Datasets:

officeblues
/

misery-data

license: cc-by-4.0
language:
  - en
pretty_name: Office Blues — Misery Data
tags:
  - labor-statistics
  - wages
  - bls-oews
  - burnout
  - open-data
  - united-states
  - economics
size_categories:
  - n<1K
source_datasets:
  - original
configs:
  - config_name: occupation_wages
    data_files: data/occupation_wages.csv
  - config_name: city_burnout_index
    data_files: data/city_burnout_index.csv

Misery Data — the cost of work, in numbers

Two small, clean, public datasets on what work actually costs you — maintained by Office Blues, a tools-and-data project for the unhappily employed. Built from federal sources, free to reuse, easy to cite.

CC BY 4.0 — quote a figure, build on it, cite us. No scraping mystery; every number traces to a public source.

📦 What's inside

Config	Rows	What it is
`occupation_wages`	823	Annual wage percentiles for U.S. detailed occupations (BLS OEWS, May 2025)
`city_burnout_index`	50	A composite burnout score (0–100) for the 50 largest U.S. metros

Both ship as CSV and JSON, plus a Frictionless Data datapackage.json (typed schemas + licensing).

🚀 Quickstart

from datasets import load_dataset

wages = load_dataset("officeblues/misery-data", "occupation_wages", split="train")
burnout = load_dataset("officeblues/misery-data", "city_burnout_index", split="train")

print(wages[0])
# {'soc': '15-1252', 'title': 'Software Developers', 'median_usd': 132270, ...}

Prefer pandas?

import pandas as pd
url = "https://huggingface.co/datasets/officeblues/misery-data/resolve/main/data/occupation_wages.csv"
df = pd.read_csv(url)
df.nlargest(10, "median_usd")[["title", "median_usd"]]

🗂️ Schema

`occupation_wages` — U.S. occupation annual wages (n = 823)

field	type	meaning
`soc`	string	Standard Occupational Classification (SOC 2018) code
`title`	string	Occupation title
`p25_usd` / `median_usd` / `p75_usd` / `p90_usd`	int	Annual wage percentiles, USD
`source_url`	string	The Office Blues page for this occupation

Source: U.S. Bureau of Labor Statistics, Occupational Employment and Wage Statistics (OEWS), May 2025 release. National figures.

`city_burnout_index` — City burnout index (n = 50)

field	type	meaning
`cbsa_code`	string	Census CBSA code
`city` / `state`	string	Metro area / state
`population`	int	Metro population
`burnout_score`	int	Composite score, 0–100 (higher = worse)
`rank` / `of_cities`	int	Rank (1 = worst) of 50
`source_url`	string	The Office Blues page for this metro

Methodology: the burnout score composites public metro-level indicators (pay-to-cost-of-living gap, commute time, unemployment). Full method: officeblues.net/methodology. Treat it as an editorial index (v1), not an official statistic.

📑 Citation

DOI: 10.5281/zenodo.20534485 — the concept DOI, always resolving to the latest version (archived on Zenodo).

@misc{officeblues_misery_data,
  author       = {Office Blues},
  title        = {Misery Data: U.S. occupation wages and a city burnout index},
  year         = {2026},
  publisher    = {Office Blues},
  doi          = {10.5281/zenodo.20534485},
  url          = {https://officeblues.net},
  note         = {CC BY 4.0}
}

ℹ️ About

Maintained by Office Blues — https://officeblues.net. The methodology and the tools behind these numbers (the Meeting Tax Calculator, the Salary Negotiation Script, the daily pulse) live there. Aggregates only — no personal or per-visitor data. This is the canonical mirror of github.com/officeblues/misery-data.

Tools, data & receipts for the unhappy employed.