"Surprises" in BLS Jobs Revisions Became More Frequent After 2020
3 days ago
- #BLS
- #Data Analysis
- #Economic Indicators
- BLS jobs revisions have become more frequent and unpredictable post-2020, indicating a more difficult estimation landscape.
- A two-component mixture model identifies two distributions for revisions: 'surprise' revisions and normal revisions, with the former having a much larger standard deviation.
- The proportion of surprise revisions is estimated at 11.3%, translating to about 1.4 surprises per year, challenging the notion of them being rare 'inflection points.'
- Post-2020, distinguishing between surprise and normal revisions has become harder, with surprise probabilities more evenly distributed and slightly higher on average.
- Former BLS Commissioner Erica Groshen emphasizes the importance of BLS surveys for public knowledge and decision-making but acknowledges challenges like declining response rates.
- The article suggests the BLS could improve communication about the nature and challenges of its jobs reports to maintain public trust and understanding.