Predicting Recessions in Real-Time: Mining Google Trends and Electronic Payments Data for Clues
Why this work is in the frame
A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.
Bibliographic record
Abstract
Many official economic indicators are released with a time lag, released infrequently and often require revision. In this Commentary, I discuss new sources of electronically recorded data that are both timely and reflect the real-time intentions of millions (or billions) of agents. Specifically, I consider whether Google searches and the growth of electronic payments variables, such as debit and credit card transactions, would have predicted the 2008 – 2009 recession. Not too long ago, Canadian empirical macroeconomic researchers would have to wait two months for the release of the monthly National Accounts in order to update their models and forecasts. However, in the last 10 years to 20 years, technological innovations have resulted in vast amounts of other data being recorded electronically and stored. New data series are now being generated at a rate faster than analysts can study them. Due to the emergence of Google as the dominant search engine, its search-term usage can provide a snapshot of current group interests in numerous issues, such as economics, politics, health, etc. In principle, if many people are entering the same economic search terms, this could provide a clue about changing conditions, such as the onset of a recession. I find that the usage of Google search terms “recession” and “jobs” could have predicted the recession up to three months in advance of its onset. However, since Google query data are only available from 2004, the time span studied in this Commentary is very short in the context of business cycles. Consequently, our study should be viewed as illustrative of the potential uses of electronic data. I also highlight the benefits and pitfalls that users of Google data may encounter in the context of economic monitoring.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.001 |
| Open science | 0.000 | 0.001 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it