Green Innovation Needs Better Metrics

Climate change, pollution, and the economy’s sensitivity to energy prices all make continued environmental innovation a key priority. In a recent paper titled “Environmental Policy and Innovation: A Decade of Research,” David Popp of Syracuse University summarizes what economists have learned over the last 10 years about the link between environmental policy and innovation.

Popp is one of the leading scholars of the economics of environmental innovation and has written many of the most important papers in the field. He begins his latest summary of the literature by reviewing the market failures that public policy tries to correct. The first is that many activities generate environmental externalities. The cost of environmental damage is borne by the general population, so the firms undertaking the activity do not include them in their calculations and we get too much pollution. Second, environmental research generates positive spillovers. Because the firms conducting research do not capture all the benefits, they engage in too little of it. There is some evidence that the spillovers from clean innovation are larger than those for dirty technology, making research in this field especially valuable. Finally, imperfections in the market for knowledge can also inhibit the spread of the best technologies, delaying their diffusion into the markets.

However, scholars have trouble measuring the relationship between the myriad of possible environmental policies and their outcomes across many countries and industries. For lack of better metrics, they often focus on research spending and patents as key measures of the amount of innovation in an industry. Several studies have shown a clear connection between higher energy prices and induced innovation in different energy markets. One study showed that a 10 percent increase in fuel prices increased clean innovation by 10 percent and decreased innovation in dirty technologies by 6 percent. Another showed that, not only does heightened regulation increase innovation, the induced innovation also lowers the cost of complying with the stricter regulation.

Surprisingly, some studies show that, while stricter regulation has an effect on innovation in energy efficiency, research subsidies do not. But innovation need not be limited to regulated entities. It can also occur among energy users. Simply looking at the direct effect of innovation by regulated firms can underestimate the total amount of induced innovation in the broader market by 71 percent.

A key lesson is that environmental policy needs to be consistent and credible, so that companies can calculate the costs of not doing anything as well as the benefits of innovating. Consistent policy can also increase a nation’s innovative capacity and thereby enhance its ability to develop new innovations. Stringent regulation and strong patent protection attract R&D to a country. New innovation (as opposed to the importation of existing innovation) tends to be concentrated in those countries with the greatest capacity for innovation.

Another line of inquiry looks at the relative effectiveness of different environmental policies. Simply correcting prices for market failures is not enough. Clear performance standards are also important. However, policymakers must ensure that the benefit of these policies exceeds their cost. The structure of regulation also matters. One study shows that market-based policies consistently induce greater innovation, whereas non-market-based policies do not. Deregulation can therefore have a positive effect if it improves the overall performance of energy markets. Technology-“push” policies that encourage research and development often lead to both short-run inventions and longer-run cost reductions. Priced-based policies such as feed-in tariffs that guarantee a minimum price for new technologies also perform well in the long term since they encourage cost reductions over time.

In some cases, knowledge market failures may halt the spread of the best technology. These can include learning-by-doing, in which costs fall as producers learn how to manufacture and implement the technology, as well as capital-market failures such risk aversion. Learning-by-doing may justify subsidies to new users, thereby helping the technology get through the learning stage and hasten price decreases. In general, governments should supplement broad-based environmental policies with limited subsidies for those technologies furthest from the market.

With regard to failures in the capital markets, investors may shy away from risky new technologies even when the expected rewards are positive. The U.S. Small Business Innovation Research Program has been extremely effective as an alternative source of capital, reducing the need for private capital and doubling the probability of firm survival and successful exit. This may indicate that more research dollars should be targeted to smaller companies in the early stages of research. (A forthcoming ITIF paper will examine the evidence for focusing SBIR on companies with a potential for rapid growth.)

Finally, there is strong evidence that government research has a positive impact on innovation. However, because this research tends to be earlier stage, it often takes several years for the results to show. Research performed at government labs also plays an important role in spurring environmental innovation. There is also emerging evidence that collaborative research, which may also involve government resources, produces higher-quality output.

In summary, although research in this area is difficult, some firm results are emerging. Government policies to correct prices for market failure are important, but they are not enough. Clear guidelines on environmental standards are also needed, as is strong funding for both government and private research on emerging technologies.

Joe Kennedy is a senior fellow at the Information Technology and Innovation Foundation. He focuses on economic policy.

This article originally appeared in the ITIF’s Innovation Files blog. It is used with permission.