- What is the R-Word Index?
- The R-Word Index is a daily 0–100 signal that estimates how strongly recession-related concern is appearing in everyday online conversation and search behavior. It is designed to reflect the social pulse around recession, not to declare whether an official recession has started.
- What does it actually measure?
- It measures how frequently and intensely people mention or search for recession and related economic distress in their everyday online activity. The system draws from multiple sources — including community discussions about economics, jobs, personal finance, and daily life, as well as search trend data for recession-related terms — scoring each relevant piece of content for both relevance and concern intensity.
- Is this an official economic indicator?
- No. The R-Word Index is an experimental public signal. It is not a government statistic, not a recession declaration, and not investment advice. It should be read as an indicative human-language index.
- How does this relate to The Economist's R-Word Index?
- The Economist's R-Word Index, introduced in the early 2000s, counts how often the word "recession" appears in major newspapers over a given quarter. It demonstrated a powerful insight: media frequency of the word "recession" often rises ahead of officially declared downturns. Our Daily R-Word Index builds on that same core idea but takes it further. Instead of counting mentions in a handful of elite publications on a quarterly basis, we scan multiple sources of everyday human online activity every day, and we go beyond simple word counts to measure how intense and urgent the concern actually sounds. The result is a higher-resolution daily signal that captures the lived texture of recession anxiety across ordinary people's conversations and search behavior.
- What sources do you use?
- The index draws from sources that reflect authentic human expression rather than official news or institutional commentary. This includes community discussion platforms where people talk about their economic lives — jobs, personal finance, housing, daily expenses — and search trend data that shows how actively the broader population is searching for recession-related topics. The mix captures both the depth of individual experience and the breadth of public awareness.
- How is the index calculated?
- Each piece of collected content is analyzed by an AI model that evaluates two factors: how directly it relates to recession (relevance) and how worried the author sounds (concern intensity). The AI filters out historical references, metaphorical uses, and general economic commentary that isn't about recession concern. Each source is scored independently, then combined using configured weights. The daily score is the normalized result, smoothed to reduce noise.
- Why do past index values sometimes change?
- We continuously refine our scoring — tuning AI prompts, adjusting weights, and fixing bugs. When we make a meaningful change, we re-score the entire historical record so the chart reflects a single consistent methodology end to end. A value you saw for a past date may read slightly differently when you revisit it — we consider this more honest than freezing an outdated number. Every row carries a methodology version stamp, and our public methodology changelog lists every re-score with the date and how much historical values moved.
- Does a high reading mean a recession has started?
- Not necessarily. A high reading means recession concern is widespread and intense in public online discussion and search behavior. That can happen before, during, or after an official downturn.
- Why can the index move even if the economy has not obviously changed?
- Because the index measures public expression and search behavior, not only official macroeconomic facts. It can respond to layoffs, inflation fears, election uncertainty, tariff announcements, or changing narratives before those show up clearly in official data.
- How do you prevent one source from dominating the index?
- Each data source is scored independently, producing its own average score for the day. The final daily index is computed using configured weights for each source, regardless of how many individual data points it contributes. Search trends carry the highest weight as they reflect the broadest population signal, while community discussions provide depth of personal experience. This ensures that raw volume does not determine influence.
- How often is it updated?
- The public site updates daily. Under the hood, collection and scoring run in batches, but the public index is published once per day for consistency.
- Is there a mobile app?
- Yes! R-Word Index is available as a free app on iOS. You can download it from the App Store. An Android version is currently under development.