26/5/16: After the Crisis: Why the Slowdown in Productivity Growth?

My article for Cayman Financial Review 2Q 2016 is out, covering the structural nature of labour productivity growth decline in post-crisis economy: see here http://caymanianfinancialreview.cay.newsmemory.com/ pages 66-67 or click on images below to enlarge:

17/4/16: Start Ups, Manufacturing Jobs and Structural Changes in the U.S. Economy

In the forthcoming issue of the Cayman Financial Review I am focusing on the topic of the declining labour productivity in the advanced economies - a worrying trend that has been established since just prior to the onset of the Global Financial Crisis. Another trend, not highlighted by me previously in any detail, but related to the productivity slowdown is the ongoing secular relocation of employment from manufacturing to services. However, the plight of this shift in the U.S. workforce has been centre stage in the U.S. Presidential debates recently (see http://fivethirtyeight.com/features/manufacturing-jobs-are-never-coming-back/).

An interesting recent paper on the topic, titled “The Role of Start-Ups in Structural Transformation” by Robert C. Dent, Fatih Karahan, Benjamin Pugsley, and Ayşegül Şahin (Federal Reserve Bank of New York Staff Reports, no. 762, January 2016) sheds some light on the ongoing employment shift.

Per authors, “The U.S. economy has been going through a striking structural transformation—the secular reallocation of employment across sectors—over the past several decades. Most notably, the employment share of manufacturing has declined substantially, matched by an increase in the share of services. Despite a large literature studying the causes and consequences of structural transformation, little is known about the dynamics of reallocation of labor from one sector to the other.”

“There are several margins through which a sector could grow and shrink relative to the rest of the economy”:

1. “…Differences in growth and survival rates of firms across sectors could cause sectoral reallocation of employment”
2. “…differences in sectors' firm age distribution could affect reallocation since firm age is an important determinant of growth or survival behavior” 
3. “…the allocation of employment at the entry stage which we refer to as the entry margin could contribute to the gradual shift of employment from one sector to the other.”
4. “…because the speed at which differences in entry patterns are reflected in employment shares depends on the aggregate entry rate, changes in the latter could affect the extent of structural transformation.”

Factors (1) and (2) above are referenced as “life cycle margins”.

The study “dynamically decomposed the joint evolution of employment across firm age and sector”, focusing on three sectors: manufacturing, retail trade, and services.

Based on data from the Longitudinal Business Database (LBD) and Business Dynamic Statistics (BDS), the authors found that “…at least 50 percent of employment reallocation since 1987 has occurred along the entry margin.” In other words, most of changes in manufacturing jobs ratio to total jobs ratio in the U.S. economy can be accounted for by new firms creation being concentrated outside manufacturing sectors.

Furthermore, “85 percent of the decline in manufacturing employment share is predictable from the average life cycle dynamics and the early 1980s distribution of startup employment across sectors. Further changes over time in the distribution of startup employment away from manufacturing, while having a relatively small effect on manufacturing where entry is less important, explain almost one-third of the increase in the services employment share.”

Again, changed nature of entrepreneurship, as well as in the survival rate of new firms created in the services sector, act as the main determinants of the jobs re-allocation across sectors.

Interestingly, the authors found “…little role for the year-to-year variation in incumbent behavior conditional on firm age in explaining long-term sectoral reallocation.” So legacy firms have little impact on decline in manufacturing sector jobs share, which is not consistent with the commonly advanced thesis that outsourcing of American jobs abroad is the main cause of losses of manufacturing sector jobs share in the economy.

Lastly, the study found that “…a 30-year decline in overall entry (which we refer to as the \startup deficit) has a small but growing effect of dampening sectoral reallocation through the entry margin.”

These are pretty striking results.

The idea that the U.S. manufacturing (in terms of the sector importance in the economy and employment) is either in a decline or on a rebound is not as straight forward as some political debates in the U.S. suggest.

Reality is: in order to reverse or at least arrest the decades-long decline of manufacturing jobs fortunes in America, the U.S. needs to boost dramatically capex in the sector, as well as shift the sector toward greater reliance on human capital-complementary technologies. It is a process that combines automation with more design- and specialist/on-specification manufacturing-centric trends, a process that is likely to see accelerated decline in lower skills manufacturing jobs before establishing (hopefully) a rising trend for highly skilled manufacturing jobs.

24/11/15: Europe's Dead Donkey of Productivity Growth

Remember the mythology of European productivity miracles:

1. The EU is at least as competitive as the U.S. (with Lisbon Agenda completed, or rather abandoned);
2. The EU growth in productivity is structural in nature (i.e. not driven by capital acquisition alone and not subject to cost of capital effects); and
3. The EU productivity growth is driven by harmonising momentum (common markets etc) at a policy level, with the Euro, allegedly, producing strong positive effects on productivity growth.

Take a look at this chart from Robert J. Gordon's presentation at a recent conference:

The following observations are warranted:

• EU convergence toward U.S. levels of productivity pre-dates major policy harmonisation drives in Europe and pre-dates, strongly, the creation of the Euro;
• EU productivity convergence never achieved parity with the U.S.;
• EU productivity convergence was not sustained from the late 1990s peak on;
• The only period of improved productivity in the EU since the start of the new millennium was associated with assets bubble period (interest rates and credit supply).

Darn ugly!

But it gets worse. Since the crisis, the EU has implemented, allegedly and reportedly, a menu of 'structural' reforms aiming at improving competitiveness.  Which means that at least since the end of the crisis, we should be seeing improved productivity growth differentials between Europe and the U.S. And the EU case for productivity growth resumption is supported by the massive, deeper than the U.S., jobs destruction during the crisis that took out a large cohort of, supposedly, less productive workers, thereby improving the remainder of the workforce levels of productivity.

Here is a chart from the work by John Van Reenen of LSE:

Apparently, none of this happened:

• EU structural reforms have been associated (to-date) with much lower productivity growth post-crisis than the U.S. and Japan;
• EU jobs destruction during the crisis has been associated with lower productivity increases than in the U.S. and Japan;
• All EU programmes to support growth in productivity, ranging from the R&D supports to investment funding for productivity-linked structural projects have produced... err... the worst outcome for productivity growth compared to the U.S. and Japan.

And the end result?

I know, I know... a Genuine Productivity Union, anyone?...

15/6/15: Euro Area Labour Productivity: It's Low and Lagging

Euro area's problem in one chart? Might sound like a bit of an over-simplification, but here is a summary of labour productivity index simply constructed as real GDP per employee:

The chart shows several facts:

1. Euro area labour productivity is currently low, despite massive uplift in unemployment (which should have increased output per employee more substantially).
2. Euro area labour productivity has grown faster than that in the U.S. in the period of 1986-1995, but has been growing at a slower rate for some twenty years now.
3. Post-2010, euro area productivity has been lagging all groups of advanced economies.

Now, remember, no one talks as much about carrying out labour markets reforms as euro area leadership. In a way, this might be warranted, given poor performance, but in a way it also might suggest that the reforms are not working. After all, since the start of the Great Recession, allegedly, we had plenty of these reforms, and we had a 'productivity-enhancing' rise in unemployment, reduction in labour force and wages moderations galore. And productivity is not really expanding much. Secular stagnation, anyone?

24/6/2014: US Productivity Slowdown: It's Structural & Nasty

"Productivity and Potential Output Before, During, and After the Great Recession" a new paper by John Fernald (NBER Working Paper No. 20248, June 2014) looks at the U.S. labor and total-factor productivity growth slowdown prior to the Great Recession in the context of the slowdown "located in industries that produce information technology (IT) or that use IT intensively, consistent with a return to normal productivity growth after nearly a decade of exceptional IT-fueled gains". In a sense, the paper reinforces the point of view that I postulated in my TEDx talk last year dealing with the 'end' of the Age of Tech (here: http://trueeconomics.blogspot.ie/2013/11/14112013-human-capital-age-of-change.html).

Fernald opens the paper with a set of two quotes. One brilliantly describes the core question we face:
"When we look back at the 1990s, from the perspective of say 2010,…[w]e may conceivably conclude…that, at the turn of the millennium, the American economy was experiencing a once-in-a-century acceleration of innovation….Alternatively, that 2010 retrospective might well conclude that a good deal of what we are currently experiencing was just one of the many euphoric speculative bubbles that have dotted human history." Federal Reserve Chairman Alan Greenspan (2000)

Fernald argues that "The past two decades have seen the rise and fall of exceptional U.S. productivity growth. This paper argues that labor and total-factor-productivity (TFP) growth slowed prior to the Great Recession. It marked a retreat from the exceptional, but temporary, information-technology (IT)-fueled pace from the mid-1990s to the early 2000s. This retreat implies slower output growth going forward as well as a narrower output gap than recently estimated by the Congressional Budget Office (CBO, 2014a)."

Figure 1 from the paper illustrates how the mid-1990s surge in productivity growth indeed ended prior to the Great Recession. The rise in labor-productivity growth, shown by the height of the bars, came after several decades of slower growth. But, notes Fernald, "in the decade ending in 2013:Q4, growth has returned close to its 1973-95 pace. The figure shows that the slower pace of growth in both labor productivity and TFP was similar in the four years prior to the onset of the Great Recession as in the six years since."

And things have been bad since. Labour productivity growth (slope of liner trend below) is now on par with what we have been witnessing in 1973-1995, and shallower than in 1995-2003. But the trend is still close to actual performance, which signals little potential for any appreciable acceleration:

Beyond labour productivity, things are even messier. Charts below plot the Great Recession against other recessions in terms of productivity, output and labour utilisation:

Notes: For each plot, quarter 0 is the NBER business-cycle peak which, for the Great Recession,
corresponds to 2007:Q4. The shaded regions show the range of previous recessions since 1953. Local
means are removed from all growth rates prior to cumulating, using a biweight kernel with bandwidth of 48 quarters. Source is Fernald (2014).

All of the above show the cyclical disaster that is the current Great Recession, but crucially, they show poor recent performance in Labour Productivity, exceptionally poor performance in Hours of Labour used, disastrous performance in Total Factor Productivity… in other words - historically problematic trends relating to productivity, labour utilisation and tech-related productivity in the current recession compared to all previous recessions.

But more worrying is that, as Fernald notes: "That the slowdown predated the Great Recession rules out causal stories from the recession itself. …The evidence here complements Kahn and Rich’s (2013) finding in a regime-switching model that, by early 2005—i.e., well before the Great Recession—the probability reached nearly unity that the economy was in a low-growth regime."

So what's behind all of this slowing productivity growth? "A natural hypothesis is that the slowdown was the flip side of the mid-1990s speedup. Considerable evidence… links the TFP speedup to the exceptional contribution of IT—computers, communications equipment, software, and the Internet. IT has had a broad-based and pervasive effect through its role as a general purpose technology (GPT) that fosters complementary innovations, such as business reorganization. Industry TFP data provide evidence in favor of the IT hypothesis versus alternatives. Notably, the euphoric, “bubble” sectors of housing, finance, and natural resources do not explain the slowdown. Rather, the slowdown is in the remaining ¾ of the economy, and is concentrated in industries that produce IT or that use IT intensively. IT users saw a sizeable bulge in TFP growth in the early 2000s, even as IT spending itself slowed. That pattern is consistent with the view that benefiting from IT takes substantial intangible organizational investments that, with a lag, raise measured productivity. By the mid-2000s, the low-hanging fruit of IT had been plucked."

This a hugely far-reaching paper with two related implied conclusions:

1. Prepare for structurally slower growth period in the US (and global) economy as the last catalyst for growth - tech - appears to have been exhausted; and
2. The Age of Tech is now in the part of the cycle where returns to innovation and technology are falling, while returns to financial assets overlaying tech sector are still going strong. The classic bubble scenario is being formed once again, as always on foot of disconnection between the real economic returns to the assets and asset valuations. This bubble will have to deflate.

12/2/2014: ICT, Productivity & Employment in the US Manufacturing

More recent research, to follow up on previous post (which dealt with jobless recoveries). This time around on the key issue of workers displacement by technology.

"RETURN OF THE SOLOW PARADOX? IT, PRODUCTIVITY, AND EMPLOYMENT IN U.S. MANUFACTURING" by Daron Acemoglu, David Autor, David Dorn, Gordon H. Hanson and Brendan Price (NBER Working Paper 19837, http://www.nber.org/papers/w19837 from January 2014) looks into the validity of the 'technological discontinuity' paradigm - the one that "suggests that IT-induced technological changes are rapidly raising productivity while making workers redundant."

Here's the justification of the tested thesis: "An increasingly influential “technological-discontinuity” paradigm suggests that IT-induced technological changes are rapidly raising productivity while making workers redundant. This paper explores the evidence for this view among the IT-using US manufacturing industries."

Basic argument here is that modern workplace is continuing to become more automated, transformed by the ICT capital. Two implications of this are:

"First, all sectors—but particularly IT-intensive sectors—are experiencing major increases in productivity. Thus, Solow’s paradox is long since resolved: computers are now everywhere in our productivity statistics."

"Second, IT-powered machines will increasingly replace workers, ultimately leading to a substantially smaller role for labor in the workplace of the future. Adding urgency to this argument, labor’s share of national income has fallen in numerous developed and developing countries over roughly the last three decades, a phenomenon that Karabarbounis and Neiman (forthcoming) attribute to IT-enabled declines in the relative prices of investment goods. And many scholars have pointed to the seeming “decoupling” between robust U.S. productivity growth and sclerotic or negligible growth rates of median U.S. worker compensation (Fleck, Glaser and Sprague 2011) as evidence that the “race against the machine” has already been run—and that workers have lost."


Top conclusion of the paper: "There is some limited support for more rapid productivity growth in IT-intensive industries depending on the exact measures, though not since the late 1990s."

But there are some serious nuances involved.

"We find, unexpectedly, that earlier “resolutions” of the Solow paradox may have neglected certain paradoxical features of IT-associated productivity increases, at least in U.S. manufacturing." Of these, the paper highlights two:

"First, focusing on IT-using (rather than IT-producing) industries, the evidence for faster productivity growth in more IT-intensive industries is somewhat mixed and depends on the measure of IT intensity used. There is also little evidence of faster productivity growth in IT-intensive industries after the late 1990s.

"Second and more importantly, to the extent that there is more rapid growth of labor productivity (ln(Y=L)) in IT-intensive industries, this is associated with declining output (ln Y ) and even more rapidly declining employment (lnL). If IT is indeed increasing productivity and reducing costs, at the very least it should also increase output in IT-intensive industries. As this does not appear to be the case, the current resolution of the Solow paradox does not appear to be what adherents of the technological-discontinuity view had in mind."

In other words: "Most challenging to this paradigm, and our expectations, is that output contracts in IT-intensive industries relative to the rest of manufacturing. Productivity increases, when detectable, result from the even faster declines in employment."

Goes some miles explaining the declining role of primary labour… 

11/02/2012: Labor Productivity - some cross-EU comparatives

There has been much of talk about Euro area (and EU27) competitiveness trends recently in the media. Some of the commentary I've seen references the issue of decoupling in competitiveness across EU states. In light of this, I decided to take a look at productivity trends. Using eurostat data, I was able to:

1. Take eurostat main series for per person aggregate (total) productivity index that sets 2005=100
2. Rebase the index to Q1 2000=100 and recompute entire set of EU27 countries, plus EA17 and EU27 aggregates
3. Obtain via (1) and (2) above new set of productivity indices that reflect dynamics in per person productivity since Q1 2000 through Q3 2011
4. Note: data is seasonally adjusted and I am only reporting countries where data is hours adjusted as well.

Here are the core charts (I added Ireland and EU 27 average in every chart):

So few trends are apparent:

• Ireland performs - since the beginning of the crisis extremely well in terms of productivity improvements and levels - much due to the massive destruction of its employment base (I commented on this effect a number of times before). Overall - this is remarkable performance albeit at huge cost.
• Spain has posted some significant increases as well, mostly due to destruction of employment - much more so than Ireland.
• Italy is performing poorly as does Greece. In fact, Greece is the third worst performer in the entire EU27 in terms of productivity growth since the beginning of the crisis (Q1 2008).
• Portugal improvements appear to be largely consistent with the pattern for Spain.
• Finland clearly leads the pack (after Ireland) in the group of Small Open Economies (SOEs)
• Strong trends in growth in East-Central Europe (ex Hungary and Slovenia) 

Now, let's take a look at cumulative growth in productivity since the beginning of the crisis - note: green boxes mark countries that outperform EU27 average by more than 1/2 STDEV, while red boxes mark those countries that underperform the EU27 average by more than 1/2 STDEV:

And similar analysis for cumulative growth in productivity since Q1 2000:

 So is there 'decoupling' going on in terms of labor productivity? Not really. Here's what's happening:

• Spain shows highest gains in total productivity since Q1 2008 but weak (roughly average) gains since Q1 2000
• Ireland shows second highest gains since Q1 2008 and above average (6th highest in EU27) gain since Q1 2000
• Slovakia doing spectacularly well, albeit, of course, from low levels, as is Estonia (though not too great during the crisis period)
• During the crisis, Belgium, UK, Greece, Hungary, Italy, lux & Sweden all posted below average (more than 1/2 STDEVs) performance
• Since Q1 2000, Italy and Lux were the only two statistical underperformers.

So unless we go beyond Q1 2000 (the period for which we don't really have coherent comparable data) there is no 'decoupling' going on in labor productivity. There is shallow growth in it on average, but no dramatic 'decoupling'. In other words, much of core Europe is pretty poor in terms of labor productivity growth, while East-Central Europe and Ireland are performing pretty well.