A Blog by Jonathan Low


May 22, 2017

Is US Productivity (Finally) Picking Up?

It is conceivable, as the following article suggests, that 'the productivity puzzle,' is not so puzzling after all, but a reflection of risk aversion and misallocation of capital investment driven by financialization of the economy. JL

Gavyn Davies reports in the Financial Times:

Structural forces have reduced productivity growth since the 1960s. These include slower rates of technological advance, ageing of the population and slower advances in education. Strained corporate balance sheets and a dysfunctional banking sector have led to slower implementation of technical advances and misallocation of capital. Risk aversion has added to the decline in investment. Productivity growth derived from “capital services per hour worked” has begun to rise, implying that investment may be improving, as it normally does late in the economic cycle.
The dire performance of labour productivity growth in the global economic recovery – the so-called “productivity puzzle” – is by far the most depressing feature of recent economic performance in the big economies.Growth in total-factor productivity in the advanced economies (i.e. the efficiency of labour and the capital stock combined) has fallen to zero in recent years, compared with a little below 1.0 per cent a year in the decades immediately before the Great Financial Crash. Growth in labour productivity (i.e. output an hour worked) has fallen even more. This collapse in TFP and labour productivity has been by far the main cause of the disappointing growth in GDP in the advanced economies since 2008.Actually, the productivity “puzzle” is not so much of a puzzle, as a new study by IMF economists clearly demonstrates. In the long run, structural forces have been reducing productivity growth since the 1960s. These include slower rates of technological advance, the ageing of the population and slower advances in education. It is also possible that there has been some increased mismeasurement of inflation that has in turn resulted in underestimates of real output and productivity.The effects of these forces were in evidence well before the GFC, and they are still inexorably getting worse. There is no good reason to expect any improvement in these structural trends in the near future.However, there are also many reasons why productivity growth may have been temporarily affected by changes in economic behaviour since the GFC. Strained corporate sector balance sheets and a dysfunctional banking sector have weakened capital investment and led to slower implementation of technical advances and a misallocation of capital to zombie companies.Elevated policy uncertainty and risk aversion among corporate management has added to the decline in capital investment. The recent IMF paper provides evidence that many of these forces have been significant in the advanced economies since 2008.Some of these headwinds caused by the GFC should have become less powerful as the economic recovery has progressed, especially in economies such as the US and the UK, where labour markets have fully returned to normal. Until recently, however, there has been barely any sign of productivity turning a corner, and recent downgrades to global GDP projections have been driven entirely by renewed disappoints on the productivity front.With US labour utilisation having little further room to grow, productivity has now, for the first time since 2008, become the main constraint on the GDP growth rate. US Treasury secretary Steven Mnuchin has set a target for “sustainable” GDP growth at 3 per cent a year, but this certainly cannot be attained without a big recovery in productivity growth. Fortunately, there are finally a few early signs that productivity might be starting to recover in the US, which is the economy where we would expect to see the headwinds disappear first.Trend growth in global GDP stabilisingThe dynamic factor models used by Fulcrum to estimate “nowcasts” for the big economies also produce real-time estimates of long-term underlying growth rates in real GDP. The latest results for the main advanced economies are as follows:The decline in the long-run trend in GDP growth, driven by lower productivity growth, is clearly visible. The good news is that the advanced economies’ trend seems to have stopped falling in 2012, and recently there have been some very tentative indications that the trend may be starting to rise slightly. In some countries, notably the eurozone, the improvement in trend GDP growth has come mainly from a cyclical recovery from the recession during the euro crisis in 2011-12. This improvement is demand rather than supply driven, and has had little to do with any recovery in productivity.US trend productivity growth improving slightlyIn the US, however, there have been some signs that productivity growth may be starting to recover from the low points reached a few years ago. Here is the latest estimate of total factor productivity in the business sector, taken from official data:Jan Hatzius at Goldman Sachs has derived an alternative estimate of labour productivity growth, using official data and the ISM purchasing managers’ index. The official data are fairly volatile over short periods, but the series derived from the ISM survey shows a smoother picture, with a clear improvement since the beginning of 2016:Finally, it is instructive to look at the latest estimates of trend productivity taken from the nowcasting model estimated by Juan Antolin-Diaz et al [1], which is due for publication in the Review of Economics and Statistics next month. This is our preferred up-to-date estimate for underlying productivity, and it shows that trend growth has risen from a low point of 0.3 per cent in 2012 Q2 to 0.7 per cent now:Clearly, these signs of improvement in the US are very tentative [2], but they support the theory that productivity will recover now that the headwinds from the GFC have started to abate. In particular, the contribution to productivity growth derived from “capital services per hour worked” has begun to rise, implying that investment may finally be improving, as it normally does late in the economic cycle.Disputes about data measurement problemsIt also seems that the US official statisticians may now be willing to admit that they have been underestimating productivity growth because inflation is overstated, and real output is understated, in the GDP data. Until recently, economists in the Federal Reserve and other official sources have strongly rejected claims that there is any downward bias in the productivity data, and the latest IMF paper (see above) agrees with that conclusion. But economists in other countries, including the UK, have had a more open mind on this question.Jan Hatzius is one of the economists who believes that US data are understating the true rate of productivity growth by as much as 0.5-0.75 percentage points a year, a figure that is probably greater than the mismeasurement in earlier decades. If so, then the measurement problem may explain part of the slowdown in the official productivity data.Hatzius argues that the sharp increase in the quality of services provided by new products in consumer technology has not been picked up in the data, so real output is higher, and inflation lower, than shown in the government’s data. A large part of the problem stems from the introduction of entirely new products (such as the iPhone). These products may greatly enhance the real value of services to consumers in ways that are not included in consumer price data [3]. Hatzius believes that the US Bureau of Labor Statistics is now more willing to allow for these effects in calculating the CPI, especially in technology and health products.Mnuchin’s ambitious growth targetWhat is the conclusion from all this? There has probably been some recovery in US productivity growth in the recent past, driven by a reduction in the temporary headwinds that appeared after the GFC. Further gains are probable as the headwinds fade further into the distance.But so far this is a tentative conclusion, and the improvement has not taken the productivity trend back to the levels seen just before the GFC, still less to the much higher rates seen in the golden years in the 1960s. Mismeasurement may explain part of this underperformance, but economists are sharply divided on that point.Mr Mnuchin’s 3 per cent GDP growth target requires labour productivity growth to rebound to 2.25 per cent a year, double or triple the current trend. Given that the supply side benefits of reducing marginal tax rates are shown in many economic studies to be fairly small, it is not clear how he expects to achieve that.———————————————————————————————-Footnote:[1] Antolin-Diaz, J., Drechsel, T., and Petrella, I., (2017) “Tracking the Slowdown in Long-Run GDP Growth”, The Review of Economics and Statistics 99:2, May.[2] The drop in GDP growth in 2017 Q1, officially reported last Friday, may result in a temporary dent in the productivity growth rate. However, the GDP data have clearly been distorted downwards by residual seasonality that is not identified in the first estimates of GDP. In reality, output and productivity growth in Q1 are almost certainly much higher than shown in the offcial data.[3] The key research paper on the effects of new products on the CPI estimates that productivity growth may be understated by 0.5-1.25 per cent a year from this source alone, up from 0.4-0.9 per cent a year before the mid-1990s.


Post a Comment