Strong link between high emissions taxes and investments in research and development
feb. 15, 2022
New publication in the Review of Financial Studies by SHoF researchers: Investigating the correlation between pollution taxes and R&D investment by polluting companies, experts from the US and Sweden say that rather than employing R&D expenditure to launch new innovation, such firms use it to improve their ability to use and assimilate external knowledge.
Higher taxes on SOx (sulphur dioxide) emissions are associated with a substantial increase in firm-level R&D spending. The pollution taxes have relatively stronger effects on R&D in sectors with dirtier production technologies, as expected if the relation between pollution taxes and R&D is causal. In contrast to R&D, pollution taxes do not lead to more patenting in high-pollution firms, suggesting that firms increase R&D to improve their ability to use and assimilate external knowledge rather than to develop new innovations. Consistent with this idea, the R&D response to pollution taxes is concentrated in sectors where external knowledge is easier to acquire. Overall, research findings suggest that investment in technological absorptive capacity is a first-order response when noninnovative firms with dirty production technologies face higher emissions taxes.
Policies that encourage firms to shift to cleaner production technologies can mitigate climate change risks and other environmental concerns without significantly slowing long-run economic growth. This potentiality has motivated a prominent literature to study how environmental policy affects the development of new “clean” products and technologies. But environmental policy can influence the path of technical change through another, less studied, mechanism: inducing polluting firms to make the new technology investments that enable them to fundamentally transform their production processes. Although several studies highlight the key role technology spending plays in firms’ efforts to re-engineer production and reduce pollution at the source, there is little systematic evidence that environmental policy can encourage polluters to make these investments.
“Technology investments that facilitate the transition to cleaner production have a higher marginal payoff to polluting firms when countries impose taxes on dirty emissions.”
This gap in the literature was filled by studying how country-level taxes on dirty manufacturing emissions affect technology spending in high-pollution firms. The main idea is that emissions taxes make it more expensive for polluting firms to continuing using their existing production technologies. The taxes thus incentivize polluting firms to make the investments that allow them to adopt and implement cleaner production processes. In this sense, the technology investments that facilitate the transition to cleaner production have a higher marginal payoff to polluting firms when countries impose taxes on dirty emissions.
To test this idea, an estimate was made of how cross-country differences in taxes on sulphur dioxide emissions affect firms’ investment in research and development (R&D). The focus was on SOx taxes because, in addition to SOx being a major air pollutant, the country-level variation in SOx taxes over time is considerable. In addition, information on SOx emissions at the industry level was sufficiently disaggregated to enable categorization of firms based on how heavily “treated” they are by higher pollution taxes. The fous was on R&D spending because R&D is the only widely available and internationally comparable measure of technology investment in use. Beyond this practical consideration, extensive evidence shows that R&D plays a central role in facilitating corporate efforts to overhaul production processes and reduce environmental impact. In particular, firms invest in R&D as a means to expand their capacity to use and absorb external knowledge about cleaner production techniques. Thus, R&D spending captures technology investments that determine both the speed and the path of technical change.
To better understand the R&D response identified in polluting firms, the “two faces” of R&D spending was explored: firms invest in R&D not only to develop new products and innovations but also to expand their capacity to understand and assimilate the external knowledge and technical know-how needed to fundamentally transform the way they produce. Broadly, the first face of R&D corresponds more closely to studies on the development of clean products and technologies, whereas the second face of R&D relates to the literature on re-engineering production and adopting cleaner production processes.
“No evidence was found that the high-pollution firms most affected by the higher emissions taxes drive the increase in clean patenting. Thus, the typical high-pollution firm responds to pollution taxes by increasing R&D, but not new invention.”
For these tests, information was used from the European Patent Office’s Worldwide Patent Statistical Database (PATSTAT). There was an evaluation of the new invention face of R&D by studying whether it is associated with more patentable innovative output. No relation was found between higher pollution taxes and patenting activity in all technology classes; however, pollution taxes are positively related to new patents in air pollution abatement technologies. Thus, higher pollution taxes appear to encourage the development of clean technologies, broadly consistent with the evidence in prior studies linking environmental policy with new invention. However, in sharp contrast to R&D, no evidence was discovered that the high-pollution firms most affected by the higher emissions taxes drive the increase in clean patenting. Thus, the typical high-pollution firm responds to pollution taxes by increasing R&D, but not new invention.
Why then do polluting firms increase R&D? A plausible answer is that they invest in R&D to expand their technological absorptive capacity. There is no proxy that directly measures output from the “second face” of R&D. A series of cross-sectional tests was thus derived to evaluate this mechanism. To start, firms were categorized based on two ex-ante proxies for the likelihood that any given dollar of R&D spending represents new product innovation. The R&D response to higher pollution taxes is concentrated in firms from sectors with low levels of new product innovation and in sectors in which R&D spending is more focused on process innovation rather than new product development. At a minimum, these findings are consistent with the idea that a key reason polluting firms invest in R&D is to expand absorptive capacity rather than to develop new patentable innovations.
“It is not the case that polluting firms are simply poor innovators who are unable to convert the higher R&D into new patentable output.”
Based on further tests and research, it was discovered that in settings in which new knowledge is difficult to appropriate and, as such, more readily spills across firms, firms have more incentive to invest in R&D to expand absorptive capacity but less incentive to invest in R&D for new invention. This insight suggests that to the extent the R&D response in polluting firms reflects investment in absorptive capacity rather than new invention, the R&D response should be relatively stronger among firms operating in sectors where external knowledge is easier to acquire. On the other hand, if the R&D response is about new invention, the effects should be relatively stronger in settings in which new knowledge is easier to appropriate.
It was found that polluting firms operating in sectors with low appropriability (high knowledge spillovers) account for essentially all of the R&D response documented. Moreover, higher pollution taxes are associated with an increase in the marginal value of R&D investment in the subsets of firms where R&D is most closely tied to absorptive capacity rather than new invention. Thus, it is not the case that polluting firms are simply poor innovators unable to convert the higher R&D into new patentable output.
----------------------------------------------------------------------------
The above is a synopsis of the original paper published by: