Clean energy: Learning by doing only happens if we DO!

A policy of aggressive deployment of low-emission energy technologies is the path most likely to help the world meet its climate objectives, argues Jonathan Koomey, author of Cold Cash, Cool Climate: Science-based Advice for Ecological Entrepreneurs. Koomey says it’s misguided to think that clean energy technologies can be made cheaper than conventional alternatives on an internal cost basis without aggressive deployment of existing technologies. He also states that the goal of cost competitiveness without considering externalities is itself the wrong metric for success.

The following is a guest post, in response to our question.

Climate Science Watch:

Your most recent book lays out the urgency and enormous challenge of global climate disruption and argues that policy to jump-start immediate reduction of carbon emissions using technologies that are already on the market is the most effective approach to driving the innovation needed for a decarbonized energy transformation. How would you compare your analysis with an approach that separates energy policy from climate change and emphasizes instead the development of breakthrough technologies toward the goal of ultimately making clean energy cheaper than the alternatives?

Koomey response:

It would be terrific if clean energy were cheaper than conventional alternatives on an internal cost basis, because then the fight to promote low emissions technologies would be a lot easier (and the political conflicts less contentious, which seems to be what people who take this position yearn for).  But using this goal as an excuse to avoid aggressive deployment of existing technologies is misguided on at least two levels.

First of all, there’s nothing sacrosanct about current comparisons of internal economic costs, which ignore externalities and reflect historical subsidies as well as arbitrary choices about how to structure property rights and markets.  If we measure economics based on societal costs including externalities, energy efficiency, wind, small hydro, and some other renewables are already wildly cost effective compared to coal and oil-fired electricity generation, even including very generous allowances for backup costs.[1]  For example, even the lowest estimates of external costs from existing coal fired generation (ignoring the climate externalities entirely) are about 8 cents per kWh, according to Epstein et al. (2011).  And Muller et al. (2011) conclude that coal and oil fired generation contribute net negative value added to the US economy when all external costs are included.  So a proper societal cost comparison already justifies replacing these polluting resources with widely available renewable technologies, and with solar photovoltaic (PV) module costs falling 80% in the past five years, PVs aren’t far off from wild cost effectiveness as well, as long as you do a proper societal cost comparison.

Second, somebody who believes that fostering a clean energy transition is primarily a question of inventing new technologies in the near and medium terms doesn’t really understand what the latest literature says about these issues.  Costs can come down because of breakthrough technologies (what economists call “invention”) but much more often they come down because of learning effects for technologies that are already on the market, and these learning effects tend to be sustained and powerful over long periods.  Most conventional economists downplay the importance of learning effects because they are difficult to model (Arthur 1990), but I’m convinced these effects are critical for understanding our options for climate mitigation.

In addition, we just don’t have time to wait for new innovations, because of the nature of the climate problem (more on that below).  Fortunately, we have lots of currently available options from which to choose.

Neoclassical economics has taught us a lot about how economies work, but that discipline is based on a set of assumptions that often don’t reflect economic decision making in the real world.  For example, most economic models assume perfect & costless information, perfect competition, no externalities, no transaction costs, and constant or decreasing returns to scale.  In reality, however, information is imperfect and costly, transaction costs can be large, and increasing returns to scale are pervasive.  These (and other) factors lead to what’s called “path dependence”, meaning that our choices now affect our options later.

For example, if we invest in deploying mass produced technologies (like solar panels and wind turbines) we move down the learning curve, thus reducing the costs of those technologies five or ten years hence.  If we deploy fewer of those devices, we don’t move as far down the learning curve and their costs in 2020 will be higher than they would be in the case where we more actively promote deployment of these technologies.

Learning effects in particular are dependent primarily on deployment, not on new inventions.   Learning by doing only happens if we DO, and lots of the needed innovations are in our institutional structures, legal arrangements, and social norms, not just in particular widgets. In a very real sense the future is ours to create—our choices now affect our options later.

This point of view is well supported by recent history.  A recent example is that of photovoltaic installations in Germany vs. the US, where the feed-in-tariffs (and associated policy changes) in Germany made the installed costs for PVs in that country 30 to 50% lower than for those installed in the US.  The chief difference is related to the so-called “soft costs”, which are affected by the maturity of the industry installing the products. Significantly greater cumulative sales (3.6X bigger for residential installations in Germany compared to the US) led to learning effects that substantially reduced those soft costs.

The power of these learning effects is reflected in the Fourth Assessment Report of the Intergovernmental Panel on Climate Change (IPCC 2007, WG3, section 2.7.1 Technology and climate change). In particular, that report states (on p. 151) that

The treatment of technological change in an emissions and climate policy modeling framework can have a huge effect on estimates of the cost of meeting any environmental target.  Models in which technological change is dominated by experience (learning) curve effects show that the cost of stabilizing GHG concentrations could be in the range of a few tenths of a percent of GDP, or even lower (in some models even becoming negative)—a finding also confirmed by other modeling studies (eg Rao et al., 2005) and consistent with the results of the study by Gritsevskyi and Nakicenovic (2000) reviewed above, which also showed identical costs of “high” versus “low” long-term emission futures.  This contrasts with the traditional view that the long-term costs of climate stabilization could be very high, amounting to several percentage points of economic output (see also the review in IPCC 2001).

So the cutting edge economics literature that acknowledges the power of the learning effects comes to a very different conclusion from those who advocate R&D and new inventions as the main response to climate change.  No one disagrees that we should continue aggressive R&D on climate solutions, but the idea that R&D should be the primary focus for climate mitigation is deeply misguided.

Another source of path dependence is the nature of the climate problem itself.  Because the most important greenhouse gases stay in the atmosphere for a long time, it’s the cumulative emissions of greenhouse gases that matter.  That means that we can emit only a fixed amount of carbon (our “carbon budget”) if we want to stay under the 2 Celsius degree warming limit that the US and other major nations accepted at Copenhagen in 2009.  If we burn more high carbon fuels now, we commit ourselves to even faster reductions in emissions later (because the total carbon budget over the next century is fixed).

It is this reality that yields the urgency for climate action in the near term.  As I discuss in Cold Cash, Cool Climate, choosing a warming limit like 2 C implies the need for immediate emissions reductions, not just R&D into technological options. [2]   And the powerful learning effects related to technology adoption make a path with aggressive deployment the only one likely to yield emissions reductions that are large enough and fast enough to make a real difference for the climate.

References

Arthur, W. Brian. 1990. “Positive Feedbacks in the Economy.” In Scientific American. February. pp. 92-99.

Epstein, Paul R., Jonathan J. Buonocore, Kevin Eckerle, Michael Hendryx, Benjamin M. Stout Iii, Richard Heinberg, Richard W. Clapp, Beverly May, Nancy L. Reinhart, Melissa M. Ahern, Samir K. Doshi, and Leslie Glustrom. 2011. “Full cost accounting for the life cycle of coal.”  Annals of the New York Academy of Sciences.  vol. 1219, no. 1. February 17. pp. 73-98. [http://dx.doi.org/10.1111/j.1749-6632.2010.05890.x]

Gritsevskyi, Andrii, and Nebojsa Nakicenovic. 2000. “Modeling uncertainty of induced technological change.”  Energy Policy.  vol. 28, no. 13. November. pp. 907-921.

Rao, Shilpa, Ilkka Keppo, and Keywan Riahi. 2006. “Importance of Technological Change and Spillovers in Long-Term Climate Policy.”  The Energy Journal.  vol. 27, pp. 25-42.

IPCC. 2007. Climate Change 2007: Mitigation of Climate Change–Contribution of Working Group III to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change [Metz, B., O. Davidson, P. Bosch, R. Dave, and L. Meyer (eds.)]. Cambridge, United Kingdom and New York, NY, USA.: Cambridge University Press. [http://www.ipcc.ch/publications_and_data/publications_and_data_reports.shtml]

Krause, Florentin, Wilfred Bach, and Jon Koomey. 1989. From Warming Fate to Warming Limit:  Benchmarks to a Global Climate Convention. El Cerrito, CA: International Project for Sustainable Energy Paths.  Republished by John Wiley and Sons in 1992. [http://www.mediafire.com/file/pzwrsyo1j89axzd/Warmingfatetowarminglimitbook.pdf]

Muller, Nicholas Z., Robert Mendelsohn, and William Nordhaus. 2011. “Environmental Accounting for Pollution in the United States Economy.”  American Economic Review vol. 101, no. 5. August. pp. 1649–1675. [http://nordhaus.econ.yale.edu/documents/EnvAccount_MMN_AER0811.pdf]

Footnotes