What it is<\/strong> Source<\/strong> Size\/scope<\/strong> How it was built<\/strong> Links<\/strong> Dataset: https:\/\/www.kaggle.com\/datasets\/adizz321\/us-job-market-shift-since-2020-postings-by-sector<\/a> Notebook with visualizations (recovery-speed chart, sector winners\/losers, state-level breakdown): https:\/\/www.kaggle.com\/code\/adizz321\/the-job-market-inversion-what-changed-since-2020<\/a><\/p>\n<\/div>\n submitted by \/u\/Firm_Ad9420 <\/a>
Daily and sector-level tracking of the US job market since February 1, 2020 (pre-pandemic baseline) \u2014 how hiring demand shifted through COVID, the 2021\u201322 boom, and the tech slowdown after. Includes a “sector summary” file with the headline numbers already computed for you: net change since baseline, pandemic trough\/peak, and days to recovery per sector.<\/p>\n
Underlying job postings data is from Indeed Hiring Lab’s Job Postings Tracker<\/a>, licensed CC BY 4.0. I built and included an original sector-to-SOC-2018-major-group crosswalk (Indeed’s category system doesn’t map 1:1 to the standard occupational taxonomy, so this fills that gap) \u2014 that mapping is my own work, not part of Indeed’s data.<\/p>\n
44 occupational sectors, all 51 states (incl. DC), daily observations from Feb 1, 2020 through today. National, sector-level, and state-level files, plus the pre-computed summary stats file.<\/p>\n
Fully scripted pipeline (Python\/pandas) \u2014 one command downloads the raw data and rebuilds every output file from scratch. Nothing manually edited or copy-pasted, so it’s reproducible and easy to refresh later since Indeed updates their source weekly.<\/p>\n
[link]<\/a><\/span> [comments]<\/a><\/span><\/p>
