With new gizmos, drugs and financial instruments appearing all the time, it certainly seems like technical and scientific innovation continues fast and furious from the 20th century.

However, in the keynote address I posted, Collapse of Complex Societies, anthropologist Dr Joseph Tainter was doubtful about the effectiveness of our current pace of innovation to keep solving problems. As if everyone should know this, he offhandedly mentioned studies indicating that the pace of innovation had declined mid-20th century. He expanded upon that idea in this interview:

KV: What about technological innovation? The spread of digital technology, the Internet, and mobile technology contributed to the economic recovery during the last fifteen years. There has been a bit of talk about innovating our way out of this recession too... Is such optimism in technological solutions warranted? ...

JT: Short answer: It’s complicated. Long answer: Technological-innovation-as-savior is part of our cosmology. It is a fundamental part of our beliefs, so frequently we don’t think about it rationally. Relying on technological innovation to find some solution is what I call a faith-based approach to the future. There are two things about technological innovation that concern me. The first is that, like other endeavors, research grows complex and costly and can reach diminishing returns. This is covered in the Collapse book so I won’t elaborate here. The second problem is what is known as the Jevons Paradox. William Stanley Jevons, a 19th century British economist, pointed out that in the long run technological innovations aimed as at using less of a resource actually lead to even more of the resource being used. His example was coal, but the principle applies across the board. As technological innovation leads to economy in using a resource, people respond to the lower cost by using even more. I conclude from this that technological innovations can offer only short-term advantages. They quickly become outdated, then the next round of innovations may be harder to achieve.

I don't know what studies Tainter meant, but there appears to be a history of controversial theories of long economic cycles, such as 50 to 60 year Kondratiev waves, which are driven by, "wealth, capital accumulation and innovation," but eventually stagnate. Kondratiev was executed by Stalin for his troubles, but his ideas were championed by the Austrian economist Joseph Schumpeter, who proposed that 'creative destruction' was necessary to get out of the troughs. Schumpeter's American student Robert M Solow, won a Nobel Prize in Economics for his model of economic growth, which is still used, though modified. Solow wrote the forward for the second of the papers below by economist Robert J. Gordon, which concerns a long wave in the American economy. Gordon is the author of Macroeconomics (a textbook that seems to baffle students), a fellow at the National Bureau of Economic Research and was a member of the Boskin commission. The first paper supports Tainter's assertion that innovation is less pervasive now. I've added paragraph breaks to the abstracts:

Does the "New Economy" Measure up to the Great Inventions of the Past?
Journal of Economic Perspectives, 2000, v14(4,Fall), 49-74.

During the four years 1995-99 U. S. productivity growth experienced a strong revival and achieved growth rates exceeding that of the 'golden age' of 1913-72. Accordingly many observers have declared the 'New Economy' (the Internet and the accompanying acceleration of technical change in computers and telecommunications) to be an Industrial Revolution equal in importance, or even more important, than the Second Industrial Revolution of 1860-1900 which gave us electricity, motor and air transport, motion pictures, radio, indoor plumbing, and made the golden age of productivity growth possible.

This paper raises doubts about the validity of this comparison with the Great Inventions of the past. It dissects the recent productivity revival and separates the revival of 1.35 percentage points (comparing 1995-99 with 1972-95) into 0.54 of an unsustainable cyclical effect and 0.81 points of acceleration in trend growth. The entire trend acceleration is attributed to faster multi-factor productivity (MFP) growth in the durable manufacturing sector, consisting of computers, peripherals, telecommunications, and other types of durables. There is no revival of productivity growth in the 88 percent of the private economy lying outside of durables; in fact when the contribution of massive investment in computers in the nondurable economy is subtracted, MFP growth outside of durables has actually decelerated.

The paper combines the Great Inventions of 1860-1900 into five clusters and shows how their development and diffusion in the first half of the 20th century created a fundamental transformation in the American standard of living from the bad old days of the late 19th century. In comparison, computers and the Internet fall short. The rapid decline in the cost of computer power means that the marginal utility of computer characteristics like speed and memory has fallen rapidly as well, implying that the greatest contributions of computers lie in the past, not in the future.

The Internet fails the hurdle test as a Great Invention on several counts. First, the invention of the Internet has not boosted the growth in the demand for computers; all of that growth can be interpreted simply as the same unit-elastic response to the decline in computer prices as was prevalent prior to 1995. Second, the Internet provides information and entertainment more cheaply and conveniently than before, but much of its use involves substitution of existing activities from one medium to another. Third, much internet investment involves defense of market share by existing companies like Borders Books faced with the rise of Amazon; social returns are less than private returns. Fourth, much Internet activity duplicates existing activity like mail order catalogues, but the latter have not faded away; the usage of paper is rising, not falling. Finally, much Internet activity, like daytime e-trading, involves an increase in the fraction of work time involving consumption on the job.

In these three papers, the term MFP comes from the Solow growth model, in contrast with LP which is labor productivity.

Multifactor productivity (MFP) is output divided by a weighted average of capital and labor inputs. The weights used are usually based on the aggregate input shares either factor earns. This ratio is often quoted as: 33% return to capital and 66% return to labor (in Western nations), but Robert J. Gordon says the weight to labor is more commonly assumed to be 75%.

Interpreting the "One Big Wave" in U.S. Long-Term Productivity Growth

Productivity growth, inflation, and unemployment: The collected essays of Robert J. Gordon, With a foreword by Robert M. Solow. Cambridge; New York and Melbourne: Cambridge University Press, 2004.

This paper assesses the standard data on output, labor input, and capital input, which imply one big wave' in multi-factor productivity (MFP) growth for the United States since 1870. The wave-like pattern starts with slow MFP growth in the late 19th century, then an acceleration peaking in 1928-50, and then a deceleration to a slow rate after 1972 that returns to the poor performance of 1870-1891. A counterpart of the standard data is a mysterious doubling in the ratio of output to capital input when the postwar era is compared with 1870-1929.

Three types of measurement adjustments are applied to the standard input data. Following the lead of Denison and Jorgenson-Griliches, adjustments for the changing composition (or quality') of labor and capital, currently published by the BLS back to 1948, are estimated for 1870-1948. These composition adjustments take into account the shifting mix of the labor force along the dimensions of education and age-sex composition, and of the capital stock between equipment and structures. Further adjustments are made to capital input data to allow retirement to vary with gross investment rather than to follow a fixed pattern depending only on age, and to add types of capital owned by the government that are particularly productive in the private sector. A new MFP series taking account of all these adjustments grows more slowly throughout, and the big wave' phenomenon is both flatter and extends back further in time to 1891.

However, there is no solution to the post-1972 productivity slowdown, and in the new data MFP growth during 1972-96 proceeds at a pathetic 0.1 percent per year. A byproduct of the measurement adjustments is to solve completely the previous puzzle of the jump in the output-capital ratio; in the new data this ratio is actually lower in 1996 than in 1870. The primary substantive explanation for the big wave lies in the timing of inventions. MFP growth during the big wave' period benefited from the diffusion of four great clusters of inventions that dwarf today's information technology revolution in their combined importance.

A complementary hypothesis is that the partial closing of American labor markets to immigration and of American goods markets to imports during the big wave period gave an artificial and temporary boost to real wages which fed back into boosting productivity growth, followed by a reopening that contributed to the post-1972 productivity slowdown.

Revisiting U. S. Productivity Growth over the Past Century with a View of the Future

2010

The statistical trend for growth in total economy LP ranged from 2.75 percent in early 1962 down to 1.25 percent in late 1979 and recovered to 2.45 percent in 2002. Our results on productivity trends identify a problem in the interpretation of the 2008-09 recession and conclude that at present statistical trends cannot be extended past 2007. For the longer stretch of history back to 1891, the paper provides numerous corrections to the growth of labor quality and to capital quantity and quality, leading to significant rearrangements of the growth pattern of MFP, generally lowering the unadjusted MFP growth rates during 1928-50 and raising them after 1950. Nevertheless, by far the most rapid MFP growth in U. S. history occurred in 1928-50, a phenomenon that I have previously dubbed the “one big wave.”

The paper approaches the task of forecasting 20 years into the future by extracting relevant precedents from the growth in labor productivity and in MFP over the last seven years, the last 20 years, and the last 116 years. Its conclusion is that over the next 20 years (2007-2027) growth in real potential GDP will be 2.4 percent (the same as in 2000-07), growth in total economy labor productivity will be 1.7 percent, and growth in the more familiar concept of NFPB sector labor productivity will be 2.05 percent. The implied forecast 1.50 percent growth rate of per-capita real GDP falls far short of the historical achievement of 2.17 percent between 1929 and 2007 and represents the slowest growth of the measured American standard of living over any two-decade interval recorded since the inauguration of George Washington.

This is way too much economics before lunch, but I think it explains where Tainter gets his assertions that our feeling that there is still a lot of innovation may be more a result of marketing than new invention. Econ is not my field, so I welcome constructive comments from those of you with more of an economics background.

Here's an update with another take on waves from the Economist in 1999. Things haven't turned out as they expected:

Catch the wave
The long cycles of industrial innovation are becoming shorter

By the time Schumpeter died in 1950, the third cycle of his “successive industrial revolutions” had already run its course. The fourth, powered by oil, electronics, aviation and mass production, is now rapidly winding down, if it has not gone already. All the evidence suggests that a fifth industrial revolution—based on semiconductors, fibre optics, genetics and software—is not not only well under way but even approaching maturity. This may explain why America shrugged off its lethargy in the early 1990s and started bounding ahead again, leaving behind countries too preoccupied with preserving their fourth-wave industries. If so, then Schumpeter’s long economic waves are shortening, from 50-60 years to around 30-40 years.

There is good reason why they should. It was only during the third wave, in the early part of the 20th century, that governments and companies began to search for new technologies in a systematic manner. One of the oldest, Bell Laboratories at Murray Hill in New Jersey, was founded in 1925. Rather than leave the emergence of “new-wave” technologies to chance, all the major industrial countries nowadays have armies of skilled R&D workers sifting the data in pursuit of blockbuster technologies capable of carving out wholly new markets. The tools they use—computer analysers, gene sequencers, text parsers, patent searchers, citation mappers—are getting better all the time, speeding up the process. The productivity of industrial laboratories today is twice what it was a couple of decades ago.

So the fifth industrial revolution that started in America in the late 1980s may last no more than 25-30 years. If, as seems likely, we are already a decade into this new industrial cycle, it may now be almost too late for the dilatory to catch up. The rapid-upswing part of the cycle—in which successful participants enjoy fat margins, set standards, kill off weaker rivals and establish themselves as main players—looks as though it has already run two-thirds of its course, with only another five or six years left to go. Catching the wave at this late stage will depend on governments’ willingness to free up their technical and financial resources, invest in the infrastructure required and let their fourth-wave relics go. Failing that, latecomers can expect only crumbs from the table before the party comes to an end—and a new wave of technologies begins, once again, to wash everything aside.