Environmental innovation and environmental performance
Introduction
Does environmental policy spur innovation in environmental technology? Conversely, does environmental innovation lead to a tightening of environmental standards, reflecting the lower pollution abatement costs associated with better technologies? Recent empirical work focuses on the first question, finding evidence of induced innovation. In particular, higher pollution abatement expenditures (PAE) – attributable to tighter environmental policy – are estimated to increase rates of environmental patenting [12], [24]. However, in principle, causal effects may go in both directions: environmental policy may spur innovation, and innovation may spur tightening of environmental policy.
This observation is important for at least three reasons. First, we would like to understand the effects of environmental innovation on environmental performance. The most likely channel for effect is due to innovation-induced tightening of government standards. However, innovation may also spur at least temporary over-compliance with government pollution standards by lowering costs of meeting demands of environmental NGO's and green consumers [5], [7], [21]. Identifying the qualitative and quantitative impact of innovation sheds light on the potential benefit of promoting environmental research in order to reduce toxic pollution.
Second, regardless of which direction of effect one wishes to study – how policy affects research or how innovation affects policy-driven environmental performance – one needs to account for the other direction of causal effect. That is, innovation and policy are, at least in principle, jointly determined.1 Hence, estimates of induced innovation effects that fail to account for the joint endogeneity of innovation and policy are likely to be biased.
Third, ultimately one would like to understand whether, and to what extent, tightened environmental policy can stimulate innovation and thereby yield additional long-run environmental dividends – long-run pollution reductions beyond those required by the initial tightening of standards. Because these additional pollution reduction effects multiply the initial pollution reduction, they represent what we call an environmental policy multiplier. To identify such benefits requires studying both directions of causal effect between policy and research outcomes, the object of our study.
In this paper, we examine 127 manufacturing industries over the 16-year period (1989–2004). Changes in environmental technologies, as measured by the number of environmental patents, can lead to changes in producing firms’ pollution targets, which in turn drive their observed emissions. Emissions in turn proxy for the changes in pollution targets that drive environmental R&D and, hence, resulting patents. In view of the joint determination of research and pollution outcomes, we estimate two simultaneous equations, using appropriate instruments to identify each endogenous variable.
This paper contributes to a surprisingly small empirical literature on environmental innovation.2 This literature focuses on the effects of pollution abatement expenditures (PAE) on innovative activity. Jaffe and Palmer [24] find evidence for the induced innovation hypothesis in US industry-level panel data on total (environmental and non-environmental) R&D expenditures and patent counts. Lanjouw and Mody [28] also find informal evidence that environmental innovation is induced by higher PAE, presenting tabular data on environmental patents and control costs from the US, Germany and Japan. Brunnermeir and Cohen [12] are the first to estimate a model that links PAE to US environmental patent counts, again finding evidence in support of the induced innovation hypothesis.3
Our work differs from previous studies in a number of respects. To our knowledge, this is the first paper to directly estimate the impact of environmental innovation on pollution. Relative to the induced innovation literature, we study a model of bi-directional links that explicitly accounts for the joint determination of policy-induced pollution outcomes and environmental R&D, and we use a more direct measure of policy stringency, emissions as opposed to PAE. PAE costs are problematic when one is interested in bi-directional effects. The reason is that innovation can be expected to lower PAE costs directly, but indirectly raise them due to a stimulated tightening in emission standards. Our more direct measure of regulatory stringency enables us to identify the latter link between innovation and policy.
Because the two directions of causal effect are expected to be reinforcing – both negative, with higher emissions lowering research incentives, and greater research output leading to tighter environmental targets – one expects that our accounting for joint endogeneity will dampen estimated impacts in both directions. We nevertheless find policy-induced innovation and innovation-induced pollution effects that have the predicted negative sign, and are statistically significant. In quantitative terms, our estimates of policy effects on innovation are small, but much larger than suggested by prior work [12]. However, our estimated effects of patent counts on environmental performance are large by any measure, revealing the central role of innovation in achieving reductions in toxic pollution.
Section snippets
Empirical model
We envision an underlying structural model that determines four outcomes, our two observable variables (emissions and patents) and two unobservable variables (effective industry pollution targets and environmental R&D). We assume that this model takes the following simple form, reflecting intuitive dynamics and causal relationships that we describe below: where Pit is
Data
Our sample is a balanced industry-level panel of 127 manufacturing industries (SIC codes 200–399) over the period 1989–2004. Because we focus on toxic emissions, we restrict attention to manufacturing industries that are the principle sources of such pollutants. Table 1, Table 2 present variable definitions and descriptive statistics for our sample.
Using the EPA's Toxic Release Inventory (TRI), we construct an industry-level measure of regulated toxic air releases (Emissions) by aggregated
Econometric methods
We have two simultaneous equations which we estimate equation-by-equation.14 A number of econometric issues arise. First, we have a panel data structure and, hence, need to account for individual effects. Second, we have endogenous
Empirical findings
Before turning to our two equations, we note that a key issue motivating our work is the prospective joint endogeneity of emissions and patent outcomes. Given endogeneity tests available to us, we are able to provide some preliminary evidence that we indeed have simultaneity in our sample. In particular, for our IV fixed effects emissions equation, we can test for the endogeneity of patents (ENVPATENTSMA) with a standard Hausman statistic; the resulting (Chi-square (1)) statistic is 14.67 with
Conclusion
In this paper, we present empirical evidence of bi-directional linkages between toxic pollutant emissions, on the one hand, and environmental innovation, on the other hand. Emissions and environmental R&D are jointly determined as successful R&D prompts pollution reduction, and as the anticipated tightening of pollution targets spurs new R&D. Specifically, we examine 127 manufacturing industries over the 16-year period, 1989–2004, accounting for the joint determination of research and pollution
Acknowledgments
We thank Price Fishback, Alfonso Flores-Lagunes, Kei Hirano and Ron Oaxaca, for invaluable comments and discussions. We also thank workshop participants at the University of Arizona, the University of Rhode Island, University of Florida, Cal Poly, and the AAEA meetings. We are particularly grateful to Mary Curtis and Terrence Mackey for their help in obtaining our data. Finally, we are greatly indebted to an anonymous referee, an Associate Editor, and the Editor, Dan Phaneuf, for penetrating
References (41)
- et al.
Towards a theoretical model of voluntary over-compliance
J. Econ. Behav. Organ.
(1995) - et al.
Individual effects and dynamics in count data models
J. Econometrics
(2002) - et al.
Determinants of environmental innovation in US manufacturing industries
J. Environ. Econ. Manage.
(2003) Regulatory factors shaping environmental performance at publicly-owned treatment plants
J. Environ. Econ. Manage.
(2004)Did the EPA's voluntary industrial toxics program reduce emissions?
J. Environ. Econ. Manage.
(2006)- et al.
Compliance and enforcement: air pollution regulation in the US steel industry
J. Environ. Econ. Manage.
(1996) A simple test for exogeneity in probit, logit, and poisson regression models
Econ. Letters
(1990)- et al.
EPA's voluntary 33/50 program: impact on toxic releases and economic performance of firms
J. Environ. Econ. Manage.
(1999) - et al.
Innovation and the international diffusion of environmentally responsive technology
Res. Pol.
(1996) - et al.
Unit root tests in panel data: asymptotic and finite sample properties
J. Econometrics
(2002)
Permits, standards and technology innovation
J. Environ. Econ. Manage.
Dynamic incentives by environmental policy instruments – a survey
Ecolog. Econ.
Can voluntary pollution prevention programs fulfill their promises? Further evidence from the 33/50 program
J. Environ. Econ. Manage.
Bounding the effects of R&D: an investigation using matched establishment-firm data
RAND J. Econ.
Some tests of specification for panel data: Monte Carlo evidence and an application to employment equations
Rev. Econ. Stud.
Panel Data Econometrics
Panel data models: some recent developments
Why do firms volunteer to exceed environmental regulations? Understanding participation in EPA's 33/50 program
Land Econ.
Private politics, corporate social responsibility, and integrated strategy
J. Econ. Manage. Strategy
Econometric Analysis of Panel Data
Cited by (393)
Pollution and green innovation: Evidence from CEOs’ early-life experience
2024, International Review of Economics and FinanceHow does cultural diversity determine green innovation? New empirical evidence from Asia region
2024, Environmental Impact Assessment ReviewBelated stock returns for green innovation under carbon emissions trading market
2024, Journal of Corporate FinanceEco-innovation and financial performance nexus: Does company size matter?
2024, Journal of Open Innovation: Technology, Market, and ComplexityShifting focus from end-of-pipe treatment to source control: ESG ratings’ impact on corporate green innovation
2024, Journal of Environmental Management