Climate Policy Uncertainty Indexes

Annual Data          Monthly Data          Annotated Index

The authors construct indices of climate policy uncertainty (CPU) for 11 countries adapting the methodology of Baker, Bloom, and Davis (2016). They compute a CPU index for Australia, Canada, France, Germany, Ireland, Italy, New Zealand, Mexico, Spain, the United Kingdom, and the United States. Depending on newspaper availability across countries, the series start in the 1990s or early 2000s and go up to 2019. The authors develop a 3-block search strategy to count the number of articles talking about CPU. Each selected article must contain a word or expression in each block:

  • Climate: general environmental terms (e.g. climate change, environmental, pollution, carbon, greenhouse gas) or specific technologies (e.g. wind power, electric cars, hydro)
  • Policy: e.g. policy, regulation, tax, cap and trade, standards, subsidies
  • Uncertainty: uncertainty, uncertain, vague, unclear

Next, for each newspaper, they compute the ratio of articles talking about CPU to the number of total articles published by each newspaper in a month. For instance, the U.S. CPU index is based on counts from the Wall Street Journal, the New York Times, and the Washington Post. This ratio is then divided by the newspaper-specific standard deviation across all years. This creates a newspaper-specific time series with unit standard deviation across the entire time interval. These standardized series are then averaged across all newspapers within each country by year. Last, the country-specific series are normalized to a mean of 100 over the time interval.

In addition, the authors also develop 2 sub-indices to capture whether climate policy uncertainty stems from potential policy weakening (CPU-) or potential policy strengthening (CPU+). The authors follow the same methodology to build the subindices and add a 4th block to the search strategy to capture the direction of climate policy uncertainty. This 4th block includes terms like "loosen", "weaken" and "lower" for CPU- and "tighten", "strengthen", and "raise" for CPU+. The sub-indices are available for Australia, Canada, the United Kingdom, and the United States, for the same timeframes as the main index.

Monthly and yearly data can be downloaded above for CPU, CPU-, and CPU+. Data will be regularly updated on this page. Yearly data for CPU up to 2018 are also available on the OECD data explorer. Additional information, including the full text of search strategies in different languages, is available here for the U.S. and in this OECD Working Paper for other countries.

The dataset should be cited as follows: Basaglia, P, C Berestycki, S Carattini, A Dechezleprêtre and T Kruse (2025), 'Climate Policy Uncertainty and Firms' and Investors' Behavior', CEPR Discussion Paper No. 20160. CEPR Press, Paris & London. https://cepr.org/publications/dp20160.