Improving Technology Forecasting by Including Policy, Economic, and Social Factors

The goal of technology forecasting is to predict a technology’s future characteristics, applications, cost, performance, adoption, existence, or capabilities. Improving technology forecasting is important because these forecasts are used by public and private decision-makers to inform decisions and create policies. Technology forecasts are intended to help decision-makers anticipate future events, avoid surprises, set priorities, and allocate resources effectively. To that end, technology forecasts should be both as accurate as possible and include uncertainty, so that decision-makers know how heavily to rely on them. However, when technology forecasts have been assessed retrospectively, they are often shown to not be very accurate. Generally, only people with expertise in engineering and technology are consulted when making technology forecasts. This dissertation describes the development and demonstration of an improved method of technology forecasting that includes consideration of policy, economic, and social factors that may affect how a technology develops. It is designed to help forecasters better imagine and think more systematically about factors beyond a technology itself that may shape its future. This dissertation provides evidence for the efficacy of this method of including policy, economic, and social factors in improving technology forecasting. 

The first chapter of this dissertation presents an overview of the few publicly available retrospective assessments of technology forecasts and recommends best practices for making such retrospective assessments. These best practices include making forecasts as specific as possible, including some measure of uncertainty in each forecast, and designing forecasts so they can be assessed retrospectively. I attempted to adhere to these best practices in the following two studies in this dissertation.

The second chapter presents a pilot test of the proposed improved technology forecasting method that includes consideration of policy, economic, and social factors. A randomized controlled trial (n=133) was used to test the effectiveness of short briefings on reducing overconfidence in emerging automotive technology forecasts. The briefings warned participants to avoid overconfidence and encouraged them to think about how policy, economic, and social factors (such as government regulations, gasoline prices, and news coverage) may affect their forecasts. A retrospective assessment of the forecasts showed that the briefings reduced overconfidence in the forecasts. 

The third chapter is an application of the improved forecasting method to the future demand of power electronic applications in 2030 and which policy, economic, and social factors would be the most influential in determining that future demand. Fifteen experts were interviewed about the likely future demand of emerging power electronic applications (such as solar panels, transmission lines, and electric vehicles) with a focus on relevant policy, economic, and social factors (such as legislation, competitiveness, and public perception). The experts forecasted that electric vehicles, charging infrastructure, transmission, distribution, generation, ICT infrastructure, and the residential sector would have the highest future demand, approximately 2x-8x the demand today. The most influential policy, economic, and social factors on future demand were legislative action on climate change, the state of the global economy, sufficient resources, a reliable electric grid, public perception, and affordability