Energy and Environmental Technology

Developing new and improved clean-energy technologies is an important part of any strategy to combat global climate change. For example, generation of electricity and heat is the largest source of carbon emissions, accounting for 42 percent of carbon emissions worldwide in 2012.¹ Meeting the climate policy goals currently under consideration, such as European Union discussions to reduce emissions by 40 percent below 1990 levels by 2030 or the U.S. Clean Power Plan goal of reducing emissions from the electricity sector by 32 percent by 2030, will not be possible without replacing much of the current fossil fuels-based electric generating capacity with alternative, carbon-free energy sources.

My research focuses on the role of technology for both reducing energy consumption and providing clean energy. This work includes three main themes: empirical studies of the relationship between environmental policy and innovation, policy simulations and empirical work on ways environmental and science policies may promote energy innovation, and empirical studies of environmental technology transfer. Much of my research uses patent data to track energy innovation, thereby building on the pioneering efforts of NBER researchers such as Adam Jaffe and Bronwyn Hall, whose early forays into patent data made these data accessible to a new generation of researchers.²

Empirical Studies of Induced Innovation

My empirical work on policy-induced technological change seeks to understand how policy affects the development of new environmentally-friendly technologies. I use patent data to track changes in environmental technologies, such as pollution control devices, alternative energy sources, and technologies designed to improve energy efficiency. With this research, I aim to better inform researchers who simulate the effects of long-term policies such as climate change policy and to contribute to the broader discussion of environmental policy design.

Early work on energy innovation focused on the link between energy prices and innovation. In a 2002 paper, I use patent data to identify innovation on 11 different alternative energy and energy efficiency technologies.³ In the long run, a 10 percent increase in energy prices leads to a 3.5 percent rise in the number of energy patents. Most of the response occurs quickly after a change in energy prices, with a mean lag response time between energy prices and patenting activity of 3.71 years. My estimates controlled for the quality of knowledge available to an inventor as well as other factors influencing R&D, such as government support for energy research and technology-specific demand shifters.

Subsequent work turned attention to the incentives offered by various policy instruments, showing that the types of incentives matter. In a 2003 paper, I combine plant-level data on flue gas desulfurization (FGD) units installed at U.S. coal-fired power plants with patents pertaining to FGD devices to assess the impact of innovation before and after the 1990 Clean Air Act (CAA),⁴ which instituted permit trading for sulfur dioxide (SO2). Before this act, new plants were required to install flue gas desulfurization capacity capable of removing 90 percent of SO2. As a result, the innovations that occurred before the 1990 CAA focused on reducing the cost of FGD units, rather than on improving their environmental performance. After passage of the act, the focus of innovation became improving the ability of FGD units to remove SO2 from a plant's emissions.

While economists often favor using broad-based policies, such as a carbon tax or tradable permits, to address externalities, policy makers often use more narrowly focused options. In renewable energy, popular options include feed-in tariffs, in which governments guarantee a fixed price above prevailing market prices for energy from renewable sources, and renewable portfolio standards that require a minimum percentage of electricity be generated using renewable sources. While renewable portfolio standards leave it to market forces to decide which renewable sources are used to meet the target, feed-in tariffs may target specific energy sources. For example, at their peak, feed-in tariffs for solar energy in Germany were over seven times higher than the feed-in tariffs for wind energy.⁵

In a 2010 publication, Nick Johnstone, Ivan Haščič, and I collect data on renewable energy policies and patents across countries to assess the effect of various renewable energy policies on innovation.⁶ Figure 1 shows that patenting activity has increased rapidly over the past decade as these policies have become more prevalent.⁷ Moreover, different instruments end up promoting innovation on different types of renewable energy. Quantity-based policies, such as renewable portfolio standards, favor development of wind energy. Of the various alternative energy technologies, wind has the lowest cost and is closest to being competitive with traditional energy sources. As such, when faced with a mandate to provide alternative energy, firms focus their innovative efforts on the technology that is closest to market. In contrast, direct investment incentives are effective in supporting innovation in solar and waste-to-energy technologies, which are further from being competitive with traditional energy technologies and thus need the guaranteed revenue from a feed-in tariff to be competitive.

These results suggest particular challenges to policy makers who wish to encourage long-run innovation for technologies that have yet to near market competitiveness. Economists generally recommend using broad-based environmental policies, such as emission fees, and letting the market "pick winners." This leads to lower compliance costs in the short-run, as firms choose the most effective short-term strategy. But because firms will focus on those technologies closest to market, broad-based market policy incentives do not provide as much incentive for research on longer-term needs. There may be a complementary role for policies such as direct R&D subsidies to promote development of clean technologies further from the market.

The Roles of Environmental and Science Policy

Understanding the role of environmental policy on technological change involves the study of two market failures. Because pollution is not priced by the market, firms and consumers have no incentive to reduce emissions without policy intervention. Thus, the market for technologies that reduce emissions will be limited without policy interventions that alter these incentives. At the same time, the public goods nature of knowledge leads to spillovers that benefit the public as a whole, but not the innovator. As a result, potentially innovative private firms and individuals may not have incentives to provide the socially optimal level of research activity.

The evidence suggests that science policy plays a supporting role, but that environmental policies are most important for promoting new green technologies. Policies must be in place not only to encourage the development of cleaner technologies, but also to encourage the adoption of existing clean technologies. In a 2006 paper using ENTICE, a model of the global economy that links economic activity to carbon emissions and allows research in the energy sector to respond to policy changes,⁸ I compare long-run welfare gains from both an optimally-designed carbon tax (one equating the marginal benefits of carbon reductions with the marginal costs of such reductions) and optimally designed R&D subsidies.⁹ While combining both policies yields the largest welfare gain, a policy using only the carbon tax achieves 95 percent of the welfare gains of the combined policy. In contrast, a policy using only the optimal R&D subsidy attains just 11 percent of the welfare gains of the combined policy. This finding is confirmed by other researchers simulating U.S. and global energy policies, showing that policies directly targeting environmental damages from electricity generation better promote both emissions reductions and innovation.¹⁰ However, carbon prices and R&D subsidies can complement each other if clean technologies are less developed than existing dirty technologies. In such a case, initial R&D subsidies can close the gap between clean and dirty technologies, reducing the level of carbon taxes needed in future years to reduce greenhouse gas emissions.¹¹

To better understand the potential for future public energy R&D spending, in recent work, I use scientific publications to evaluate the effectiveness of public energy R&D expenditures.¹² Combining data on scientific publications for alternative energy technologies with data on government R&D support helps isolate the effect of public R&D and sheds light on the process through which public R&D helps develop scientific knowledge. Interestingly, unlike work on private sector innovation, other factors such as energy prices and policy have little effect on alternative energy publications. Thus, current government energy R&D efforts appear to support novel research, rather than crowding out work that would otherwise be done. I find little evidence for diminishing returns to energy R&D at current funding levels. However, patience is important for evaluating public investment in energy R&D. The ultimate goal of government energy R&D funding is not a publication, but rather a new technology. Thus I use citations these articles receive from future patents to assess the impact of basic science on new technologies. Figure 2 traces the time path of the increased citation probability for publications generated from an additional $1 million in R&D funding being cited by a patent. It may take up to a decade to realize the full effect of public energy R&D funding on publications, and even longer until these publications are cited in new energy patents. Because of the lags between initial funding and publication, there is little increase in the cumulative probability of a citation resulting from new R&D funding until approximately six years after the funding, with the effect not leveling out until almost 18 years afterwards. Allowing for a five year window for processing patents, this suggests that new patent applications citing these publications begin appearing about one year after funding and continue for 13 years.

International Technology Transfer

A third stream of my research focuses on the international dimensions of environmental technological change. This work began with a 2006 study of air pollution control equipment in the United States, Japan, and Germany.¹³ Whereas the United States was an early adopter of stringent sulfur dioxide (SO2) standards, both Japan and Germany introduced stringent nitrogen oxide (NOX) standards much earlier than the U.S. Using patent data from all three countries, I find that innovation responds to policy even in countries that adopt regulations late, suggesting that these countries do not simply take advantage of technologies "off the shelf" that have been developed elsewhere. Instead, late adopters often undertake adaptive R&D to fit previously developed technology to local markets. As evidence, I show that these later patents are more likely to cite earlier foreign rather than domestic inventions.

My more recent work on international environmental technological change explores how technology can help developing countries address environmental issues. As emerging and developing countries continue to grow, the environmental impact of their economies increases. Access to clean technologies may mitigate this impact. Mary Lovely and I flip the usual question of policy's effect on regulation around, asking instead how technology affects regulation.¹⁴ Because most pollution control technologies are first developed in industrialized countries, and because environmental regulations are needed to provide incentives to adopt these technologies, we focus on the adoption of environmental regulation as the first step in the international diffusion of environmental technologies. Using a hazard model, we study the adoption of environmental regulations for coal-fired power plants in a set of 39 developed and developing countries. While the adoption of pollution control technologies within a country responds quickly to environmental regulation, we find that adoption of the regulations themselves follows the typical S-shaped pattern noted in studies of technology diffusion. Access to technology is an important factor influencing regulatory adoption. As pollution control technologies improve, the costs of abatement, and thus the costs of adopting environmental regulation, fall. Thus, countries adopt environmental regulation at lower levels of per capita income over time. Moreover, countries that are more open to international trade have better access to these technologies, and are thus more likely to adopt regulation. While openness to world markets may also work against passage of environmental regulation by increasing competition and making it harder for local firms to pass along cost increases to consumers, we find that the access to technology effect dominates once the level of abatement technology reaches a critical level, which in our sample occurs during the early 1990s.

International climate agreements also foster access to technology. The Clean Development Mechanism (CDM) enables entities in developed countries to sponsor emission-reducing projects in developing countries. A secondary goal of the CDM is to help developing countries achieve sustainable development through the transfer of climate-friendly technologies from developed countries. If the technologies transferred via CDM projects lead to subsequent diffusions within the country, it will reduce the future abatement costs of carbon emissions and drive technological change in the energy sector of the recipient country.¹⁵ Tian Tang and I use data on wind turbine projects in China sponsored through the Clean Development Mechanism to ask whether these projects improve the technical capacity of wind projects in China.¹⁶ Using a learning curve model allowing for spatially correlated errors, we find that project costs decrease and project efficiency (measured by the capacity factor, which compares a wind farm's actual annual electricity generation to its potential annual output if the wind farm operates at its full capacity) increases with the previous experience of the project developer. The greatest efficiency gains come from repeated interactions between local project developers and foreign wind turbine manufacturers. That these improvements occur for the capacity factor as well as for cost reductions suggest that technology transfer occurs, and that the results are more than reduced transaction costs and lower contract prices for repeat customers.

Conclusion

While the papers cited here highlight the important connections between environmental policy and technological change, much work remains to fully understand the potential for technology to aid in both the mitigation of and adaptation to climate change. In addition to the research questions addressed here, the role of technology in climate change adaptation¹⁷ and the behavioral influences of clean technology adoption,¹⁸ are important areas for future work.