News
Daily News Sentiment Index – San Francisco Fed
The Daily News Sentiment Index is a high-frequency measure of economic sentiment based on lexical analysis of economic-related news articles. The index is described in Buckman, Shapiro, Sudhof and Wilson (2020) and based on the methodology developed in Shapiro, Sudhof and Wilson (2020).
Shapiro, Sudhof, and Wilson (2020, hereafter SSW), construct sentiment scores for economics-related news articles from 24 major US newspapers compiled by the news aggregator service Factiva. Newspapers cover all major regions of the country, including some with broad national coverage, such as the New York Times and the Washington Post. SSW uses articles with at least 200 words where Factiva identified the article topic as “economy” and the country topic as “United States”. Combining publicly available lexicons with a news-specific lexicon constructed by the authors, the study develops a sentiment scoring model tailored specifically for newspaper articles.
The SSW aggregates the scores of individual articles into a daily time-series measure of news sentiment, relying on a statistical adjustment that takes into account changes over time in the composition of the sample across newspapers. The Daily News Sentiment Index is constructed as a weighted time series average, with weights that decrease geometrically with time since the article was published. The data here will be updated regularly on a weekly basis.
Chart 1: Daily News Sentiment Index: Historical View
Note: Graph 1 shows moving averages of daily news sentiment scores since 1980; higher values indicate more positive sentiment and lower values indicate more negative sentiment. Gray bars indicate NBER recession dates.
Chart 2: Recent daily news sentiment
Note: Chart 2 is enlarged to show data from the most recent 18 months.
References
Buckman, Shelby R., Adam Hale Shapiro, Moritz Sudhof, and Daniel J. Wilson. 2020. “News sentiment in the time of COVID-19.” FRBSF Economic Letter 2020-08 (April 6).
Shapiro, Adam Hale, Moritz Sudhof, and Daniel J. Wilson. 2020. “Measuring news sentiment.” FRB San Francisco Working Paper 2017-01.
Download data
Daily News Sentiment Data (Excel document, 377 kb)
Replication code (zip file 1.8mb)
Contact Aditi.Poduri (at) sf.frb.org