Public Policy and Decision Science
This is the approved revision of this page, as well as being the most recent.
🌎 This article is a stub, and is currently under construction. You can help by adding to it!
This page is about the intersection of policy-making and machine learning. For an overview of policy-making and decision science, please see the Wikipedia page on this topic.
When creating policies, decision-makers must often negotiate fundamental uncertainties in the underlying data and construct mathematical models to help them assess or trade off between different policy alternatives (ex-ante policy analysis). ML can help alleviate some of this uncertainty by extracting information from satellite imagery, sensors, social media posts, policy documents, and other source and provide new techniques for working with models commonly used by decision-makers (e.g. integrated assessment models, multi-objective optimization, etc.). When deciding between different policy options, it is necessary to understand how effectively each strategy will reduce emissions, as well as how society may be affected. ML can also help retroactively, by analyzing the text of existing policies and by performing causal inference on historical data (ex-post policy analysis).
Machine Learning Application Areas[edit | edit source]
Gathering decision-relevant data[edit | edit source]
- Data management, cleaning and imputation: ML can help with cleaning, merging and completing datasets that are relevant for policy-making.
- Remote sensing: Satellite data can provide a lot of valuable information to policy makers, for example by helping map infrastructure, land use and ecosystem health. The field of remote sensing has seen large improvements with ML.
- Computational text analysis: Text documents are an important source of information to inform climate policy. Natural language processing can help to analyze large corpora of text.
Decision science[edit | edit source]
- Decision science
- Multi-criteria decision-making: Multi-criteria decision-making can also help policy-makers manage trade-offs between different policies. Computational approaches and machine learning can help with finding solutions to these optimization problems.
Modeling and planning[edit | edit source]
- Agent-based modeling: ABMs are used in simulating the actions and interactions of agents in their environment, and ML can help integrate data-driven insights into these models, for example by learning rules or models for agents based on observational data.
- Energy system modeling
- Urban planning: With new techniques such as surrogate modeling, ML can help with complex planning tasks.
- Power System Planning: Algorithms for planning new low-carbon energy infrastructure are often large and slow. ML can help speed up or provide proxies for these algorithms.
- Integrated assessment models: IAMs are large simulations used to explore future societal pathways that are consistent with climate goals, which can be improved with ML.
Ex-post policy analysis[edit | edit source]
- Causal inference with machine learning: To analyze whether a policy intervention has had the desired effect, causal inference is an important tool.
Background Readings[edit | edit source]
General[edit | edit source]
- Resources for Effective Climate Decisions. (Ch. 4) Informing an Effective Response to Climate Change (2010): A chapter from a report published after a series of five coordinated activities convened by the National Research Council in response to a request from Congress. Available here.
- Social, Economic, and Ethical Concepts and Methods. (Ch. 3). (2014) : a report by the Intergovernmental Panel on Climate Change (IPCC) regarding the social and economic aspects of climate change. Available here.
- Climate Change Policies (2016): A report by the European Environmental Agency on defining successful, impactful policies for climate change. Available here.
Online Courses and Course Materials[edit | edit source]
Educational resources for policy analysis[edit | edit source]
- Theory and Practice in Policy Analysis: Including Applications in Science and Technology (Morgan, 2017): A rich resource for teaching classes on policy analysis with a focus on science and technology. Available here.
- Policy Analysis: Concepts and Practice (Weimer & Vining, 2017): Essential primer on policy analysis. Available here.
Conferences, Journals, and Professional Organizations[edit | edit source]
Major conferences[edit | edit source]
- Data for Policy: A global forum for interdisciplinary and cross-sector discussions around the impact and potentials of the digital revolution in the government sector (international conference, UK-based).
- International Association for Energy Economics (IAEE) Conferences: Main venue for academic and professional energy analyst, with a strong policy focus (international and regional).
- Association for Public Policy Analysis and Management (APPAM) Conferences: Conferences and events dedicated to improving public policy and management (international and regional).
- International Conference on Public Policy (ICPP):
Major societies and organizations[edit | edit source]
- International Association for Energy Economics (IAEE): A major society for academic and professional energy analyst, with a strong policy focus.
- Association for Public Policy Analysis and Management (APPAM): Society dedicated to improving public policy and management by fostering excellence in research, analysis, and education.
- International Public Policy Association (IPPA): A non-profit organization with the aim of promoting scientific research in the field of Public Policy, and to contribute to its international development.
Libraries and Tools[edit | edit source]
Given the importance of representing the impacts of decision-making and market-based strategies, interactive simulation tools and packages for multi-objective optimization are particularly useful in this application. Some of these are listed below:
- Python packages for multi-objective optimization:
- Evolutionary Multi-Objective Optimization (EMOO)
- Platypus - Multi-objective Optimization in Python: Platypus is a framework for evolutionary computing in Python with a focus on multiobjective evolutionary algorithms (MOEAs), providing optimization algorithms and analysis tools for multiobjective optimization.
- Matlab package for multi-objective optimization: Matlab package for multi-objective optimization, accompanied by tutorial videos and explanations.
Data[edit | edit source]
There are several sources of data at various global, regional, and national levels, all of which are useful for modeling the impact of policies.
Climate policy databases[edit | edit source]
- Policies database of the International Energy Agency (IEA): one of the largest international climate and energy policy databases, integrating the IEA/IRENA Renewable Energy Policies and Measures Database, the IEA Energy Efficiency Database, the Addressing Climate Change database, and the Building Energy Efficiency Policies (BEEP) database.
- Climate Policy Database of the New Climate Institute: covering policies from top 30 emitting countries which cover 82% of global GHG emission.
- Database of State Incentives for Renewables & Efficiency (DSIRE): US subnational climate policy database.
- ClimActor, harmonized transnational data on climate network participation by city and regional governments: includes more than 10,000 city and regional governments.
- Environmental Treaty Status Data Set, 2012 Release (1940–2012): provides information on the status of country participation in international environmental agreements.
Climate change impacts and adaptation[edit | edit source]
- Vulnerability to Climate Change Dataset: quantifies the vulnerability of 233 countries to three major effects of climate change (weather-related disasters, sea-level rise, and reduced agricultural productivity).
Carbon price data[edit | edit source]
- CO2 “price” in European ETS: Data about the European Union Emissions Trading System (ETS), coming mainly from the EU Transaction Log.
Data by or relevant to international organizations[edit | edit source]
- World Bank ClimateSmart data portal: focuses on the needs of practitioners working in developing countries on low-emission development and climate resilient projects.
- IPCC Socio-Economic Baseline Data, v1 (1980, 1990, 1991, 1992, 1993, 1994, 1995, 2025): dataset for the evaluation of climate change impact curated by the Intergovernmental Panel on Climate Change (IPCC)
- IPCC Fourth Assessment Report (AR4) Observed Climate Change Impacts, v1 (1970–2004): database with observed responses to climate change for multidisciplinary studies curated by the IPCC.
- OECD Data: The OECD publishes many economic and social indicators that are relevant for climate policy.
References[edit | edit source]
- Council, National Research (2010-07-21). Informing an Effective Response to Climate Change. ISBN 978-0-309-14594-7.
- Social, Economic, and Ethical Concepts and Methods, Cambridge University Press
- Del Río, Pablo. "Climate Change Policies and New Technologies". Climate Change Policies. doi:10.4337/9781781000885.00016.
- Morgan, Granger (2017). Theory and Practice in Policy Analysis: Including Applications in Science and Technology. Cambridge: Cambridge University Press. doi:10.1017/9781316882665. ISBN 978-1-316-88266-5.
- Weimer DL, Vining AR. Policy analysis: Concepts and practice. Taylor & Francis; 2017 Mar 31.