Field Experiments for Net Zero
Table of contents
Event description
This session presented field experiments on achieving net-zero emissions. It introduced innovation and behavioral economics approaches that have been tested through empirical experiments. The event demonstrated how randomized controlled trials can be used in the climate context to measure the effectiveness of various interventions. Practical examples of how behavioral change can be systematically promoted to achieve decarbonization goals and derive evidence-based policy recommendations have been discussed.
Recording
Written summary of the input presentations
Overview
This session introduced randomised controlled trials (RCTs) and field experiments as tools for measuring and promoting cost-effective and equitable decarbonisation. The speakers demonstrated how behavioural and innovation economics approaches can generate credible causal estimates of intervention effectiveness, and how evidence from well-designed experiments can directly inform climate policy design and prioritisation. The session spanned both solar PV home installations and firm-level innovation, highlighting the breadth of settings in which experimental methods are now being applied to climate, innovation and sustainability challenges.
Why experiments? The credibility revolution in climate policy evaluation
Prof. Kenneth Gillingham opened with the motivation for experimental methods in the climate context. Standard empirical strategies, such as comparing energy consumption before and after a policy change, or across treated and untreated regions, face well-known identification challenges: selection into treatment, anticipation effects, spillovers, and confounding trends. Randomisation solves these problems at the source: by generating treatment and control groups statistically identical in expectation, a well-powered RCT allows clean attribution of observed outcome differences to the intervention. This matters greatly for policies such as solar adoption or internal carbon pricing: an overestimate of effectiveness due to selection bias can lead to large-scale deployment of programmes that deliver far less than expected.
Solar installations: prices, complementarities, and their limits
Gillingham presented experimental evidence from the residential solar adoption. Information and social-norms interventions generated contagion in the adoption of solar home installations, arguably because they are very visible to neighbours in the form of « credibility-enhancing displays». Also, self-interest messaging was more effective in more affluent neighbourhoods, while pro-social messaging was more effective in lower income areas, and also creating more long-term satisfaction among recipients. For internal carbon pricing, Ken ran an RCT on internal carbon pricing at Yale, with an information provision having a lower impact than a control group, and actual carbon pricing having a more pronounced effect. Ken emphasized field experiments as a powerful tool for generating new knowledge for policy-design.
Innovation experiments
Albert Bravo-Biosca presented the application of experimental methods to innovation and firm behaviour, since European governments spend around 150 bln euros per year on innovation support programmes. Randomised evaluations of programmes providing firms with information, AI tools, grants, or technical assistance can help scale what actually works. To become more experimental, institutions need to embark on a journey that starts with a new mindset asking « what if », but also accept failure as part of a learning process. An experiment in this sense is more than just trying something new, but rather it implies intentionally testing hypothesis within a set timeframe. While in a UK evaluation of local economic development programs, a mere 3 percent had a credible control group. Experiments can be leveraged to test, refine or optimize innovation policies. There tends to be a tension between policy and research trials in terms of the questions they seek to address, which typically requires a certain level of compromise. Albert highlighted how little A/B testing is done in governments vis-a-vis the private sector for example. A prominent example of a randomized innovation trial facing firms is one with innovation vouchers that are handed out to firms to increase their innovative activity. Further, competitive funding calls are another area where more trials could be conducted to help tailor them more granularly for reaching their beneficiaries more effectively.
What experiments can and cannot tell us
The session closed with an honest assessment of limitations. While RCTs provide unparalleled internal validity, external validity, that is whether findings from one context generalise to others, remains a serious challenge: political expediency, long time horizons, and risk of failure, expecially in the political realm, but also career risk. Also, not all relevant questions can be adequately answered by randomized controlled trials. But also : you cannot always randomize who gets a treatment and who does not, especially when you are working with public funds. With carbon pricing for example, you oftentimes cannot discriminate who has to pay it and who does not. What worked on the Yale campus might not work at the level of the entire economy. An optimal first step, according to Albert, to make institutions embrace an experimental mindset is to create a small fund to help fund trials signaling that trials are supported.
Portrait Albert Bravo-Biosca (speaker)
Albert Bravo-Biosca is the Founding Director of the Innovation Growth Lab (IGL), a global partnership based at Nesta that brings together governments, foundations, and researchers to generate rigorous evidence on what policies and programs best support innovation and business growth. He is also a Research Fellow at the Barcelona School of Economics (BSE). Since joining Nesta in 2007, his work has been at the intersection of innovation, growth, and evidence-based policy. The Innovation Growth Lab works to improve the quality of policy decisions by promoting the use of randomized controlled trials and other rigorous evaluation methods to test innovation and growth policies. His research focuses on understanding what drives productivity growth, innovation, and business dynamism, and how policy can support these outcomes. He has worked with governments and organizations around the world to design and evaluate policies that support entrepreneurship, innovation, and economic growth.
Portrait Gerhard Fehr (moderator)
Gerhard Fehr is a delegate of the board of directors, partner, and behavioral designer at FehrAdvice & Partners AG. His personal mission is to inspire and empower businesses and political bodies to experiment. As a behavioral designer, he serves as a personal advisor to numerous politicians, board members, CEOs, and top managers. Gerhard Fehr holds a degree in business administration from the University of Vienna and is a trained journalist with over ten years of experience.
Portrait Prof. Kenneth Gillingham (speaker)
Professor Gillingham is the Senior Associate Dean of Academic Affairs at the Yale School of the Environment and a professor of economics at Yale, with a primary appointment at the School of the Environment and secondary appointments in the Department of Economics and School of Management. He is also a faculty research fellow at the National Bureau of Economic Research. In 2015-2016, he served as the Senior Economist for Energy and the Environment at the White House Council of Economic Advisers. He is an energy and environmental economist drawing from the fields of applied microeconomics, behavioral economics, industrial organization, and integrated assessment modeling of climate change. He has published widely on consumer decisions and policy in transportation, energy efficiency, and renewable energy. His work has been published in top-tier journals such as Science, Nature, PNAS, Journal of Political Economy, AEJ: Economic Policy, RAND Journal of Economics, Quantitative Economics, Management Science, Marketing Science, Journal of the Association of Environmental & Resource Economists, and the Journal of Environmental Economics & Management. He holds a Ph.D. from Stanford University in management science & engineering and economics.