Voluntary Programs to Encourage Diffusion: The Case of the Combined Heat-and-Power Partnership.

• Author: Ferrara, Andreas

1. INTRODUCTION

   Since the start of the first voluntary environmental programs in the early 1990s, governments have increasingly used this type of policy tool to achieve emissions reductions, raise firms' environmental awareness or improve information provision to the public. This trend has consequently led to a growing importance of measuring those programs' success (Brouhle et al., 2005 and EPA, 2007). The majority of the programs call on participating firms (known as partners) to commit to an action such as a reduction of emissions. A more novel use of voluntary programs has involved the acceleration of technology diffusion to overcome problems like asymmetric information, principal-agent issues or to lower the threshold of network externalities. The U.S. Environmental Protection Agency's (EPA) Combined-Heat-and-Power Partnership (CHPP) was established in 2001 and represents this application of voluntary programs. Combined heat-and-power (CHP) adoption in the U.S. has always responded notably to incentives as well as to disincentives generated by policy makers.

   The Public Utilities Regulatory Policy (PURPA) of 1978 led to increased installation of CHP plants, though this trend stopped in the early 1990s when the Energy Policy Act of 1992 altered the incentives for wholesale power purchases (Fox-Penner, 1990; Dismukes and Kleit, 1999). PURPA had given CHP generators wholesale generation privileges (utilities were required to buy from CHP generators if their costs were low enough), however the deregulation of electricity markets allowed independent producers to enter the wholesale generation market and compete with CHP generators (Kaarsberg et al., 1999). In the late 1990s, it was recognized that the loss of wholesale generation privileges and other barriers to CHP adoption required new measures to encourage CHP installations. Since then the EPA, the Department of Energy, and the states have made more efforts to remove such barriers and strengthen adoption incentives (Shipley et al., 2008). Some of these trends in CHP use can be seen in Figure 1, which shows the ratio of heat used by CHP plants in the industrial and electricity sector to all heat used in these sectors over time.

   The CHPP is a new attempt to reduce installation barriers and to promote the use of CHP. Designed as a multi-sector federal voluntary program, it aims to facilitate the diffusion of CHP systems by giving early-stage consulting support to firms, providing public recognition and hosting a platform for contacts and knowledge transfer. This paper attempts to fill a gap in the literature concerning the effectiveness of a program of this nature. Two hypotheses are to be tested here: (i) whether the partnership has encouraged the installation of CHP applications in electricity and manufacturing plants and (ii) if it has assisted knowledge transfers and spillovers that helped firms to utilize CHP more efficiently. We use a nearest neighbor matching estimator to test the installation hypothesis, with a robustness check using conditional logit model, and panel data techniques to test whether utilization of CHP is greater due to CHPP spillovers. Generally, the impact of the CHPP is positive; on average partners have a 3% higher probability of installation though there was no significant effect concerning learning spillovers and higher utilization of CHP due to the partnership. These results provide some evidence for a positive impact of the program on technology diffusion.

   The trend in environmental and energy policy over the past two decades has been to use market-based instruments to encourage more efficient outcomes. While there have been many market-based regulation successes (the Acid Rain program and natural gas deregulation, for example), concern is growing that few policies have been enacted in recent years. For example, the U.S. has not passed a comprehensive environmental law since the 1990 Clean Air Act Amendments and is largely without a comprehensive energy strategy (Hayward, 2010). Of the more recent policies, many do not specify a specific instrument to reach the specified targets. The UK Climate Change Act of 2008 sets emissions targets for firms but does not provide direct instruments to meet these goals, creating the Committee on Climate Change to make recommendations instead. The European Union 20-20-20 system calls for a 20% reduction in energy use through increases in energy efficiency to be achieved with a number of initiatives across different sectors.

   Most of these policies recommend a path to meeting such targets which rely on the use of new technologies that are either low-C[O.sub.2] emitting or improve the efficiency of a given amount of energy. However, new technologies do not spread throughout industry as efficiently as they should due to diffusion externalities such as learning-by-doing, incomplete information or network effects (Jaffe et al., 2005). Indeed, this explains why a basket of policy instruments are shown to be more efficient at achieving an emissions goal than any single instrument (Fischer and Newell, 2008). Voluntary programs like CHPP might be a potentially cost effective way to overcome these adoption and diffusion externalities. This type of program can complement policies that provide a goal but do not specify actions that need to be taken to achieve the goal. However, for this argument to hold, it must be shown that voluntary programs are overcoming the externalities they are meant to address.

   The only other evaluation of a voluntary environmental program that encourages the diffusion of a new technology that the authors are aware of is DeCanio (1998). DeCanio (1998) finds that the U.S. Green Lights program, a voluntary program that encouraged firms to use energy efficient lighting, has been successful at diffusing new lighting technology. In general, the empirical evidence is mixed regarding the effectiveness of the traditional voluntary programs in the economics literature. Boyd and Mason (2011) and Lyon and Maxwell (2007) discuss a number of reasons why it is difficult to undertake rigorous evaluations of voluntary programs. (1)

   CHPP was established in 2001 with the goal of promoting the use of CHP as a means of reducing the environmental impact of power generation (CHPP, 2010). The economic rationale for a program like CHPP comes from the innovation and diffusion externalities that are common with new technologies (Jaffe et al., 2005). These externalities arise from a number of sources, such as the public good nature of knowledge, learning-by-doing effects, and/or incomplete information. Currently there are 369 partners including federal, state and local government agencies as well as private organizations like energy users and producers, service companies, CHP project developers, consultants and manufacturers. To join CHPP, firms need to fill out a short postcard and submit it to the partnership. No promise of installing a CHP system is given when firms join though they agree to designate a liaison to the partnership to provide information on any CHP decisions being made.

   The CHPP utilizes a number of methods to encourage installation of CHP such as project-specific assistance, information and knowledge exchange opportunities, and public recognition. The project-specific assistance includes a basic cost-benefit analysis to determine whether CHP potentially generates net benefits at a given plant. Comprehensive information is provided on environmental, technical or policy related questions, potential funding opportunities, the next steps in the project development and contacts to engineers, parts suppliers and project developers to finalize the project. CHPP runs a number of workshops and web-seminars (webinars) for partners to discuss their experience with CHP system. Finally, public recognition is granted by listing partners' names on the EPA's CHP website and through awards like the Energy Star CHP award.

2. DATA

   The main data set used for the analysis is the EIA Form 906/920, a sample of utility and non-utility boilers for the years 2001 through 2008 in the U.S. The data are recorded annually at the plant level, and plants may contain more than one boiler. Only plants in the electricity generation or manufacturing sector are used in this analysis, North American Industrial Classification System (NAICS) 22 and 31-33, respectively.

   The data contain information on plant specific characteristics: primary fuel consumption on site, amount of fuel used on site, average total heat consumed on site, the location and industry code of the...