Paul Krugman has recently pointed out a very pessimistic, but very provocative paper by Robert Gordon, about the possibilities of long run growth. Gordon suggests that the “rapid progress made over the past 250 years could well turn out to be a unique episode in human history.” In his view, long-term stagnation is a very possible outcome. He asserts that the reasons for this are the effects of technical progress on investment.
Gordon argues that, while the first (steam, cotton textiles, railroad) and especially the second (automobile, chemicals, electricity, oil) Industrial Revolutions (IR) led to a significant increase in investment, the third IR (information technology) has been less prone to lead to significant increases in investment. Further, the advantages of the first and second IRs were enhanced by demographic changes and the process of urbanization, which created the need for investment in infrastructure.
For Gordon the problems of the third IR can be summarized as follows:
“The computer and Internet revolution (IR #3) began around 1960 and reached its climax in the dot.com era of the late 1990s, but its main impact on productivity has withered away in the past eight years. Many of the inventions that replaced tedious and repetitive clerical labor by computers happened a long time ago, in the 1970s and 1980s. Invention since 2000 has centered on entertainment and communication devices that are smaller, smarter, and more capable, but do not fundamentally change labor productivity or the standard of living in the way that electric light, motor cars, or indoor plumbing changed it.”
While it is true that the effects of the third IR have been disappointing to say the least, and Solow famously put it “one can see the computer age everywhere but in the productivity statistics,” Gordon’s pessimism is suspect because it stems from his supply side interpretation of the process of growth and technical change. This is a problem I have discussed in a previous post. In the supply side view, technological innovation, and the investment associated with it drives the process of economic growth. Demand adapts, since the increasing abundance of the newly introduced innovations reduces their price and leads to widespread utilization.
There are several problems with the conventional supply side story of growth and technical change. The most important is that it is not clear at all why somebody would introduce new goods, new methods of production, new sources of supply, and new forms of organization in order to produce more and more efficiently if there is no demand for any those things. Adam Smith long ago noted that the division of labor, the higher productivity that explains the wealth of nations, resulted from the extension of the market. So output growth and productivity growth depended on the size of demand.
This is why weaker productivity growth since the 1970s has followed the rise of conservatism, the trickledown and austerity driven macroeconomic policies that put an emphasis on low inflation, and the reduction of taxation on wealthy “job creators” as an incentive investment. Note that productivity actually picked up in the 1990s, during the so-called New Economy, as a result of the higher growth rates during the Clinton boom. Of course the boom was based on the dot.com expansion, and once the bubble burst productivity growth fizzled. And even with the housing bubble the Bush boom was always lackluster and productivity growth, as noted by Gordon, never recovered.
Further fiscal and monetary expansion would not only lead to higher output and employment, as I noted here, but also would lead to higher productivity growth (this is often called the Kaldor-Verdoorn Law). The main constraint to further macroeconomic expansion has been political; the reasons for the current slow recovery and for the slowdown in productivity over the almost four decades has been conservatism—not something inherent in the third IR. In fact, the single most important obstacle to the expansion of demand has been the stagnation of real wages and median income, which has led to debt-driven demand expansions. In the absence of wage increases families were forced to become increasingly indebted to maintain their levels of consumption. Inequality, which has been the main cause of the increasing instability of the economic system, is also at the heart of the lower rates of productivity growth.
Hence, not only do we need more fiscal and monetary expansion (even if I’m less sanguine than DeLong and Krugman about the possibilities of monetary policy), but also the strengthening of institutions that allow for a more sustainable expansion of demand. We need stronger unions, higher taxes on the wealthy, and more and stronger regulation of financial markets. Sure enough it is possible, even likely, that the United States will maintain lower rates of growth in the future as suggested by Robert Gordon, but this is not the inevitable result of a supply side, technology-driven fate, but a political decision that can, and hopefully will, be changed by active social forces.
Triple Crisis Welcomes Your Comments. Please Share Your Thoughts Below.
Great comments to Gordon’s article. I can see how politics didn’t play a large enough role in his “exercise in subtraction” but should have.
Thanks Peter. Note that politics can only play a role if growth is not purely a supply side, exogenous change in technological knowledge, and if demand (and the problems of expanding it or not, with the possibilities of winners and losers) becomes the force behind economic it.
Robert Gordon’s paper is historically informed, but conceptually incredibly impoverished. If this is representative of the state of economic thought these days (and Krugman’s citation suggests that it is), the economics is even worse off than Krugman’s regular complaints suggest.
I’m not an economist, or academic of any kind, so I’m not going to write a rebuttal paper, but I’ll state a few points.
– Any paper purporting to be about the economics of the IT revolution that gives only one sentence to Moore’s Law can be totally ignored straightaway. Moore’s Law has made 100% value destruction of computer hardware every 4 years completely reliable for the 50 years.
– Likewise for software complexity. Gordon quotes “what Intel giveth, Microsoft taketh away” without any evidence of understanding its meaning. Many of the most important global problems that are well-defined have exponential scalability limits that not even quantum computation will be able to overcome. Cosma Shalizi at the Univ. of Michigan has shown that Stalinist planned economies fall into this class.
– When you chain an exponential growth function to an exponential slowdown function, you get chaotic sensitivity to initial conditions. More sensitive than any amount of data you can ever conceive of collecting. Economics can never be more than a pseudo-science until it is able to face this uncertainty.
– When a biologist looks at Gordon’s figs.2 and 3 showing growth in GDP per capita from 1300-2100, the resemblance to a logistic growth function with bounded resources is instantly clear. Why is this not obvious to economists? Not seeing the forest for the trees, maybe?
– The third IR (or maybe the fourth?) has changed the nature of the “product” that GDP represents, yet the validity of GDP reports remains unquestioned. When I can sit in my living room in Texas and look out an Earthcam video window on the Brooklyn Bridge or a beach on any of 3 continents, or view traffic jams on the Northern Motorway in Aukland, New Zealand, my quality of life is improved in ways that are incommensurate with the loss of GDP that could be computed by my not spending money that I don’t have (cf. “effective demand”) to those locations.
– Finally, the emergence of financial derivatives that are made possible by computerized, algorithmic asset partitioning create vast amounts of economic activity that produces nothing of value except the illusion of safety, and a few yachts and mansions that are paid for with a subtraction in the value of the dollar as the Fed pours yet another round of quantitative easing into the money supply.
I’m not sure which of these is the most important. I would probably put the externalization of the informational component of lifestyle quality, and the inability of money to distinguish between the value of experience and the value of possessions at the top.