03 January 2012

A Critique of Stiglitz on Innovation and the Economy

Writing in the current issue of Vanity Fair, Nobel laureate Joseph Stiglitz has proposed a theory to explain the current economic malaise:
For the past several years, Bruce Greenwald and I have been engaged in research on an alternative theory of the Depression—and an alternative analysis of what is ailing the economy today. This explanation sees the financial crisis of the 1930s as a consequence not so much of a financial implosion but of the economy’s underlying weakness. The breakdown of the banking system didn’t culminate until 1933, long after the Depression began and long after unemployment had started to soar. By 1931 unemployment was already around 16 percent, and it reached 23 percent in 1932. Shantytown “Hoovervilles” were springing up everywhere. The underlying cause was a structural change in the real economy: the widespread decline in agricultural prices and incomes, caused by what is ordinarily a “good thing”—greater productivity.
He proposes that the Great Depression of the 1930s was a consequence of rapidly accelerating agricultural productivity, which led to a massive transformation in the national economy:
At the beginning of the Depression, more than a fifth of all Americans worked on farms. Between 1929 and 1932, these people saw their incomes cut by somewhere between one-third and two-thirds, compounding problems that farmers had faced for years. Agriculture had been a victim of its own success. In 1900, it took a large portion of the U.S. population to produce enough food for the country as a whole. Then came a revolution in agriculture that would gain pace throughout the century—better seeds, better fertilizer, better farming practices, along with widespread mechanization. . . What this transition meant, however, is that jobs and livelihoods on the farm were being destroyed. Because of accelerating productivity, output was increasing faster than demand, and prices fell sharply. It was this, more than anything else, that led to rapidly declining incomes.
Stiglitz then draws a parallel between what happened to agriculture in the early decades of the last century to what has happened to manufacturing in recent times:
The parallels between the story of the origin of the Great Depression and that of our Long Slump are strong. Back then we were moving from agriculture to manufacturing. Today we are moving from manufacturing to a service economy. The decline in manufacturing jobs has been dramatic—from about a third of the workforce 60 years ago to less than a tenth of it today. The pace has quickened markedly during the past decade. There are two reasons for the decline. One is greater productivity—the same dynamic that revolutionized agriculture and forced a majority of American farmers to look for work elsewhere. The other is globalization, which has sent millions of jobs overseas, to low-wage countries or those that have been investing more in infrastructure or technology. (As Greenwald has pointed out, most of the job loss in the 1990s was related to productivity increases, not to globalization.) Whatever the specific cause, the inevitable result is precisely the same as it was 80 years ago: a decline in income and jobs. The millions of jobless former factory workers once employed in cities such as Youngstown and Birmingham and Gary and Detroit are the modern-day equivalent of the Depression’s doomed farmers.
Stiglitz argues that the banking crisis exacerbated the underlying problem of economic transformation caused by innovation. The hypothesis is interesting, but I don't think that the evidence backs it up.  Let's have a look.

The key variable for Stiglitz is productivity, which refers to how much output we get for a set of inputs. Productivity is also a measure of innovation. Improving productivity is a key element in increasing per capita wealth and overall economic growth. At the same time, productivity growth inevitably leads to disruptions and dislocations in the overall economy.  The example cited by Stiglitz of early 20th century agriculture is a good one - from 1900 to 1930 agriculture went from employing more than 40% of the nation's jobs to about 20%.

The causality is obvious -- suppose that it takes 2 people (plus land and capital) to produce 100 bushels of corn. Now suppose that innovations in agriculture -- say new tools, seeds, fertilizers etc. mean that those 2 people can now produce 150 bushels. The combination of labor and other inputs has improved productivity. Consequences of these improvements might be falling agricultural prices or a need for less labor or both, which is exactly what happened over much of the 20th century in agriculture. For those who purchase food, lower prices is generally a good thing, because it makes them relatively richer, freeing up resources previously spent on food that can now be used in other ways. For the farm sector, to maintain employment, displaced workers would need to find other jobs, probably necessitating new skills.

But does the Depression-era analogy hold today?

Let's first look at productivity growth in manufacturing (specifically, multi-factor productivity -- Data here from BLS). The graph above shows that productivity growth increased from the late 1980s to the mid-2000s. This trend is consistent with the loss of jobs in the manufacturing sector in recent decades from about 18 million jobs in 1988 to less than 12 million today or from about 15% of the labor force to about half that (data).

So far so good. But things get complicated when we look at the overall productivity of the economy. Let's take a look at productivity growth in the US economy from 1951-2010 using data from the Bureau of Labor Statistics (the data set I am using is Multi-Factor Productivity in the private business sector, Excel data here).
Over this longer term it is hard to conclude that the recent increased productivity in manufacturing has led to an unusual increase in overall productivity -- far from it, the economy has seen longer periods of higher productivity in the past.

Thus, it is not surprising that Stiglitz recommends a somewhat incoherent remedy:
What we need to do instead is embark on a massive investment program—as we did, virtually by accident, 80 years ago—that will increase our productivity for years to come, and will also increase employment now.
The answer to too much productivity is to increase productivity?!

Bottom line -- Increasing productivity in manufacturing has indeed contributed to a change in its role in the overall economy. The larger economic problem, however, is not increased productivity in one sector of the economy, but rather its correlates, such as poor responses to dealing with the resulting dislocated workers in manufacturing and those policies that facilitated the decline of manufacturing in the first place. Focusing on strengthening manufacturing (and a government role in that effort) may indeed make good economic sense, but I have a hard time locating the origins of today's economic woes in too much productivity.


  1. Every factory I worked in has closed down. During a recent trip to the home supply store, I saw the brand name of the last company I worked at. Made in Mexico.

    Productivity changes in the United States had nothing to do with it. Much of the work in the plant was low-skilled, and Mexican peasants could be taught to do it. That's where your jobs went - not eliminated by technological innovation, but replaced by low cost labor.

  2. Boy, just look at today's news on employment in Germany (increasing) while Spain's employment decreased slightly (the unemployment rate in Spain is over 22%). I think everyone know which of these two economies are more productive yet Spain is considering even more austerity as its tax revenues lag.

    The real problem is created by imbalances and no one creates more lasting imbalances in an economy than government industrial, monetary, or other policy. If Spain had its own currency, it would have been devalued when their economy tanked and the products made there more competitive leading to more business. In other words, self correcting. That's not possible with the Euro so Spain suffers and Germany prospers. In our own country, the desire of those in power for everyone to partake in the American dream of home ownership created a boom-bust cycle and brought down the financial system. Health care policies that insures someone will pay for all procedure has driven up costs so high that our government now spends more per capita on health care than most other countries yet we still pay into a private health care system and manage not to insure 15% of the population. The imbalances created by government policies today have replaced the imbalances in capital markets in earlier years. The scariest investment of them all is one that can't fail.

  3. -Mark B.-

    Thanks ... I get your point, but there is innovation there as well, it is just not innovation leading to increased productivity in the US.

  4. The wealth of an economy comes from its productivity. If people want salary increases every year, these can only be generated from increased productivity. However, as noted here, increased productivity means fewer employees needed per unit of good or service.

    This conundrum has generally been overcome by using that increased wealth to buy goods and services that were previously out of reach of the non-rich, e.g. appliances, cars, airplane flights, etc. The new demand for such things created new employment. That's how tourism, e.g, become one of the largest industries in the world, employing many millions.

    Most importantly, new wealth supports the creation of whole new goods and services, e.g. personal computers, programming, electronic games, cell phones, dog grooming, personal trainers, adventure tourism, etc. Govt policy should focus on making it as easy as possible to create new companies offering new products and services.

    Govt should, however, avoid the "hour-glass syndrome". I.e. lots of very small companies at the bottom, big old legacy giants at the top and very few firms in-between. I heard about this will living in Europe where growing a small company is extremely hard due to the lack of venture capital, high taxes and onerous employment rules that kick in as a company grows, heavy regulation, etc.

    There is essentially an infinite number of new goods and services that could be created as productivity growth makes the economy wealthier. We just need to make sure it's easy both to start and to grow the companies that will provide such things.

  5. The way I've always wrapped my head around productivity and unemployment is a story I think I read in the comments here about an American diplomat visiting China that may or may not be true but that doesn't matter because I think it illustrates things well.

    The diplomat sees a dam being constructed by a hundred men with shovels and asks his Chinese counterpart: "Why don't you use a bulldozer? With one bulldozer you could be doing the work of a hundred men." The Chinese diplomat replied "Yes, but the what would these men do for work? They would be out of a job!"

    Now the implication of this story is that the negative effect (lost jobs) of increased productivity on the workers is offset by their ability to be re-trained to do something else. Perhaps education is the issue?

  6. those policies that facilitated the decline of manufacturing in the first place.

    What makes you say there was a decline in American manufacturing? American output as a share of global manufacturing output has held steady for the last thirty years (with some ground lost during the recession).

    Please tell me you're not one of those who confuses decline in manufacturing employment with a decline in the sector itself.

  7. -6-Christopher

    Thanks ... yes, "manufacturing employment" would have indeed been more accurate.

  8. My "1% argument" would be that: "Improving productivity is a key element in increasing per capita wealth and overall economic growth." only is true if the gains of said productivity go largely to the workers and not mainly to the CEO or investors. In addition to policies that foster a strong middle class, one way to handle increasing productivity and still retain workers is to improve working conditions. In this country, we seem to be going in the opposite direction on that.

    It seems to me that many of the current job losses come from off-shoring, and don't directly factor into a productivity analysis in this country. Also, if we were to convert to new industries, new jobs would be created to replace obsolete older ones. Then, Stiglitz' argument would make sense to me in that investing towards building industry here in this country would help more with creating new jobs, or jobs that might otherwise be elsewhere, and less to do with taking away existing jobs by productivity improvements.

  9. ""The diplomat sees a dam being constructed by a hundred men with shovels and asks his Chinese counterpart: "Why don't you use a bulldozer? With one bulldozer you could be doing the work of a hundred men." The Chinese diplomat replied "Yes, but the what would these men do for work? They would be out of a job!""

    The potentially apocryphal ending to that potentially apocryphal story is that the American says, "Oh, you're worried about employment? Then why not force them to use spoons?"


    True or not, it's a good story with a useful lesson.

  10. A missing piece is effective demand. The loss of agricultural employment in the 30s was finally off-set in the 50s with the tremendous expansion of the consumer goods industries including automobiles. More recently education and health care employment have expanded enormously -- but these are close to saturation (not to mention close to the Chinese dam story). Efforts to create an energy revolution akin to the communications revolution of the 80s and 90s, appears to be simply failing due to the lack of actual benefits.
    I have no clue as to where the next increase in effective demand will come from to off-set our continued increase in productivity.

  11. Roger - Unfortunately, the conflation is extremely common and makes discussion of the issues rather difficult. Absolute decline in manufacturing has very different causes and potential solutions than an absolute rise in output and fall in employment.

    Manufacturing is the direct physical production of goods and is extremely sensitive to advances in technology. The large gains in the sector suggest slow technical innovation isn't the problem. The question therefore shifts to the other major sectors of the economy, services and government. Why aren't the productivity gains seen in manufacturing occurring in these two sectors (retail and finance being exceptions)?

  12. Roger – Stiglitz is basically right, but doesn't look far enough ahead:

    The fruits of our "command" of nature have been both sweet and toxic.

    If and when we successfully decontaminate our air and water – and our prejudiced minds – if we can avoid nuclear disasters, control climate and control population growth, then certain aspects of the future seem clear.

    Science and technology (and especially automation and information technology) will continuously increase the productivity per person employed.

    This must lead, either to a continually decreasing average work-week, or to massive industrial unemployment.

    To avoid economic, let alone social chaos, will require continuing redistribution of wealth to workers (who will be gradually becoming workers in name only), either through tax-supported welfare programs or through extensive redistribution of ownership of the means of production. Most of the population will be effectively "on relief", as are many of the "idle" rich. There are no reasonable alternatives.

    But when we consider the depressing failures of recent efforts to provide government relief for the impoverished:

    destroyed sense of self-worth; destroyed motivation for self-improvement; growing infant mortality; increases in teen-age, out-of-wedlock births; and growing drug dependence,

    there will be little confidence that any kind of new governmental intervention/regulation can work. It should be distressing that the children of the affluent often exhibit a significant part of the same syndrome.

    It appears that if genuine satisfaction is not derived from daily occupation, a sense of self-worth and of social responsibility withers.

    Unfortunately, there is no obvious way that either economic isolationism, or an unrestrained free market, can solve this problem.

    Perhaps our only hope is to restructure education and our schools and universities to increasingly train the next generations for "productive, self-fulfilling leisure".

    Much community service, teaching, research, personal care, participatory sports and studio/performing arts will probably never be replaced by machines, and could occupy most of the population. But measures of value will have to develop which are somewhat different from those of the present marketplace. It must become convincing, that such occupations, by positively contributing to the quality of life, are "worth" the entitlement payments or stock-distributions necessary to keep the 'economy' working, are not merely "busy work" and are important to the very survival of humanity. This is a tall order.

    Virtually no one discusses such a "brave new world".

  13. A number of others have pointed out that freeing up innovation in new sectors is key to reinvigorating a changing economy. Unfortunately, the default response is to lock down and protect existing jobs when we should be facilitating the creation of new jobs.

    If I were a gov. planning wonk, I'd point out that there are always new things for people to consume. And the newest and biggest of these is healthcare.

    There is an enormous opportunity for healthcare to consume excess wealth from the wealthy while providing loads of jobs to many others of all skill-levels. In fact, the U.S. could be a major provider of healthcare to foreign customers - if we opened up the system and encouraged the same sort of cost saving innovation that led to Dr. Shetty's hospitals in India.

    As an aside, the regulatory and legal burdens of employing others are so high that companies shirk from doing so unless they absolutely have to. This encourages productivity growth that might actually cost more than it would be worth in a loser labor market.

  14. "What makes you say there was a decline in American manufacturing?"

    Although not unique to the U.S., manufacturing has declined as a % of GDP.


    But the decline in the U.S. has been steeper (these numbers are greater than in the above graph - no idea why):

    "America's manufacturing sector has retreated faster and further in relative terms than that of any other large, affluent nation. US manufacturing as a percentage of GDP declined from 27 percent in 1950 to 23 percent in 1970 to 14 percent in 2000 to 11 percent in 2009. While manufacturing as a share of GDP has also declined in Germany and Japan, both countries have retained relatively larger manufacturing sectors at 17 and 21 percent, respectively. The contribution of manufacturing to per capita GDP is also higher in Germany ($6,900) and Japan ($8,300) than in the United States. The most shocking, but underemphasized, fact about global manufacturing is that Germany's share of global merchandise exports is actually higher than America's (9 percent vs. 8.5 percent in 2009), despite having an economy just one-quarter of the size."

  15. Received by email:

    "Dear Sir,

    Heres my comment 5/1/2012

    No one has really spelt out the problems which are as I see it:

    1. As a general rule academics are in charge of education, so naturally they like academic curricula. In the UK from Shakespeares time practical people who make things have been looked down upon. Academic learning's everything here in the UK.

    2. Practical education is expensive for the government to provide and nobody really wants to pay for it anymore, certainly in the past engineering employers with one or two honourable exceptions did not like to meet these costs.

    However their is some hope here now as the government is starting some university technical schools (organised and attached to universities, why the universities I don't know, I guess the hope is to give status) round the UK. I just hope the students get technical jobs when they leave school!! As our manufacturing base shrinks in the UK.

    3. Then you have got exam grade inflation,and general lowering of performance standards (even allowing for recalibration of exams) required of the students over the last 40 years. I don't blame the young people for this its the system that has been manipulated at the instigation of politicians, so that we celebrate success but can not admit to failure. Politicians want to get elected again so all children of electors must do well. I believe that this issue is fast coming to a head in the UK and maybe a Royal Commission will be instituted to look into the whole matter.

    4. Then there is lack of status and therefore earning power for qualified engineers (that is C.Eng. and I. Eng)is well known in the UK. Cross country train drivers are on about £50K because they are highly protected and unionised, unlike C. & I Eng's who are usually staff employees if they have a job. One could argue that trade electricians do better as much electrical work has to be done by certificated electricians to be acceptable. Would that qualified engineers had the same protection.

    5. Overseas competition(but not US, Canadian or European)where labour rates and employee costs are said to be very low, thus reducing the on costs of manufacturing in these countries so that they can sell products in the west much more cheaply than similar home produced articles. The environmental damage done by some of these companies is ignored and not on the balance sheet nor is the adverse well being of their employees.

    Now at least one major European politician is throwing doubt on whether this unequal trading should continue. Huge cargo ships are said to arrive full in Europe and go home empty, except perhaps for scrap metal and waste for recycling. Is this equal trade?

    6. I have also heard of good copies of western products being made overseas and then being sold into western countries, but nothing much appears to be being done about this by western governments, accept interminable trade talks which go on for years with very outcome.

    Do we have to wait until we have no industry left before our largely academically trained politicians and government officials with very little business experience for the most part come to realise the danger we are in.

    7. And lastly what do we do about some, only some of those entities, on our side of the fence who are making a lot of money out of the current trading set up but do not care at all where all this will ultimately lead, the demise of the west. Western companies do need to operate all over the world but one hopes that they respect the environment and their employees. I see this aspect of the problem not as either or but a question of degree.

    Just to finish, you may not believe me but I do believe in free trade, as long as it really is free, and as long as it is trade and not a one way street.

    Yours faithfully R Turner"

  16. Although not unique to the U.S., manufacturing has declined as a % of GDP.

    This is because other sectors have expanded faster (namely services), not because we produce less. American manufacturing output (measured by value of goods produced) has in fact doubled since 1980 even while the number of people employed has gone off a cliff.

  17. "This is because other sectors have expanded faster (namely services), not because we produce less. "

    Sure, but I think it is worthwhile to consider the reasons for that.

    Technological growth would be one reason.

    Other reasons:

    It's easier to turn cheap, unskilled labor into profits in the service sector.

    Manufacturing requires long-term business plans for producing quality products over long-term time frames at competitive prices - why waste time with that when investors and those trained in finance can move over to the exploding financial sector (not sure how the growth in the financial sector compares to growth in the service sector, but I'd guess that they are comparable) and capture exponential returns through short-term leveraging?

    And of course, as for increasing profitability and production without increasing employment: Manufacturing executives at the upper levels of corporate structure can increase their income relative without increasing the income of the average worker by exploiting cheap foreign labor. Certainly the disproportionate growth in executive pay is a factor in the changed ratio of productivity relative to employment.

  18. I like what Menth says in comment no. 5, but I don't think education is the problem. What I feel is that people make time and increased productivity an issue as if they were in a rush to do things when in certain areas, there is no problem as far as demand is concerned. What this means is, why use a bulldozer to do the work of a hundred men who will lose their jobs and then need to be retrained at the cost of the taxpayer when there is no real rush to build the dam. Sometimes, a slow steady managed approach is key to not rocking the boat. As long as the company is profitable, even at the expense of 100 workers salaries that thereafter contribute to the economy, why use a bulldozer so the company can have horrendous profitability margins only to add 100 workers to unemployment rolls and education benefits for retraining at the cost of the taxpayer, unless of course, the company is insanely ambitious over profit without giving any consideration to how a reduction in force is going to affect the area economy. This is what we are living in America today...