economy in the past 35 years. But the bloom may now be off the rose. Not only is productivity growth going through the usual cyclical slowing of a maturing economic expansion, but there are new reasons to question the secular story. Is it game over for US productivity?
A weaker-than-expected 4Q04 productivity report left little doubt that a classic cyclical slowing is at hand. Output-per-hour in the nonfarm business economy rose at just a 0.8% annual rate on a sequential quarterly basis — less than half the 1.8% pace of the third period and less than one-fifth the 4.4% average annualized pace of the prior 13 quarters. This slowdown is fairly typical of comparable decelerations that have occurred at similar junctures of earlier business cycles. Given the well-known lags between hiring and a cyclical rebound in aggregate demand, productivity growth usually surges in the early stages of recovery. Once, however, businesses become convinced that demand growth is sustainable, work schedules are extended and job creation usually picks up — triggering a classic cyclical slowdown in productivity growth. That’s been the pattern in every US business cycle of the post-World War II period. In that context, the productivity downshift in the second half of 2004 was hardly surprising.
[Roach WTF are you talking about? Productivity has been on the rise for 15 years, although nowhere near as much as is reported, and you are talking as if it just started occurring with this latest "recovery" Mish]
An important test will come at some point in the next six to nine months, when the cyclical noise typically subsides and the underlying productivity trend can then be observed with greater clarity. If productivity growth rebounds to the post-1995 average of 3%, then the magic will still be alive. There are three reasons why I do not believe that will be the case: First, the days of IT-enabled capital deepening could well be drawing to an end. From late 1995 through early 2004, the IT share of total capital equipment went from 36% to 58% in real terms — sufficient to have boosted the capital-spending portion of US GDP from 9.6% to 11% over that period. That also raises the so-called capital-to-labor ratio — mathematically guaranteeing an acceleration in productivity growth while this capital deepening was occurring.
With Corporate America now having made the transition from one technology platform to another, there is good reason to believe that IT saturation could be at hand. Indeed, the IT share of equipment spending actually moved down in the final three quarters of 2004; after peaking at 58% in 1Q04, it ended the year at 57%. This is an extremely rare development for the Information Age, and, in fact, marks the sharpest downturn of the IT share in the post-1980 era. The possibility of IT saturation does not mean that US businesses have changed their minds on the potential payback from these technologies. It simply implies that in the absence of any new spectacular technological breakthroughs, the IT endowment of America’s capital stock may have hit its limit. That would suggest that prospective IT demand would be driven more by obsolescence and replacement rather than by further increases in the capital-to-labor ratio.
[I have no qualms with the above statement in bold. What it means however is that capital spending will drop. Jobs required in the tech sector will be reduced. Roach fails to mention this aspect. Mish]
In the absence of further capital deepening, this impetus to productivity growth would then fade. Even if that conclusion is wrong, it seems highly unlikely that the IT share of capital spending would rise as much in the next ten years as the nearly 20 percentage point increment that occurred over the past decade. Again, that would imply a diminished assist to productivity growth from IT-enabled capital deepening.
The second reason has to do with the potential excesses of cost-cutting. With the absence of pricing leverage and in the context of globalized production platforms — not just in manufacturing but now increasingly in services — Corporate America has moved aggressively to prune excess costs. This trend may have gone too far, taking cost cutting into the slash-and-burn realm. There is strong circumstantial evidence supporting this possibility: Job creation has been far weaker in this recovery than in any other; according to our estimates, private nonfarm payrolls are currently running 9.8 million workers below the profile that would have been implied by the typical business cycle expansion. At the same time, net capacity-enhancing investment in new plant and equipment by US businesses remains about 50% below peak levels hit in 2000.
[Once again Roach misses the boat. Mergers, outsourcing to China etc are by no means ending. Those will continue to cost US jobs. Where he really misses the boat is on hedonics. The GDP simply is NOT expanding anywhere near the published figures. Finally, we are creating jobs, tons of them. Unfortunately the bulj of those jobs we are creating are in China and India. Yes we cut jobs, But.... jobs were gained elsewher in the world. I do not believe those jobs have been accurately accounted for in this "productivity miracle". IMO a completely botched analysis by Roach. Mish]
The lack of job creation and capital accumulation are understandable in the current environment. In a saving-short US economy — where the net national saving rate has plunged to a record low of 1.5% since early 2002 -- this lack of investment hardly comes out of thin air. Moreover, during periods of intense competition, businesses have great incentive to make do with less and reap the efficiency dividend of a powerful productivity windfall. But these trends may have also gone too far. If all there is to the productivity resurgence is slash-and-burn cost cutting, the endgame may be the “hollowing out” of Corporate America — in essence, reducing the scale and scope of production platforms so far that US companies end up capitulating market share in an ever-expanding global economy. A failure to control such trends is tantamount to corporate suicide. The recent pick-up in the pace of job creation, along with the related rise in unit labor costs — the 2.3% annualized increase in 4Q04 marking the sharpest increase of this recovery — suggests that the winds of cost-cutting may now be shifting direction.
[Just where is the recent pickup in the pace of job creation that Roach alluded to? Where? We are still not producing jobs gains faster than population growth. Evidence in facts suggests job growth is now stagnating rather than growing, 3 years into a "rcovery" Mish]
To the extent that businesses continue to ease up on the excesses of cost-cutting, that would be yet another reason to expect the productivity dividend to wane.
Of course, there’s always the possibility that we may have mis-measured the productivity bonanza from the start. I have long been sympathetic to that possibility, having arguing that government statisticians do an especially poor job in capturing the lengthening out of work days that is occurring in this IT-enabled era.
[This is assuredly true. The productivity miracle is in fact over-stated because of hedonics, jobs created elsewhere in the world, and US workers with jobs working far mare than the 40 hours they are being paid for. Mish]
To the extent that portable information appliances (i.e., cell phones, laptops, and Blackberries) and increasingly ubiquitous connectivity enable information workers to toil on 24x7 schedules, the official data on hours worked vastly understate labor input. My favorite example of this bias is financial services, where the hourly workweek is currently estimated at 35.8 hours — virtually unchanged since 1981. This is ludicrous — IT-enabled knowledge workers in financial services are now stretching out their workdays as never before. Productivity enhancement is not about working longer. Instead, it is all about getting more value added out of a given unit of work time. The official labor market data don’t come close to capturing this important phenomenon.
The US Bureau of labor Statistics is sensitive to this critique and recently released its first “time-use survey” that provides a detailed tally of how a typical day is divided into a wide range of activities from sleeping and eating to leisure and work (see “BLS Time-Use Survey,” September 14, 2004). This bears critically on the issue of productivity measurement. While this first survey only provides a snapshot of time use for one year — 2003 — it enables us to crosscheck other official estimates of work schedules. The results confirm my long-standing suspicions that the government is understating work schedules and, by inference, over-stating productivity. The new survey reveals that the average workweek for full-time workers was 40.5 hours in 2003 — fully 20% longer than the official estimate of 33.7 hours (establishment survey) that is embedded in the productivity calculus. That hints at a third potential reason to expect productivity growth to fade: To the extent that the recent acceleration in productivity growth has been a reflection of increased work time, there are physical (i.e., 24 hours in a day) and physiological limits (i.e., sleep) as to how far that trend can go as well. The time-use survey says that the American worker may be closer to those limits than we had thought — underscoring the potential of diminished returns of productivity growth through sheer exertion.
[This argument I can accept, but... it can also be reduced by doing fewer IT related projects, ect in the light of reduced IT expenditures that he expects. It can also be accomplished with more temps and more outsourcing of work where possible to India and China. Those outsourced workers will not show up on US payrolls. Mish]
There can be no mistaking the financial-market implications of the great productivity fade — if, in fact, that’s what now lies ahead. Inflation risks would tip to the upside, prompting the Federal Reserve to withdraw the “productivity subsidy” to short-term interest rates. Long rates would undoubtedly follow suit — probably even more so that those at the short end.
[Roach goes from double dip to inflation Hawk. This has been an almost unbelievable transition, just as the economy is weakening worldwide, US housing is slowing, and the FED is tightening. Mish]
In a weaker productivity climate, corporate earnings and the stock market would also come under pressure. And with productivity growth the key driver of trend growth in GDP and tax revenues, the budget deficit outlook would take yet another turn for the worse. Finally, to the extent that a deceleration in US productivity growth is accompanied by a long overdue acceleration in European productivity growth — a possibility that Eric Chaney embraces (see his 26 January dispatch, “Swing, Little Productivity Pendulum, Swing”) — the euro could strengthen and the dollar could come under further downward pressure.
[So what? The Euro is down 7 cents recently so nothing will happen if it rises 7 cents again. IF, BIG IF, Europe does manage to "increase productivity", it will come at a time the world is awash in goods and only add to the oversupply of stuff. It will also increase unemployment in Europe as fewer workers will be needed to produce said stuff. Once again Roach completely and totally misses what will happen should Europe finally get its act together. Mish]
America’s fabled productivity resurgence has been at the heart of its economic miracle over the past decade. The recent cyclical slowing of productivity growth doesn’t bring this story to an end. But it does raise important questions as to the trend that will emerge on the other side of this predictable slowdown. The market consensus firmly believes America’s productivity miracle is a permanent feature of the macro landscape. To the extent that the mega forces behind the productivity resurgence — technology transition, cost cutting, and extended work schedules — end up being more fleeting than permanent, the biggest surprise on the productivity front may be yet to come.
[At this point I have no idea what the market believes because jobs data, productivity data, and GDP data, and inflation data have been so grossly distorted by hedonics and other factors that there is quite simply no way that anyone knows what is actually happening with any degree of certainty on aabout anything. Mish]
morganstanley.com |
