innovation under the climate wise program schumptoberfest october 22, 2011 keith brouhle grinnell...
TRANSCRIPT
Innovation under theClimate Wise Program
SchumptoberfestOctober 22, 2011
Keith BrouhleGrinnell College
Brad GrahamGrinnell College
Donna HarringtonUniversity of Vermont
Voluntary Environmental Programs
Voluntary Environmental Programs Encourage firms to reduce toxic releases, waste, and
other adverse environmental impacts Encourage adoption/development of environmentally
friendly practices and technology Improve regulatory compliance Used with and, sometimes in place of, regulation
Advantages Low administrative costs Encourage government/industry collaboration More cost effective than command and control
Do they work?
The Climate Wise Program
The Climate Wise voluntary program ran from 1993 to 2000 under the direction of the EPA and Dept. of Energy.
The program targeted non-utilities in the manufacturing
sector, but other entities could also join.
Four broad program objectives: i. Reduce greenhouse gas (GHG) emissions
ii. Change the way companies view environmental performance
iii. Develop productive partnerships between government and industry
iv. Foster innovation
Program focused on innovation in the areas of energy efficiency, renewable energy, and pollution prevention.
Program hoped to spur innovation by Encouraging broad goals
Allowing organization to identify the most cost-effective ways to reduce GHG emissions
Providing technical assistance
Over the course of the program, 701 entities joined Climate Wise and 353 submitted an Action Plan.
Climate Wise and Innovation
Objectives of Study
1. Determine the motivations for pledging to participate in the Climate Wise program.
2. Analyze the determinants of innovative behavior and determine whether participants in Climate Wise have significantly higher measures of innovative activity (environmental and non-environmental patents) than non-participants.
Background literature
Studies evaluate the effectiveness of VAs using different environmental measures: Most use release of emissions of toxic chemicals as
measure of environmental output (33/50 program, ISO certification, Strategic Goals program).
Other environmental outputs include environmental rating (Sustainable Slopes), fuel/electricity usage (Climate Challenge), and compliance with regulations (33/50, ISO).
We propose to use patents as a measure of environmental innovation.
Background literature
The literature that explores the relationship between environmental regulation and innovation includes: Several studies hypothesize that higher costs of
pollution abatement activities (possibly due to stricter regulations) lead to innovation (Lanjow and Mody 1996, Jaffe and Palmer 1997, Brunnermeier and Cohen 2003, Carrión-Flores and Innes 2010).
The impact of command and control regulations on innovation is explored in Bellas (1998), Lange and Bellas (2005), and Popp (2002, 2003, 2006).
Fewer studies have evaluated the relationship between VAs and innovation (working paper by Carrión-Flores, Innes, and Sam (2006)).
Conceptual Framework
Potential Benefits…
Regulatory relief
Improved reputation w/stakeholders (green consumers, investors and NGOs)
Cost savings/efficiency gains
Potential Costs…
costs of GHG mitigation and innovation activities undertaken as a result of participation in Climate Wise.
Firms will voluntarily participate/engage in the Climate Wise program when the benefits exceed the costs.
Conceptual Framework
Innovativeness
Economies of scale in R&D
Scope for innovation
Market structure and industry traits
Environmental cost savings
Regulatory pressure
Why do some firms innovate more than others?
The patenting behavior of firm i at time t is
where Yit = number of patents Dit= participation in the Climate Wise
ProgramX1it= vector of exogenous explanatory
variables
itititit XDY 111
Empirical Model
As with any standard treatment problem, estimation of the above equation may lead to biased results if the treatment is not random. Do factors that affect the decision to participate in Climate Wise also affect innovation?
Potential solutions to non-random treatment?
1.If one believes treatment occurs on observables, use a matching approach
2.If one believes treatment occurs on unobservables, use an IV approach
Our default is to use an IV approach.
Empirical Model
Empirical Model
The net benefits from participating in Climate Wise for firm i at time t are
Dit* = β2 X2it + ε2it
where Dit* = net benefits from participating in
Climate Wise.
The net benefits are not observable, so we estimate
Dit = F(β2X2it) + μit
where Dit = 1 if Dit* > 0 or 0 otherwise.
Explanatory Variables
We organize our set of explanatory variables for the participation and innovation equations into the following categories:
Firm specific environmental variables
Firm specific market and financial variables
Industry variables
Location variables
Explanatory Variables
Firm specific environmental variables
Firm specific market and financial variables
Industry variables
Location variables
Explanatory Variables
Firm specific environmental variables Superfund TRI Releases Violations Spills
Firm specific market and financial variables
Industry variables
Location variables
Explanatory Variables
Firm specific environmental variables Firm specific market and financial variables
Return on Assets Profit Assets to Equity; Debt to Assets R&D Employees Patent stock
Industry variables
Location variables
Explanatory Variables
Firm specific environmental variables Firm specific market and financial variables
Industry variables Final good Concentration New Capital Expenditures Industry dummies
Location variables
Explanatory Variables
Firm specific environmental variables Firm specific market and financial variables
Industry variables
Location variables Total Energy Price Green Energy Program Fuel Mix Disclosure State regulatory stringency (Levinson index) Climate Wise Partner State
Sample: Firms that are part of the Corporate Environmental Profile Database (CEPD) from Risk Metrics as well as the NBER
Patent database and Compustat.
Sample size: 10,998 firm-year observation from 1,738 unique firms.
Sample period: 1993 to 2003
Data and Sample
Summary Statistics
Looking at the characteristics of participants, Climate Wise participants were… In industries that impact the environment
(petroleum & chemicals, construction, electric utility sectors) Dirty in the past
(higher TRI releases; more past violations; more spills; more Superfund sites)
Close to consumers(final good producer)
And innovative(more R&D; more patents)
Results: What influences participation?
Dependent variable: Climate Wise participation
Coefficient
Environmental
TRI Releases 0.036***(0.012)
Superfund 0.015***(0.004)
Violations 0.233**(1.119)
Spills -0.087*(0.045)
FinancialR&D 0.002***
(0.0004)
Return on Assets 0.008***(0.003)
LocationLevinson 1.379***
(0.508)
Climate Wise Partner State 0.500**(0.205)
Industry Final Good 0.427***(0.136)
Results: What influences participation? Among environmental variables:
Most are significant: TRI Releases, Superfund and Violations positively influence participation; Spills negative influences participation.
Among financial variables: R&D and Return on Assets positively influence participation.
Among location variables: Levinson and CW Partner significantly influence participation
(positive).
Among industry variables: Final Good and industry dummies positively influence participation
Note, year dummies and EPA region dummies are significant too.
Results: What influences patenting?
With results from participation model, we can calculate for each firm the probability that they will join Climate Wise in each year. This predicted probability serves as our instrument in
our patenting equations.
We investigate the impact of Climate Wise participation on innovation behavior (patents). We consider patents in the environmental area
(Carrión-Flores and Innes 2010), non-environmental area, and total patents.
We estimate fixed effects panel Poisson models.
Results: What influences patenting?
Env Patents
Climate Wise
Predicted CW participation
0.0254(0.100)
Environmental
Violations-0.004(0.019)
TRI Releases0.060***(0.003)
Spills-0.008(0.006)
Financial
R&D0.001***(0.00005)
Patent stock0.001***(0.00003)
Employees0.637***(0.011)
Debt to Equity-0.006***(0.001)
Results: What influences patenting?
Env Patents
Location
Fuel mix disclosure0.131***(0.019)
Energy price0.022**(0.007)
Number of programs0.008
(0.005)
Industry
Final good-0.169***(0.025)
Concentration ratio0.00004(0.0004)
New capital expenditures0.031***(0.002)
Results: What influences patenting?
Env Patents
Non-Env Patents
Climate Wise
Predicted CW participation
0.0254(0.100)
0.286***(0.047)
Environmental
Violations-0.004(0.019)
-0.168***(0.007)
TRI Releases0.060***(0.003)
0.056***(0.001)
Spills-0.008(0.006)
-0.036***(0.004)
Financial
R&D0.001***(0.00005)
0.001***(0.00002)
Patent stock0.001***(0.00003)
0.001***(0.000003)
Employees0.637***(0.011)
0.492***(0.004)
Debt to Equity-0.006***(0.001)
-0.005***(0.0002)
Results: What influences patenting?
Env Patents
Non-Env Patents
Location
Fuel mix disclosure0.131***(0.019)
0.132***(0.008)
Energy price0.022**(0.007)
-0.068***(0.003)
Number of programs0.008
(0.005)0.006***(0.001)
Industry
Final good-0.169***(0.025)
-0.275***(0.011)
Concentration ratio0.00004(0.0004)
0.005***(0.0002)
New capital expenditures0.031***(0.002)
0.004***(0.0005)
Results: What influences patenting? Across patenting, we consistently find a significant
influence of Environmental factors: TRI Releases, Superfund Financial factors: R&D, patent stock, size, debt to equity Location factors: fuel mix disclosure
For environmental patenting, we note Spills, violations, and concentration ratio are not as important Total energy price has an different impact
Interestingly, the impact of the Climate Wise program varies: Positively influences overall patenting and regular patenting, but
no detectible influence on environmental patenting.
Conclusions
Objective 1: Determine the motivations for pledging to participate in the Climate Wise program.
Participants in Climate Wise were more likely to … have higher toxic releases, more violations and responsible for
Superfund site, be from a dirty economic sector that is a final good producer, spend more on R&D and have higher return on assets, and be located in a Climate Wise Partner state.
Conclusions
Objective 2: Analyze the determinants of innovative behavior and determine whether participants in Climate Wise have significantly higher measures of innovative activity (environmental and non-environmental patents) than non-participants.
Patenting is associated with… spending on R&D, past stock of patents, the scope of a firm’s exposure to environmental issues, a firm’s financial health as well as whether the firm pledged to participate in Climate
Wise.