Forecasting Models and Value Based Decisions: Weak evidence, strong perceptions



Renewable energy plants surpassing a certain size must undergo an Environmental Impact Assessments (EIAs) prior to their permission. Within this process, the assessing authority evaluates whether the planned project will seriously harm the environment. If so, the project might only be permitted, if these harms are outweighed by possible positive effects. In the case of renewable energy plants, positive aspects might be the reduction of CO2-emissions, ensuring energy security or creating certain desired economic effects. Negative effects could be the irreversible destruction of a natural habitat, microclimatic disruptions, or biodiversity loss.

To decide which effects predominate, the assessing authorities rely on information provided by experts from different fields, ranging from hydrologists to botanists, engineers to economists. Their assessments are often based on complex forecasting models, taking into consideration diverse aspects such as expected economic development, expected changes in water levels due to climate change, expected energy consumption etc.

Forecasting models are highly useful and often necessary to support decisions. Simultaneously, decision makers should handle them with care, as forecasting models are prone to human errors and misinterpretations, depend on data quality and underlying assumptions, while simultaneously providing a sense of certainty they may not be able to deliver. Human factors, such as selection bias, the decision to (not) use a specific modelling technique or the interpretation of model results, as well as technical factors influence consequent actions. While such factors are often considered when models are developed, used, and interpreted in a scientific context, they tend to get lost once they leave the scientific sphere. While the technical aspects of models, their use and presentation as decision-support tools, and their perceived objectivity have been widely discussed in various disciplines, legal research is lacking behind, as could be seen during the COVID-19 pandemic.

A current study of EIAs in Austria shows, that the assessing authorities deem the public interest of environmental aspects least important, while public interests such as energy production or economic growth prevail. Taking into consideration these possible downfalls of models, I am asking, whether the models used to represent different interests are of equal quality, and whether the assessing authority takes this into consideration.

While acknowledging that the balancing of interests is considered a “value-based-decision”, I am asking, whether model quality should be of greater importance within processes like EIAs. I take an interdisciplinary approach on the intersection of legal research and qualitative research in the fields of forecast modelling, environmental analysis, and science and technology studies (STS) to demonstrate, that the public interests supported by the assessing authority (e.g. economic effects) are often less supported by evidence than environmental aspects.

Looking at the imminent dangers posed not by the unprecedented loss of biodiversity, water scarcity or extreme weather events, I argue that it might be necessary to stronger rely on evidence within legal procedures, while acknowledging the complex relationships between law, technology, and science.


Short biography

Annemarie Hofer has been university assistant at the Department of Innovation and Digitalisation in Law at the University of Vienna since April 2022. She studied Environment and Bio-Resources Management at the University of Natural Resources and Life Sciences in Vienna. In her dissertation, she takes an interdisciplinary approach towards the framing of scientific evidence in legal decisions.