21 May 2012

Beyond Manna from Heaven

Writing at The Breakthrough Journal blog, Ted Nordhaus and Michael Shellenberger argue that conventional economics is not up to the task of offering sound policy advice for the 21st century.  They write
In the 70 years that have passed since Joseph Schumpeter coined the term "creative destruction," economists have struggled awkwardly with how to think about growth and innovation. Born of the low-growth agricultural economies of 18th Century Europe, the dismal science to this day remains focused on the question of how to most efficiently distribute scarce resources, not on how to create new ones -- this despite two centuries of rapid economic growth driven by disruptive technologies, from the steam engine to electricity to the Internet.

There are some important, if qualified, exceptions. Sixty years ago, Nobelist Robert Solow and colleagues calculated that more than 80 percent of long-term growth derives from technological change. But neither Solow nor most other economists offered much explanation beyond that. Technological change was, in the words of one of Solow's contemporaries, "manna from heaven."
Where does that "manna from heaven" originate? In pricing incentives, of course, derived from economic theory. But once you take a closer look at both practice and the theoretical origins, you find that economics explains far less than we've been led to believe.

Nordhaus and Shellenberger revisit the climate issue to illustrate how far conventional economics has led us astray. They provide an overview of a debate that they engaged in with an economist from the Environmental Defense Fund, Gernot Wagner, who argues against evidence and common sense that by creating the right pricing incentives, drastic emissions reductions goals can be met in the near term:
 "[W]e can achieve US emissions reduction goals for 2020 and possibly even 2030 through deployment of existing technologies. . . Price goes up, demand goes down. Economists typically call it the 'law of demand'--one of the very few laws we've got."
The theory is sound, its application is not -- a point that readers of this blog and The Climate Fix will well understand.

The good news is that many are beginning to move beyond the precepts of economic theory and take a look at the simple mathematics of the real world. For instance,  Ulrich Hoffmann, an economist at the UN Conference on Trade and Development, has done the math, which is illustrated in the figure below showing how much the world would need to decarbonize its economic activity in order to stabilize carbon dioxide at 450 ppm.

Based on these straightforward mathematics he concludes (Hoffmann has a more in-depth analysis here in PDF):
The arithmetic of economic and population growth, efficiency limits related to the rebound effect, as well as systemic issues call into question the hopes of decoupling economic growth from GHG growth. One should not deceive oneself into believing that such an evolutionary (and often reductionist) approach would be sufficient to cope with the complexities of climate change. “Green growth” proponents need to scrutinise the historical macro- (not micro-) economic evidence, in particular the arithmetic of economic and population growth, as well as the significant influence of the rebound effect.
Such messages are not well-received by conventional economists. In their exchange, Wagner explains to Nordhaus and Shellenberger that economic theory trumps real world evidence, and this means that debate over such issues is not necessary:
The main points from climate economics are no longer up for debate: carbon is a pollutant; we need make polluters pay, either through a cap or a price. Virtually all economists agree--from Holtz-Eakin, Laffer and Mankiw on one side to Stiglitz, Sachs, and Krugman on the other.

Once again, this one is not up for debate. You can argue that politically we can't get there, so we need to do other things in the short term, but it's not up for debate whether this is the economically correct solution.
Like most debates on climate this one ends predictably, with Wagner appealing to the motives of Shellenberger and Nordhaus:
[Y]our entire operation seems to be geared toward propagating contrarian-sounding views that once in a while get you some attention and get picked up by an editor somewhere, but otherwise are just that: contrarian for the sake of wanting to be different from the pack
Snore. But the larger point here is that there are articles of faith in the discipline of economics which are viewed as taboo to challenge, even when they fail to represent themselves in practice with the simplicity and elegance of theory.

However, a closer look at economic theory finds a much shakier foundation than is represented within the discipline. Writing at Slate, Konstantin Kakaes has a great piece that sums up how economics went astray when it comes to innovation:
Robert Solow, winner of the 1987 Nobel Memorial Prize in Economic Sciences, is famous for, in the recent words of a high-ranking State Department official, “showing that technological innovation was responsible for over 80 percent of economic growth in the United States between 1909 and 1949.”. . Typically, technical or technological progress isn’t explicitly defined by those invoking Solow, but people take it to mean new gadgets.

However, Solow meant something much broader. On the first page of “Technical Change and the Aggregate Production Function,” the second of his two major papers, he wrote: “I am using the phrase ‘technical change’ as a shorthand expression for any kind of shift in the production function. Thus slowdowns, speedups, improvements in the education of the labor force, and all sorts of things will appear as ‘technical change.’ ” But his willfully inclusive definition tends to be forgotten.

Solow was constructing a simple mathematical model of how economic growth takes place. On one side was output. On the other side was capital and labor. Classical economists going back to Adam Smith and David Ricardo had defined the “production function”—how much stuff you got out of the economy—in terms of capital and labor (as well as land). Solow’s point was that other factors besides capital, labor, and land were important. But he knew his limitations: He wasn’t clear on what those factors were. This is why he defined “technical change” as any kind of shift (the italics are his) in the production function. He wasn’t proving that technology was important, as economists in recent years have taken to saying he did. All Solow was saying is that the sources of economic growth are poorly understood.
Instead of "technology" Solow was really talking about "innovation." That innovation need not be understood because it was "manna from heaven" is characteristic of many arguments from conventional economists and is particularly endemic in the climate debate. From such a perspective, of course anyone asking about where innovation comes from -- other than from the magic of the invisible hand -- must either be ignorant or malign.

But as Hoffmann's essay explains, once you actually do the math of energy innovation in the context of real-world social and political forces, you see that understanding processes of innovation requires more than simply understanding the "law of demand."

Kakaes contines:
The cautionary tale of Solow is emblematic of how economists get science and technology wrong. One economist creates a highly idealized mathematical model. The model’s creator is, as Solow was, honest about its limitations. But it quickly gets passed through the mill and acquires authority by means of citation. A few years after Solow’s paper came out, Kenneth Arrow, another Nobel Prize winner, would write that Solow proved the “overwhelming importance [of technological change in economic growth] relative to capital formation.” It’s a sort of idea laundering: Solow said that we don’t know where growth in economic output comes from and, for want of a better alternative, termed that missing information “technical change.” But his admission of ignorance morphed into a supposed proof that new technologies drive economic growth.
Remarkably, in the 21st century, our policy debates reflect the fact that we do not have a good idea where innovation comes from, how it is directed and how we prepare for its inevitable downsides. Too often conventional economics presents an an obstacle to debating and discussing this topic.

As Nordhaus and Shellenberger conclude,
Over the next century, global energy demand will double, and perhaps triple. But even were energy consumption to stay flat, significantly reducing emissions from today's levels will require the creation of disruptive new technologies. It's a task for which a doctrine focused on the efficient allocation of scarce resources could hardly be more ill-suited.
Read the three essays discussed here in full at The Breakthrough Journal blog, at Bridges Trade BioRes review, and at Slate.


  1. Jesus, how about a brief look at what economists have been doing in the last 20 years, as opposed to what was en vogue 60 years ago?? Solow's seminal contribution is from 1956, that is NOT the state of the art in economic research. To claim that economists have ignored the question where technological change comes from is simply absurd. The "manna from heaven" phrase is an accurate description of the field 60 years ago. Paul Romer's work in the 80s started what is known as endogenous economic growth theory that specifically deals with the question of technological change and where it comes from. That strand of research by now has made it into textbooks (a great one is Acemoglu's).

    Before you try to bash a field for ignoring a question, do yourself a favor and ask someone who knows the literature what people have been doing lately (uhh, or for a long time in this case).

  2. -1-dave

    Thanks for the comment, and you are of course correct that this critique is not of an entire discipline, but a pretty strong thread running through it.

    In the fields of climate economics and science and technology policy, I am afraid that this critique is accurate from where I sit. There are of course exceptions (like Hoffmann cited above).

    However, if you'd like to add another perspective, you are most welcome. Thanks!

  3. Dave, I'm a mature student doing a Masters in Economics at the LSE. I was staggered at the lack of advance since I did my undergrad in late 70's. I wish I'd done Byzantine Art or something. The Solow Residual is well-named, as the post says 'Solow said that we don’t know where growth in economic output comes from'. Pace Romer and Acemoglu, I'd say that's still the case.

    (It might be that the LSE is way behind the times, but that seems improbable.)

    I'd have to agree with Roger / Breakthrough 'that economics explains far less than we've been led to believe' based on my real-world experience of 30 years in markets coupled with my year-long glimpse at what I'd hoped would be a far advanced social science.

    The Wagner / Breakthrough specific example in this blog post is, imho, a pretty good one of economists refusing to believe they don't know.

    Anyway, I shall now head off and read the linked essays.

  4. dave, specifically, in your comment you say 'To claim that economists have ignored the question where technological change comes from is simply absurd.'

    I don't think Roger / Breakthrough claimed the question had been ignored? The claim was more that no answers had been found. Nobel prizes await!

  5. If the state of economics is as inept as portrayed, it would simply be joining cancer research, climate science, biotech research, and most of the social science presently done in the academy. Or the risk models used by financial "geniuses" at banks such as J P Morgan.

    Hubris runs amok. Never have so many been so afflicted with confidence in models and theories which accurately explain so little.

  6. Technological/scientific advance is a bit of an uncomfortable area for economics due to it falling outside of the normal supply/demand framework. There are too many exogenous influences for it to be reliably accounted for.

    If something is in demand, but physically impossible, no price will bring it into existence. Conversely, an advance may come even if no one previously demanded it (most never anticipated they might even want it until it existed).

    I believe that Roger is pointing to this issue as the flaw in cap and trade or a carbon tax. It's making the implicit assumption that sufficient innovation will occur (and therefore the actual costs will be limited) as the price rises purely based on supply and demand.

  7. Technological change comes in many varieties:

    To cover the Sahara and Australian Outback with irrigated forests:


    and to sustainably, eco-neutrally harvest old-growth, tropical forests:


    are enormous technological changes, that only require well-established, RATHER THAN NEW technologies.

    Combined, they could not only slow down increase in atmospheric CO2, but in fact draw down the level to about 280 ppm AND provide a sustainable source of negative, or near-zero CO2-footprint bio-fuel, to sustain global energy needs 'indefinitely' IF population growth stops near 10 billion!

    Different ways of using EXISTING technology also constitute "technological change". Such change doesn't necessarily require new 'invention', in the sense used by economic theorists who never invented a thing in their lives.

  8. Via Twitter Richard Tol says:

    "@RogerPielkeJr Stop saying that economics doesn't understand innovation when there's a large body of literature that you're unfamiliar with."

    I have extended to him an invitation to author a guest post here explaining how economics does get innovation.

  9. As a card-carrying economist, I am reluctant to get into this.

    Nordhaus and Schellenberger state that Solow is an exception, one of the few economists interested in technological progress. Nothing could be further from the truth. Since Dixit and Stiglitz (1977), technological progress has been at the forefront of economics.

    It is unfortunate that Nordhaus and Schellenberger cite Slate Magazine rather than the large body of academic literature on how to accelerate and direct technological progress in energy and environment. Leading lights are Carolyn Fisher, Adam Jaffe, Richard Newell, David Popp, with a much larger cast of supporting figures. Theoretical, econometric and qualitative studies agree that a carbon tax is the preferred instrument (as innovation and diffusion are key, rather than invention) but there is disagreement about the desirability of auxiliary instruments.

  10. Hi,

    I maintain that economics doesn't understand what causes economic growth. And I've got $200 for any economist who begs to differ (and who has published predictions for economic growth during the 21st century). :-)


    Here are IPCC projected growth rates for the 21st century:




    Note that they predict (oops..."project") that economic growth will dramatically *slow down* relative to the growth from 1950-1990. I predict it will dramatically increase, due to increases in machine intelligence:


    Why does this matter? It matters because, even if ambient air capture of CO2 costs $1000 per ton of CO2 ($272 per ton of carbon), it will be an absolutely trivial matter (less than a few percent of GDP for a period of 1-2 decades) for the world to lower the CO2 in the atmosphere to the pre-industrial concentration of 280 ppm by the second half of this century.

    Poof! Global warming (and ocean acidification) are completely eliminated by the investment of a few percentage points of world GDP for 1-2 decades.

    P.S. Details: $272 per ton of carbon. There are 2 gigatons of carbon per ppm of CO2 in the atmosphere. Therefore, it costs $545 billion to reduce CO2 in the atmosphere by 1 ppm. By the year 2050, world GDP will be on the order of $1 quadrillion. (That's 10 billion people, with a per capita GDP of $100,000.) Assuming the CO2 concentration in 2050 is 480 ppm, the reduction to 280 ppm would be 200 ppm, and the cost would be 200 x $545 billion = $109 trillion. That would be about 11 percent of the GDP of 1 quadrillion in 2050. So the world could spend 1 percent of GDP for 11 years, and get the concentration back down to the pre-industrial concentration of 280 ppm.

    P.P.S. When I say "economics" doesn't understand what causes economic growth, I'm not including Robin Hanson, who predicts that, circa the 2020-2030 time frame:

    "Without machine intelligence, world product grows at a familiar rate of 4.3% per year, doubling every 16 years, with about 40% of technological progress coming from ordinary computers. With machine intelligence, the (instantaneous) annual growth rate would be 45%, ten times higher, making world product double every 18 months!"


  11. -9-Richard Tol

    Thanks ... please share 3-5 academic papers that you think represent the best examples of work on innovation in the literature, and I will be happy to read and summarize on this blog. Not names, but papers.


  12. -9-Richard Tol

    Would you please explain what you consider to be the difference between "innovation" ans "invention"?

  13. 13, Len
    Technological progress comes in three parts.

    Invention: a new idea, a new blueprint. This requires clever and creative people going about their peculiar ways until that rare Eureka moment.

    Innovation: take a new idea to its first sell. This requires cleverness, creativity, but most of all hard graft, many trials and hopefully not as many errors.

    Diffusion: from the first sell to substantial market penetration. This requires incremental improvement in design, manufacturing, and marketing.

    As many products and processes are composites, the distinction is not always clear in practice.

    The crucial point of this classification is that the different stages of technological progress require different personalities -- say, nutty professor / engineer / manager -- and different policy interventions -- say, freedom and prospect of fame / prospect of lucrative IPO or buy-out / profits.

  14. -13, 14-Len and Richard

    Richard is sharing the Schumpeterian characterization of technological progress, which is a form of the well worn and long critiqued "linear model of innovation."

    We also know it as basic-applied-development (or the BAD model;-)

    There is a large literature on why this model fails to accurately represent technological innovation in theory or practice. Schumpeter got many things right, including most importantly the significance of innovation -- but his description of process of technological innovation is not why he ought to be remembered. More on this soon!

  15. -15- Roger
    There was no "linearity" (why do pol sci abuse math?) implied.

  16. -16-Richard

    I didn't name it, but if you prefer, it has also been called the "stages of technological innovation" model (to use the term you introduced).

    Whatever you'd like to call it, it has been roundly critiqued, and it is the critique rather than the semantics that I am most interested in.

    Also, given your deep interest in academic pedigrees, you'll like to note that I also have a degree in math. (Why do economists abuse disciplines?;-)


  17. Also, a good lit review of the "linear model of innovation" and its characterization in the discipline of economics and S&T policy:


    Godin cites Schumpter as saying that invention “is without importance to economic analysis” which gets right back to the theme of this post.


  18. -17- Roger
    I did not implicate you in the awful term "linear" to describe a one-directional sequence.

    In reality, there is a lot of back and forth and no neat separation of roles.

    That does not make the Schumpeterian classification any less useful.

  19. -19-Richard

    Sorry if I misunderstood, I took the snide comment about poli sci to be about me ;-)

    I do not find the Schumpeterian classification useful for much, other than obscuring the innovation process into unhelpful categories.

    But I am sure that our different views derive from the same disciplinary perspectives that lead us to different views on the utility of economics to provide insight into innovation.

    More on that soon, drawing upon the nice set of readings that you sent along ... Thanks!

  20. @Roger: Have a look at Acemoglu's "Introduction to Modern Economic Growth" book if you want a feel where economists stand at this point in time. You can pick five papers, but that won't give you a sense for an area that has been heavily researched for more than 20 years with a very large body of published papers. Acemoglu's book is up to date and is the best summary you can get when you want to know where economists are on this topic.

  21. -21-dave

    Thanks, several people have suggested Acemogolu's textbook, and I have been reading it ... quite good, but nothing in it appears to contradict this post.

    There are some semantic issues at play here involving different conceptions of "innovation" as well.

    I'll follow up with a post next week. Thanks!

  22. Hi,

    I've recently realized a huge mistake in my analysis in #11. I wrote "...$1000 per ton of CO2 ($272 per ton of carbon),..."

    However, $1000 per tonne of CO2 is $3667 per tonne of carbon, not $272 per tonne of carbon. In other words, I should have multiplied by 3.67, not divided by 3.67.

    Therefore, my calculation was too low by a factor of 3.67 squared. So the real calculated cost to reduce the atmosheric concentration by 1 ppm of CO2 (for an ambient capture cost of $1000 per tonne of CO2) would be 2.13 GtC x 3.67 tonnes CO2/tonne C x $1000 per tonne CO2 = $7.8 trillion.

    However, if the global GDP hits $1 quadrillion somewhere in the 2040 to 2100 time frame, as I expect it to, even ambient air capture at $7.8 trillion per ppm of CO2 removed from the atmosphere wouldn't be out of the question. Imagine, for example, if the world GDP was $1 quadrillion, and 5% of that was set aside for CO2 capture. That would be $50 trillion per year. At $7.8 trillion per ppm, that would result in removal of 6.4 ppm per year. So if we assume the CO2 concentration in 2050 is about 500 ppm, and being added to at a rate of 3 ppm per year, it would take about 5 decades to get down to 350 ppm. Spending of 5% of GDP per year is probably less than the world economies spent on the military in the 50 years from 1935 to 1985.

    So even my huge mistake doesn't change the conclusion that, if the world economy grows as I expect it to, it makes much more sense to simply let the richer world of the latter half of the 21st century scrub CO2 from the atmosphere.