Blogs

Energy, Security, and Climate

CFR experts examine the science and foreign policy surrounding climate change, energy, and nuclear security.

Latest Post

REUTERS/Amit Dave
REUTERS/Amit Dave

Why We Still Need Innovation in Successful Clean Energy Technologies

Today is my last day at CFR. I’m joining ReNew Power, India’s largest renewable energy firm, as their CTO. I’m excited for a new adventure but sad to leave the Council, which has given me support and autonomy to study the innovations needed for global decarbonization. Read More

Climate Change
Solar Power’s Paradoxical 2015 in Three Charts
This post is co-written with Sagatom Saha, research associate for energy and foreign policy at the Council on Foreign Relations. In his final State of the Union address, President Obama celebrated the remarkable growth of clean energy, particularly solar power, which in 2015 added 7.4 GW of capacity in the United States and 55 GW globally. However, he also omitted an equally remarkable trend: over the same year, the Global Solar Index, which tracks the overall industry, collapsed, losing nearly half its value from a mid-year high. The divergence between solar market growth and solar industry market value began in July, perplexing analysts and investors. By December, company stocks may have finally bottomed out after news broke that Congress had extended generous U.S. solar tax credits for five years. Despite record global installed capacity and falling costs, solar companies either stagnated or fell in value in 2015 depending on the measuring stick applied. What explains solar’s paradoxical year? The three charts below demonstrate how falling crude oil prices and overzealous expansion combined to undermine solar companies. For an in-depth look at other factors—like the glut of public stock offerings and fears about interest rates—that affected renewable energy Yieldcos in particular, see my piece in Fortune. Since the end of 2015, SunEdison’s stock has fallen even further, losing approximately half of its market capitalization as of mid-January 2016.
Climate Change
Now Comes the Hard Part: India’s Scope for Emissions Mitigation
This guest post is co-authored by Joshua Busby, Associate Professor, and Sarang Shidore, Consultant and Visiting Scholar, at the LBJ School of Public Affairs at the University of Texas at Austin. For further analysis from the blog, see: "How India Could Achieve Its Audacious Solar Ambitions" The Paris climate negotiations produced an agreement that was satisfactory to all the major parties, including India. While much has been made of its negotiating position during the climate negotiations, less analysis has been dedicated to the implementation challenges going forward for the ambitious targets in India’s Intended Nationally Determined Contribution (INDC), namely the commitments to improve emissions intensity and scaling up of non-fossil energy. In addition to the generic challenge of affordable energy storage for intermittent renewable power, there are four India-specific challenges related to solar scale-up including (1) viability of the current bidding process, (2) the major challenges of grid integration, (3) the persistent financial crisis of distribution companies, and (4) multiple barriers to rooftop solar roll-out. India’s Emissions Profile and Trajectory Though its historic contribution to greenhouse gases is small, India is the fourth largest emitter of carbon dioxide, responsible for 7 percent of global emissions. However, India’s growth potential is enormous. India has more than 240 million people without access to electricity. Much of the rest of the country has intermittent power. As a consequence, the country’s energy use (and emissions) per capita are the lowest of all major economies and will inevitably grow as the country gets richer and people acquire more access to energy. According to the International Energy Agency, under current policies, India’s electricity generation will increase by 250 percent by 2040. In 2013, more than 70 percent of India’s electricity was generated by coal. Under business as usual policies, that would only reduce slightly by 2040, but in the post-INDC world the anticipated solar scale-up would reduce this number very substantially to about 50 percent of net generation. In its INDC, India committed to reduce its emissions intensity by 33 to 35 percent below 2005 levels by 2030. India also announced its intent to increase the non-fossil share of the country’s electricity to 40 percent by 2030, with an explicit commitment to scale-up wind to 60 GW and solar power generation to 100 GW by 2022, split nearly evenly between large-scale solar parks (60GW) and rooftop solar (40GW). With these more aggressive mitigation strategies, India could reduce that increase in carbon dioxide emissions by as much as 57 percent compared to the baseline. However, that would require India overcome many of the obstacles to implementation of its INDC targets. The Obstacles Solar scale-up is the principal hope for non-fossil energy. A significant portion of the wind potential has been exploited, and nuclear energy and large hydropower have stalled in the face of high costs and public opposition. The partial good news is that land acquisition, which has historically been a major challenge for Indian infrastructure projects, may be less of an obstacle this time around with the concept of solar parks – such as Charanka in western India – taking off. But key obstacles remain. Solar Bidding: How Low Can You Go? The Indian government has adopted the lowest-bid model for solar bids. This is due to the drive for greater transparency in the wake of the major scandals on coal blocks and telecom bandwidths during the previous government. However, companies that are making these bids may not be able to generate power at the price they have offered on a sustainable basis. There is precedent for problematic bids in India in the power sector. Nearly a decade back, India adopted a similar model for building large coal-power plants known as UMPPs (ultra mega power projects) of 4 GW capacity. The low bids these attracted made electricity rates unviable, especially when imported coal costs rose in the later part of the decade, and the contractual design had no allowance for passing on these costs to consumers. If the current solar bids result in electricity rates that cannot be profitable for the generators, then the anticipated solar scale up would fail. Fuel costs are zero for solar projects, but cost escalations routinely occur in Indian infrastructure projects. Part of the blame also lies with the private sector which submitted unrealistically low bids, hoping for upward adjustments in due course, thus inviting accusations that it has acted in bad faith. Most infamously, Reliance Power won two UMPP projects with extraordinarily low bids of Rs. 1.77 and Rs. 1.196 (approximately $0.039 cents and $0.026 cents per kWhr respectively at the then-prevailing exchange rates) for its projects in Tilaiya and Sasan in central India. The courts subsequently detected irregularities and invalidated a part of the contract terms for Sasan triggering a company lawsuit, while Reliance abandoned the Tilaiya project accusing the regional government of delays.  Reliance has also halted construction for the third UMPP in Krishnapatnam in southern India while challenging the bid rates with the central regulator. The final UMPP contract, won by Tata Power in Mundra in western India, is also mired in a legal dispute over the original bid rates, which Tata wants increased. With solar bids recently plunging below Rs. 5 per unit, a possible repeat of the UMPP failures looms on the horizon. The DISCOM Crisis: Who’s going to buy all this power? Tied to flaws in the contractual process is the financial ill-health of India’s mostly state-owned distribution companies, commonly known as DISCOMS, who are in debt to the tune of $66 billion. DISCOMS tend to under-buy power to reduce losses, a major barrier to large new additions of solar capacity. The Indian government has unveiled a rescue package for DISCOMS, but some analysts are still not convinced this will solve the problem. The Grid: Will the power flow? Another challenge is the inadequacy of the existing grid to be able to move renewable power from sites of generation to sites of consumption. This is already hampering scale-up in wind – for example, the southern state of Tamil Nadu wastes a substantial fraction of its generated wind power due to grid challenges. Moreover, the amount of electricity lost during transmission is still too high and needs to be reduced substantially. Although the Indian government has announced  spending of nearly $16 billion for this purpose, much more will be needed. Rooftop Stall: Why can’t I generate my own power? 40 GW of the 100 GW solar target is slated for rooftop solutions. At one level, this is a no-brainer from the economic standpoint because commercial and industrial users pay electricity rates in India that are above costs of solar generation. However, net metering, the technological and policy framework that is required for rooftop solutions to work, has not yet been implemented in a majority of Indian states. The financing environment for rooftop installations is far from being in place. Finally there is significant resistance from India’s largely state-owned distribution companies, many of which are deep in the red, towards widespread net metering - for the entirely legitimate reason that they will lose their best-paying customers. A win-win revenue model that does not penalize these companies is essential for net rooftop solar to be politically viable. Conclusion In sum, India is going to build significant solar generation capacity in the next few years, but unless these obstacles are overcome, that capacity may only be a fraction of the intended 100 GW target in the country’s INDC.
Economics
Budget Deal Oil-for-Renewables Trade Would Substantially Reduce Carbon Emissions
This post is coauthored by Varun Sivaram and Michael Levi. Congress is set to vote on a budget deal that would permanently end the long-standing ban on crude oil exports in exchange for temporary extensions of tax credits that support solar and wind energy. Michael wrote on Tuesday about the market, climate, and geopolitical impact of lifting the oil export ban. In this post we’re going to estimate the climate impact of the renewables tax credit extensions. We focus on 2016-2020 for three reasons: (a) it’s the period for which we have the best data; (b) beyond 2020, complex interactions with the Clean Power Plan make things much tougher to model; and (c) most important, beyond 2020, the primary effect of the ITC/PTC extension should be to make reducing emissions cheaper, and thus enable stronger policy, something that can’t be quantitatively modeled. Our bottom line: Extension of the tax credits will do far more to reduce carbon dioxide emissions over the next five years than lifting the export ban will do to increase them. While this post offers no judgement of the budget deal as a whole, the deal, if passed, looks like a win for climate. What the Budget Deal Includes The tax credit extensions would be a big deal for the renewable energy industry. The solar investment tax credit (ITC) is especially lucrative—new solar installations that begin operating before 2020 will continue to receive a tax credit equal to 30 percent of their system cost. The ITC steps down to 10 percent by the end of 2021, but projects that commence construction in 2020 and 2021 are still eligible for 26 and 22 percent tax credits, respectively. Given that solar industry leaders like SolarCity and First Solar have tailored their business models to withstand the impending ITC cliff (without this deal, the ITC would plunge down to 10 percent for any project completed after 2016), the six-year extension/phasedown is an unexpected Christmas present. The revival of the currently expired PTC is also a welcome development for the wind industry. The PTC—which compensates wind generators for ten years after they begin operating for the power they produce—would return to its full value of 2.3 cents per kilowatt-hour (kWh) for any project under construction by the end of 2016. The proposed budget bill articulates a five-year phase-out (1.84 cents/kWh for projects commencing construction in 2017, 1.38 cents/kWh in 2018, 0.92 cents/kWh in 2019, and nothing thereafter) that gives the industry visibility into the future. This is important because over the last decade, the wind industry has been plagued by boom and bust cycles driven by uncertainty over the future of the PTC. Even though the PTC would phase out faster than the ITC, then, the wind industry arguably needs policy stability as much as policy support. That should make the PTC deal a welcome development for the wind industry. Impact of the Budget Deal on Wind and Solar Deployment The first step toward estimating climate benefits is to project the effect of the new tax credit policies on renewable energy adoption. GTM Research published a helpful research note projecting solar adoption through 2020 with and without the proposed ITC extension. The contrast is stark—whereas installed solar capacity was set to peak in 2016 and then plunge over a cliff as the ITC expired, under the budget proposal there is a smaller pause in solar deployment in 2017 as the glut of projects in the current pipeline get built. From 2017 onward, all three industry segments—residential, commercial, and utility-scale solar—grow faster than without the ITC extension. This leads to around 25 gigawatts (GW) of additional capacity under the budget proposal by 2020. To estimate the impact of the PTC extension, we used analysis released this week by Bloomberg New Energy Finance comparing wind deployment under the proposed PTC phase-out to deployment without any PTC support at all for the next five years. Again, there is a stark contrast between the two projections. Without the extension, wind deployment is projected to peak in 2016, as developers rush to take advantage of the Internal Revenue Service’s determination that projects operational before 2017 will be eligible for the full 2.3 cent/kWh PTC that expired in 2014. Now, under the budget proposal, the steep cliff in capacity coming online in 2017 is replaced by a gentle hill and then a flurry of new construction to take advantage of the PTC extension. By 2020, around 19 GW of incremental wind capacity is projected to come online because of the budget proposal. The resulting projections are displayed in Figure 1. The two panels at left show the cumulative installed capacity of solar and wind in a world without tax credits and under the budget proposal. The middle panel plots the annual capacity that is incremental to the budget proposal—that is, all new solar and wind excluding projects that would be built anyway without the budget proposal. The rightmost panel shows the cumulative capacity additions due to the new policy. Two trends are important to note. Both incremental solar and wind deployment are actually negative in 2016, reflecting the forecast that under an extended ITC and PTC, there would no longer be a mad rush to build projects before a 2016 cliff. However, solar and wind deployment trends diverge in later years. Through 2020 incremental solar construction accelerates as solar becomes cheaper while the tax credit remains at 30 percent. But incremental wind deployment peaks in 2017–2018, because developers that begin construction on projects in 2016 will be eligible for the full value of the PTC (even if the projects are only operational in subsequent years). Then as the PTC phases out to zero in 2020, additional wind capacity spurred by the budget deal will decline, so by 2020 the additions under the budget proposal are roughly similar to the additions without any tax credit extensions. Emissions Impact We can use these projections of incremental capacity additions to estimate the climate benefits of the new renewable energy. We conservatively focus on the 2016–2020 period before Clean Power Plan incentives kick in fully; this means that we’re going to underestimate the climate benefits of the ITC/PTC extension. Figure 2 details the assumptions in this calculation. First, we assume that new solar and wind will on average displace a mix of fossil-fuels—coal and natural gas—as well as nuclear power in some cases. (All other sources have zero marginal cost and therefore won’t be displaced.) Nuclear power will be displaced in big chunks (if it is displaced at all), corresponding to retirements of entire plants, because as a baseload, low marginal cost resource, nuclear plants either run at near-full capacity or not at all (and some argue that zero-marginal cost renewable energy like wind and solar can make a nuclear plant’s operation unprofitable enough that it may have to shut down). To be safe, we investigated a range of values of the carbon intensity—or the emissions per unit power—of the electricity sources that solar and wind displace. At the low end, we considered the emissions intensity of a mix of nuclear, coal, and gas, weighted by their generation. And at the high end, we assumed that no nuclear reactors close because of renewable energy and instead stipulated that renewable energy displaces a mix of coal and natural gas, again weighted by generation. Second, we assume that since the solar and wind plants that will be incentivized by the tax credit extensions will be new projects, they will have high capacity factors—that is, they will be relatively efficient at generating power compared with older counterparts. New utility-scale solar plants now boast capacity factors of 30 percent. New residential solar installations deliver half that. New wind plants perform at a capacity factor of around 37 percent. Third, to be conservative, we consider an effect of up to 10 percent additional emissions from integrating renewable energy into the grid. Because renewable energy is intermittent—that is, solar and wind only produce when the sun shines and the wind blows—the rest of the power plant fleet must compensate for this added unpredictability. This leads to natural gas plants changing their power output rapidly, which reduces their efficiency and requires them to emit more CO2 per unit of generated electricity. Moreover, more power plants may have to be kept running on standby as “reserve margin” to compensate for any unanticipated shortage of renewable energy. Our final assumption deals with an early compliance mechanism for the Clean Power Plan, the Clean Energy Incentive Program, which kicks in from 2020 onward. Under the proposed program, which has not yet been finalized, states can claim credits for renewable energy projects commencing construction after September 2018, and they can use these credits to avoid equivalent emissions cuts from 2022 onward when the Clean Power Plan takes full effect. This suggests that any savings in emissions from renewable energy that gets built thanks to the tax credit extensions may result in a future emissions increase because states will have additional emissions headroom to comply with the Clean Power Plan. Because this early compliance mechanism has yet to be finalized, we excluded the effect of future offsetting emissions from our central 2020 estimate of CO2 reductions; we do, however, include it in assessing the full range of possible impacts. To express the sensitivity of our central emissions estimates to the assumptions outlined above, we have added uncertainty ranges to the bars in in Figure 2 to indicate uncertainty. The pronounced uncertainty range in 2020 reflects the open question of whether emissions saved by renewable energy will be offset by future Clean Power Plan compliance headroom. The take-home point from this figure is that the emissions reductions from solar and wind energy through 2020 is substantial, reaching as much as 90 million metric tons of avoided CO2 per year in 2020. The average annual emissions reduction over the 2016-2020 period is 25-46 million metric tons with a most likely value around 40 million metric tons. For reference, the Obama administration’s Clean Power Plan is projected to reduce CO2 emissions by about six times that level, or 240 million metric tons per year, in 2025. Putting It All Together: Renewable Energy Climate Impact Overwhelms That of Oil Exports In contrast to the considerable emissions savings from renewable energy, the climate impact of lifting the crude oil export ban is likely to be small. A previous post estimated the average annual emissions impact of lifting the oil export ban as around 10 million metric tons of CO2 per year over 2016-2025 (with a possible range of 0–20). Over the time period we have examined in this post, 2016–2020, the same methodology yields an estimate of 2 million metric tons of CO2 annually (with a possible range of 0–5). Figure 3 extends the previous figure by adding the range of positive emissions from crude oil exports to the emissions savings from new renewable energy incentivized by the budget deal.  The diamonds represent the central estimates of the net emissions impact of oil exports and new renewable energy, and the dotted bars represent the uncertainty range of how much oil exports could increase emissions. The net impact of the exports-for-renewables-credits trade, then, is to reduce carbon dioxide emissions by at least 20-40 million metric tons annual over the 2016-2020 period. The most likely emissions reduction in our estimate is around 35 million metric tons. The climate benefit of the tax credit extension is over a factor of ten larger than the climate cost of removing the oil export ban over this period. What About the Longer Run? This of course does not answer the question of what will happen over the longer run. The impact of lifting the oil export ban will persist while the ITC/PTC will be phased out. One could, in principle, extend the analysis above through 2025. (A very simple extension of the central estimates through 2025, assuming no emissions savings from the PTC/ITC after 2020, leaves one with an ITC/PTC impact considerably outweighing that of oil exports.) This would, however, be misleading. Extension of the PTC/ITC should drive down zero-carbon energy costs and reduce business as usual emissions beyond 2020. Both factors should enable stronger rules under the Clean Power Plan (or other policies that are additional to the plan). This is part of the potential payoff of an ITC/PTC extension. This can’t, of course, be modeled quantitatively. But it is the right way to think about the longer run impact of the budget deal.
  • Fossil Fuels
    Oil Exports Budget Deal? Market, Climate, and Geopolitical Consequences
    News outlets are reporting that a congressional budget deal could end the oil export ban in exchange for extension of the Investment and Production Tax Credits (ITC and PTC) that support solar and wind energy. Here I want to lay out what ending the oil export ban could mean for markets, climate, and geopolitics. (I suspect Varun may weigh in later on the ITC/PTC extensions if and when details emerge.) Short version: Little immediate impact on anything; a possible boost on the order of a few hundred thousand barrels a day to U.S. oil production over the longer run; a factor of perhaps fifty smaller impact on carbon dioxide emissions than the Clean Power Plan and CAFE (fuel economy standards); and a mixed bag for geopolitics and trade talks. Markets There is currently little if any incentive for U.S. oil producers to export crude oil even if the ban is lifted. Light sweet crude oil for January delivery in Northwest Europe (Brent) – the destination most commonly envisioned for U.S. crude oil exports – is currently selling for less than similar oil delivered on the Louisiana coast (LLS). It costs in the neighborhood of three dollars a barrel to ship oil from the U.S. Gulf coast to Europe. Spending three dollars in order to lose money is not something sane people do. (The spread could, of course, change, particularly during refinery turnarounds.) The result is that large sustained oil export volumes are unlikely to materialize soon. This should remain unchanged until U.S. oil production recovers to substantially above previous highs. U.S. refineries were previously able to accommodate U.S. production despite the export ban; they will presumably continue to be able to unless and until production rises above historic highs. And then what? To get a sense of the possible longer term impact of lifting the ban, take a look at the Energy Information Administration (EIA) modeling of what would happen to markets if the ban were lifted. In its reference case, which is its best guess of reality, lifting the ban has no impact on U.S. production or prices because U.S. refineries can absorb all the oil that the United States produces. In its high resources case, where U.S. oil resources turn out to be considerably more plentiful than the EIA currently estimates, it estimates that U.S. production would be 220 thousand barrels a day higher, on average, between 2016 and 2025, without the oil export ban than with it. The EIA modeling also projects that world oil prices (Brent) would drop by roughly a dollar on average over the next decade, while domestic oil prices (WTI) would rise by about four dollars on average over the same period, if resources were abundant and the ban were lifted. The EIA is of course fallible. But these numbers comport well with basic intuition. Were the ban to remain in place, U.S. oil could still make it to world markets if demand were there; it would just need to be refined and exported as product (gasoline, diesel, etc.), which is already legal. Refining light sweet crude oil costs a few dollars a barrel. Domestic prices would therefore need to be a few dollars lower with the ban than without it in order to make the economics of exports work. It’s tough to see how a few dollar difference in oil prices would alter domestic oil production by more than a few hundred thousand barrels a day. (The studies that claim much larger consequence for oil production assume that domestic prices can detach from international ones by twenty dollars a barrel or more without prompting any investment in refining capacity; I find that sort of departure from rational profit-seeking incredibly unlikely.) And an additional few hundred thousand barrels a day on the global market shouldn’t lower world prices by more than something on the order of a dollar a barrel. That’s basically the same as what the EIA model says. Climate All of this has consequences for climate change too. Let’s use the EIA modeling again. In the reference case, lifting the ban has essentially no impact on production, prices, or actual exports. That means it has no impact on carbon dioxide emissions either. What about in the high resource case? The EIA estimates that U.S. production rises by an average of 220 thousand barrels a day over the next decade. This will be offset in part by lower production (as a result of lower prices) elsewhere. It’s reasonable to assume that the offset is on the order of 50 percent of the increase in U.S. production, which leaves us with a net increase of about 110 thousand barrels a day, or about 40 million barrels a year. If you figure a barrel of oil, when produced, refined, transported, and consumed, generates about half a ton of carbon dioxide emissions, this works out to about 20 million tons of additional carbon dioxide emissions annually. The number could be lower if the international production response was stronger than I’ve assumed. To put this all together, assume that there are 50/50 odds of being in the reference case world or the high oil resource world. (Alternatively, one could imagine that we’re in the no-exports world for five years because of weak global demand, and then in a modest-exports world after that.) Then, on average, we’d expect 10 million tons a year of additional carbon dioxide emissions on average over the next decade. Now put that in context: Ten million tons is roughly 0.2 percent of annual U.S. emissions. The Clean Power Plan (PDF) is estimated to reduce U.S. emissions by about 240 million tons a year by 2025 (and about double that by 2030). The fuel economy standards (CAFE) are estimated to reduce U.S. emissions by about 320 million tons a year by 2025 (and by about 170 million tons a year on average between 2016 and 2025). These policies dwarf the impact on carbon emissions of allowing oil exports. Geopolitics What about the impact on geopolitics? Allowing oil exports shouldn’t hit world oil prices much. The price collapse over the last year has far bigger consequences for oil exporters’ budgets than removing the oil export ban would. Allowing oil exports would, however, marginally strengthen the U.S. position in arguing for more liberal energy trade worldwide. It also would be welcomed by allies who believe (usually without justification) that access to U.S. oil will improve their national security. On the flipside, several European countries were hoping to gain access to U.S. oil exports though a Transatlantic Trade and Investment Partnership (TTIP). This prospect would have given U.S. negotiators some leverage to get other concessions in return. Allowing oil exports now would retire that card. Bottom Line An oil-exports-for-renewables-tax-credits deal looks likely to be a win-win. (I’ll reserve final judgment pending details of an ITC/PTC extension.) Removing the oil export ban is good policy. Supporting zero-carbon energy innovation, including through appropriate deployment subsidies, is good policy. Readers of this blog know that I’ve been arguing for a “most of the above” energy and climate strategy that supports some fossil fuels and some zero-carbon energy for several years. Congress has been too deadlocked to make the sorts of deals such a strategy prescribes. But the rumored budget deal looks like it would fit the “most of the above” approach nicely. [Post-publication update: The folks at Oil Change International reasonably point out that I’m considering only the supply side market response. So here’s a slightly more full-blown analysis. Let’s assume that oil supply elasticity is 0.5 and demand elasticity is 0.2. Then supply side policies get hit with a rebound of around 70% and demand side ones suffer about 30%. The impact of lifting the oil export ban becomes 0-15 million tons annually with a central estimate around 7 million tons. The impact of CAFE becomes 120 million tons annually. The impact of CPP is unchanged at 240 million tons in 2025 (leakage through coal markets has been small so far; leakage through gas markets is ambiguous, and could actually reduce world emissions in some cases). One still finds roughly a 50:1 ratio of impacts between the policies cited.]
  • Europe and Eurasia
    Lessons in Cleantech Success from Scandinavia (Pt. 2): The Importance of the Danish Manufacturing Revival
    This post is co-written by Ben Armstrong and Varun Sivaram. Ben is a Ph.D. Candidate at MIT focused on Political Economy and a researcher at the MIT Governance Lab. In Part 1 of this series, we posed a puzzle: why has Denmark had more success at clean tech innovation than its neighbor, Sweden? Neither demand-pull conditions, which provide a sales environment that invites innovation, nor technology-push factors, which directly support technology research, development, and demonstration, appears to favor Denmark over Sweden. Both countries have similar environmental policies and environmentally conscious populaces, and Sweden has actually been more successful than Denmark in inducing other forms of innovation, especially in information and communications technology (ICT). But Denmark leads by a substantial margin in patents for climate change mitigation and the commercialization of eco-friendly technology. What explains Denmark’s outperformance in cleantech? As the nations of the world convened in Paris last week to conclude an accord on climate change, cleantech innovation emerged as a priority. Bill Gates and 27 other billionaire investors pledged to ramp up cleantech funding for emerging technologies, and 20 countries including the United States, India, and China signed on to the “Mission Innovation” pledge to double their funding for cleantech research. As many begin chasing after capital, it becomes even more paramount for countries to figure out how to provide the best environment for a cleantech sector to thrive. Denmark’s manufacturing revival led its success in cleantech We argue that Denmark’s cleantech development has been led by legacy manufacturing companies that were founded decades before “cleantech” was even discussed. These corporate actors, who pivoted toward cleantech during a period of economic crisis, are the missing puzzle piece; when paired with government demand-pull and (more importantly) technology-push policies, they transformed into globally competitive cleantech firms. And this phenomenon could recur elsewhere—legacy manufacturing regions in the U.S. Rust Belt could become cleantech hubs. The lesson is that cleantech hubs require far different ingredients than those required to develop a software-led innovation economy. Denmark’s manufacturing firms began by producing more traditional technology for energy, construction, and home appliances. But when the Danish economy struggled with high unemployment and the energy crisis of the 1970s, these companies began investing in new products that promoted alternative energy sources and increased energy efficiency. The emergence of the Danish cleantech economy was the product of repurposing from large, old companies that needed to innovate to stay relevant. These companies drew on existing manufacturing infrastructure and worker retraining programs (now part of the government’s “Flexicurity” program) to upgrade their businesses. Five companies listed among Denmark’s top cleantech innovators follow this pattern. Novozymes began in the 1920s as a biotechnology company focused on detergents, but early investments in clean energy research in the 1980s helped one branch of the business transition to focus on producing biofuels. Grundfos began its business in the early 20th century producing electric pumps for water wells. It transitioned in the 1980s to conduct research on sensors for improving pump efficiency and solar technology that would power its pumps. Rockwool started in the early 20th century as a company manufacturing tiles and digging for coal. Wool was a minor part of the business that grew until the oil crisis in the 1970s when Rockwool expanded its insulation business to reduce heating costs. Today it provides stone wool products for insulation that improve energy efficiency. Danfoss began building small valves to control heating and cooling systems in the 1930s. Over time, they re-invested in creating electrical control technologies that help improve the efficiency of power systems. Danfoss provides a particularly clear case of repurposing. While the rest of the economy was experiencing crisis during the 1970s, Danfoss expanded its operations to factories that had been vacated by textile companies and a bankrupt telephone manufacturer. In one case, they retrained the workers previously employed in textiles and began expanding the production of coils, relays, and frequency converters to improve energy efficiency. Vestas is the leading provider of wind turbines in Denmark and has become a global leader in wind power. It was founded by blacksmiths in the early 20th century, then transitioned to produce household appliances in the 1940s, agricultural equipment in the 1950s, and was most successful with the production of hydraulic cranes in the 1960s. It was only when the hydraulic crane business took off that Vestas began investing in wind technology in the late 1970s. During its initial roll-out of wind technology, one of the turbines broke and the company almost went bankrupt due to uncertain quality control. Vestas later recovered by bringing the end-to-end production of the wind power system in-house. Today it exports much of its wind technology to establish on-shore and off-shore wind farms around the world. The same cleantech innovation probably did not happen in Sweden because large Swedish businesses did not experience the same pressure; unemployment stayed low through the 1970s, and Sweden’s energy giant Vatenfall made an early investment in nuclear energy in the 1960s that locked Swedish power generators into a particular energy path. When Sweden faced its crisis in the early 1990s, it implemented a sweeping transition into services that gave birth to its more recent success in ICT. It makes sense, then, that manufacturing in Copenhagen is nearly twice as large as a portion of the economy as it is in Stockholm. Advanced manufacturing is four times as prevalent in Copenhagen (again as a proportion of overall employment). Denmark has outperformed Sweden in Cleantech by repurposing legacy manufacturing. Sweden’s large diversified firms include Electrolux, which produces home appliances like Vestas did, but continued along the same product trajectory for decades without changing course to focus on cleantech. Atlas Copco is a large Swedish producer of power tools that continued to upgrade and produce the same types of products over the course of the 20th Century. Sweden’s largest energy providers include Vattenfall, mentioned above as an early investor in nuclear, which has since diversified into wind and solar—as well as E. On Sverige, which is a subsidiary of the German energy conglomerate E. On. Although there are numerous startups in Sweden investing in improving its renewable energy portfolio and upgrading existing clean technologies, none of them has had the impact that any one of these large Swedish producers might have had if they diversified and invested in clean technology innovation. Your turn: Most readers support a demand-pull solution to the puzzle In the previous post, we invited reader comments, and below we’ll feature some of your explanations for Denmark’s comparative success. In contrast to our manufacturing revival theory, several readers described a more straightforward demand-pull story to explain Denmark’s comparative success over Sweden in cleantech and in wind energy in particular. David B. Benson notes that the wind from the Atlantic Ocean not only keeps Denmark warm but also “goes a long way towards explaining the Danish interest in wind derived power.” Similarly, Anonymous argues that Denmark is “windier than Sweden, which is more protected [from ocean winds].” And since “Denmark is right next to massive energy user Germany, which…if happy to buy the power from Denmark,” wind power in Denmark can serve both domestic and foreign markets. Ola proposes that higher electricity prices in Denmark compared with Sweden made renewable energy more competitive in Denmark. Geoff Dabelko suggests that Denmark’s choice to forego domestic nuclear power freed up resources for “state subsidies for renewable energy internally.” Echoing this policy argument, Fionn Rogan contends that favorable public opinion in Denmark has supported aggressive domestic procurement of clean energy technologies “so that the wind turbine that’s invented and developed in Denmark gets bought in Denmark too. At least initially, until there’s enough of an export market to take it to the next growth stage.” Indeed, Denmark is a leader in wind energy, and on some days the 4.8 GW of installed wind capacity in Denmark account for over 100 percent of Danish electricity consumption (the surplus is exported to neighbors). But the domestic market for wind turbines does not neatly explain the rise of Denmark’s globally competitive wind industry—not even initially. In particular, the ascent of Danish wind turbine manufacturer Vestas, which has the largest share of the global wind market, depended on its international expansion much more than its domestic sales. Vestas’ first major order came from California in 1985, when its transition to producing wind turbines rather than agricultural machinery was incipient. By the early 1990s, through exports, joint ventures, acquisitions, and subsidiaries around the world, Vestas was selling turbines to the United States, the United Kingdom, Germany, Spain, Australia, New Zealand, and—yes—to Sweden. Although the Danish government did support the deployment of wind at home—from 1981 to 2000, Denmark installed 1.4 GW of wind thanks largely to domestic mandates—the domestic market always accounted for a minority of Vestas’ sales (for reference, Vestas sold around 1.3 GW in 2000 alone, a third of the global market). And domestic policy support was far from unwavering in creating demand for wind turbines. The steady progress that Denmark made in sustaining a growing domestic market from 1981 to 2000 was abruptly undone by the incoming Conservative Party government, which slashed support from 2001 to 2007. Nevertheless, companies like Vestas continued to expand internationally, despite waning support at home; this suggests that its ability to capitalize on favorable policy toward wind abroad, in much bigger markets than Denmark (like the United States, Spain, and Germany), drove its success. Our favorite reader comment: Cleantech software The only technology-push explanation came from Graham Pugh (which is fitting since he was formerly a director at the Department of Energy, which implements technology-push policy in the United States). Graham “would posit that a combination of support for hardware startups from universities like DTU [Danish Technological University] and perhaps incubators has made the difference.” Moreover, he argues, “I think hardware IS different and requires different strategies [from supporting software].” We agree with this view, and we concede that in addition to our manufacturing revival theory, targeted government support for hardware-specific innovation in Denmark could help explain why Denmark has succeeded in cleantech. Importantly, we do not contend that Denmark is more innovative than Sweden. After all, Sweden benefitted from the success of high-growth software companies like Ericsson and Spotify and successful global producers like IKEA. Rather, both Denmark and Sweden boast innovative economies, but Graham pushes us to disaggregate what we mean by innovation—success in cleantech looks very different from success in other sectors. And the surprising conclusion from the Danish example is that the intuitive demand-pull and technology-push levers that policymakers have to support the cleantech sector may not suffice to make it globally competitive. Danish environmental regulations, labor market reforms that improved job retraining, and early-stage research support for hardware innovation helped to varying degrees. But the crucial factor that differentiates Denmark from neighboring Sweden is that Danish business captured an opportunity during crisis to reorient their business to focus on an emerging industry. And it paid off. So, who’s next?