Tuesday 16 April 2013

Robert Gordon on Productivity


Many consider economic growth a given, practically a law of nature.  It isn't.  In fact, before 1750, there was effectively no growth, and scant improvement in living standards.  In a paper published in 2012, Robert Gordon, an economist at Northwestern University, wonders whether the period from 1750 to the present was merely a one-time episode in human history, thus dooming us to return to the low-growth past.  He suggests that annual growth - for most of the population, at least - could fall to just 0.5% in the coming decades, down sharply from the 1.9% that we've enjoyed for more than a century.  (He is specifically focusing on the US economy, which may or may not reflect subdued growth in the rest of the world).

Productivity growth - "productivity" broadly measures the ratio of output per unit of input, in this case GDP per man hour worked - is what leads to improved standards of living, and Gordon breaks the post-1750 period into three Industrial Revolutions.  The first, lasting from 1750 to 1830, included inventions such as the steam engine, railroads and cotton spinning, with the economic effects lasting at least until 1900.  The second and most productive, spanning three astonishingly fruitful decades from 1870 to 1900, saw the arrival of electricity, petroleum, chemicals, the internal combustion engine, indoor plumbing, the telephone, phonograph and motion pictures.  The third IR, beginning in 1960 and running to the present, includes all of the wonders of the information age, such as computers, the internet and mobile devices.

While the inventions and innovations of the second IR arrived in the short span of just thirty years, the follow-on inventions (air travel, reliable indoor heating and air conditioning, sewer systems, electrical and indoor appliances, highways) and the long period of time it took to bring all of these enhancements across America, meant that the period of blazing growth lasted from about 1890 to 1972.  Since then, alas, growth has slowed considerably.  While the electronic age has brought amazing, gee-whiz additions to daily life, its effect on growth have so far have paled in comparison to the earlier period.  Allowing users the ease to consume anytime, anywhere is appreciated, but it doesn't bring many new labor-saving benefits.  Interestingly, the largest increase in productivity brought about by computers occurred before the arrival of the internet, when a number of clerical and administrative tasks were automated, thanks to calculators, bar-code scanners, word processors, ATMs, and other electronic marvels.

Gordon points out that the major reason that IR #2 will be difficult to one-up is the one-time nature of many of the improvements.  There's a stark difference between travelling by horse or ship one decade, and by automobile and airplane soon afterward.  However, there's no practical difference between flying at 500 miles an hour or 700.  

Moreover, it's easy to overlook the vast improvements that occurred as indirect consequences of changes in transportation, notably the gains to public health that came from replacing disease-carrying manure from horses by cars and roads, and from cleaning up the smoky, poisoned interiors that prevailed before electricity.  Similarly, once heating and cooling achieves a constant temperature in most houses, there's nowhere to go but sideways.  Other one-time changes include the mass migration from rural to urban centers, the elimination of most "brute force" manual labor, and the filling of homes with work-saving modern appliances. 

In addition, there have been a number of non-repeatable changes since the dawn of IR#3 that are already behind us, such as moving physical card catalogues onto internet-accessible databases, and the ability to read material on-screen rather than on dead trees.  Gordon speculates that the fall in growth from IR#2 has more than offset any gains that have been made since.

There are potential problems with Gordon's work.  As he freely admits, during the glorious bonanza of IR#2, there was a long period of slow growth from 1906 to 1928, a period that coincided with the accelerated roll-out of electricity, automobiles, roads and plumbing.  While this may reflect difficulties with data collection and measurement, it could also suggest a certain lumpiness in the process of improving the economy, and the current slower growth period could be a trough on the way to a new peak.

And, as Yogi Berra wisely noted, predictions are difficult - especially about the future.  A futurist in 1945 who didn't foretell the invention of the transistor a few years later also could not have predicted the rise of the electronic age.  Now, as then, it's nearly impossible to know what innovations the future holds.  Paul Krugman, for one, is more optimistic about the potential of self-driving cars and speech recognition software to significantly raise productivity growth.  However, Gordon is not closed-minded about these possibilities and points to several notable examples of predictions that turned out to be laughably pessimistic.  On the other hand, there are counter-examples of unfulfilled hope.  Who knows?

Even if Gordon's downbeat forecasts are correct, they must be kept in perspective: real GDP per capita in the US would still double by 2100.  Considering how extraordinarily well the average American lives, a 100% increase from the current standard of living, even if it takes nearly a century to attain, is nothing to sneeze at, and would be far higher than almost all of human history.

Disclaimer: The host of this blog shall not be held responsible or liable for, and indeed expressly disclaims any responsibility or liability for any losses, financial or otherwise, or damages of any nature whatsoever, that may result from or relate to the use of this blog. This disclaimer applies to all material that is posted or published anywhere on this blog.

No comments:

Post a Comment