The RAND Corporation is a nonprofit institution that helps improve policy and
decisionmaking through research and analysis.
This electronic document was made available from as a public service
of the RAND Corporation.
Skip all front matter: Jump to Page 16
Support RAND
Browse Reports & Bookstore
Make a charitable contribution
For More Information
Visit RAND at
Explore the Pardee RAND Graduate School
View document details
Limited Electronic Distribution Rights
This document and trademark(s) contained herein are protected by law as indicated in a notice appearing
later in this work. This electronic representation of RAND intellectual property is provided for noncommercial use only. Unauthorized posting of RAND electronic documents to a non-RAND website is
prohibited. RAND electronic documents are protected under copyright law. Permission is required from
RAND to reproduce, or reuse in another form, any of our research documents for commercial use. For
information on reprint and linking permissions, please see RAND Permissions.
This product is part of the Pardee RAND Graduate School (PRGS) dissertation series.
PRGS dissertations are produced by graduate fellows of the Pardee RAND Graduate
School, the world’s leading producer of Ph.D.’s in policy analysis. The dissertation has
been supervised, reviewed, and approved by the graduate fellow’s faculty committee.
Policy Impacts on Wind
and Solar Innovation
New Results Based on
Article Counts
This document was submitted as a dissertation in April 2013 in partial fulfillment
of the requirements of the doctoral degree in public policy analysis at the Pardee
RAND Graduate School. The faculty committee that supervised and approved
the dissertation consisted of Siddhartha Dalal (Chair), Nicholas Burger, and
Robert Lempert.
The Pardee RAND Graduate School dissertation series reproduces dissertations that
have been approved by the student’s dissertation committee.
The RAND Corporation is a nonprofit institution that helps improve policy and
decisionmaking through research and analysis. RAND’s publications do not necessarily
reflect the opinions of its research clients and sponsors.
R® is a registered trademark.
Permission is given to duplicate this document for personal use only, as long as it is unaltered
and complete. Copies may not be duplicated for commercial purposes. Unauthorized
posting of RAND documents to a non-RAND website is prohibited. RAND documents are
protected under copyright law. For information on reprint and linking permissions, please
visit the RAND permissions page (
Published 2013 by the RAND Corporation
1776 Main Street, P.O. Box 2138, Santa Monica, CA 90407-2138
1200 South Hayes Street, Arlington, VA 22202-5050
4570 Fifth Avenue, Suite 600, Pittsburgh, PA 15213-2665
To order RAND documents or to obtain additional information, contact
Distribution Services: Telephone: (310) 451-7002;
Fax: (310) 451-6915; Email: [email protected]
Summary The extent to which technological change will help mitigate climate change is a subject of some debate among climate policy researchers and stakeholders. In particular, models predicting the costs and effectiveness of climate policies rarely include the effects of technological change because it is difficult to predict how much change will occur. This study takes an empirical approach to filling this gap in the case of solar and wind energy research. In order to do so, a new set of data is collected to serve as a proxy for technological change, focusing on research production: the number of technical journal articles on solar or wind energy published each month. A combination of hand‐sorting and Bayesian logistic modeling is used to identify these articles, which number in the tens of thousands. The resulting monthly article counts are used to assess how solar and wind research production may have responded to major U.S. renewable energy policies. The first essay focuses on these empirical findings, the second on methodological comparisons, and the third on subcategories of solar energy and their potential relationships with policy. Both direct and indirect subsidies are found to be associated with increases in research production, as described in the first essay. Indirect subsidies are represented by the largest U.S. federal support for renewable energy, the Production Tax Credit, considered in combination with the renewable energy Investment Tax Credit. Most of the value of these credits is used for wind, although they are also available for solar. These credits are widely considered to drive installations of wind turbines, but their effects on research were previously unmeasured. Increasing the tax credits by $20 million per year (1% of their 2008 value) is found to be associated with a 1% increase in solar article counts and positive but statistically insignificant effects on wind article counts. To the author’s knowledge, this is the first study to successfully quantify these tax credits’ relationship with innovation. Also investigated is the impact of federal solar or wind research funding, i.e. direct subsidies for research. For every $1 million dollars spent on solar or wind research, solar or wind article counts increase by 1‐2%. Although the impact of research funding on research is intuitive, this result contradicts the hypothesis that public funds might simply displace private research funding with no net ix
increase in research. These findings can be used as inputs to climate policy models in order to calibrate the effects of renewable energy production subsidies and research subsidies on solar and wind energy research. The second essay focuses on methods of measuring innovation and how they affect the results. Three measures of innovation—article counts from the first essay, article counts based on keyword selection, and patents—are constructed and subjected to the same analysis as in the first essay. Patents, selected using either patent classes or by searching for keywords, are the leading measure of innovation for economic and policy analyses. Similarly, keyword selection is the usual method of finding and counting articles, which has been done in many other contexts including examination of topical trends within wind and solar energy research. Testing against a hand‐sorted sample of articles indicates that Bayesian regression and keyword selection are roughly similarly successful at identifying relevant articles, although assessment methods used favor keyword selection. Bayesian article selection identifies more articles and may be more reliable, but keyword selection is faster and in this case, both types of article counts give substantively analogous results when regressed with tax credits and subsidies. These findings suggest that it is at least sometimes possible to choose keywords which work approximately as well as the Bayesian regression models used. Thus, reasons for selecting one method over the other in the future may be case‐specific. For both wind and solar energy, patents’ associations with research funding are very similar to the associations for articles, while their associations with the tax credits are not statistically significantly different from zero or from article count results. Patents’ failure to find an association with tax credits when article counts do find such a result may be caused in part by patents being far fewer in number than articles. These findings can be interpreted to suggest that article counts are as valid a measure of innovation impacts as patent counts, and are more effective at measuring small effects. Further suggestions for future research methods are discussed. Again using Bayesian regression and keyword approaches, the third essay investigates trends over time in the two main subcategories of solar energy research: monocrystalline silicon and thin film. While demonstrating the methods with more challenging article topics, x
this study also offers preliminary findings on whether public policies have affected the distribution of research effort between monocrystalline silicon and thin film. Monocrystalline silicon, known also as the “first generation” of solar cells, is the technology used in most commercially available silicon cells to date. Thin films are a competing technology which, since they use thinner materials, are often less efficient but potentially cheaper. They are known as the “second generation” of solar technologies. Predictions that thin films imminently will dominate the market date to at least 1985. This study finds that thin film articles collected by either method have far outnumbered monocrystalline articles consistently throughout 1985‐
2010, the entire time period considered. U.S. research subsidies appear to have potentially favored monocrystalline silicon over thin film research in the more applied database, demonstrating that research subsidies can be used to steer net solar research. The ratio of monocrystalline silicon to thin film research appears to have been otherwise unaffected by the U.S. tax credits and research subsidies, consistent with the philosophy that policies should avoid being technology‐specific. Taken together, these essays provide both substantive and methodological results. Firstly, they demonstrate the substantial impacts that public policies can have on the volume of solar and wind innovation, even when innovation is not the policy’s direct target. They provide detailed results on the relative sizes of these impacts, which can be used as inputs to subsequent analyses. Simultaneously, they show that hand‐sorting followed by Bayesian regression is an effective way to count large numbers of articles for the purposes of social science research, although carefully chosen keywords may sometimes perform similarly well. The solar and wind article counts thus created are also suitable for many more studies than could be conducted here. This dissertation provides new solar and wind article count data, the methods used to create it, and information on how direct and indirect subsidies can drive solar and wind research. xi

The RAND Corporation is a nonprofit institution that helps improve... decisionmaking through research and analysis.