MSA Safety — U.S. Public Procurement Analysis

Research project analyzing MSA Safety Incorporated (NYSE: MSA, formerly Mine Safety Appliances Co.) participation and wins in U.S. public procurement (federal contract tenders) over 2021–2026. Built as supporting evidence for an investment thesis.

🔗 Live site: https://slogatskiy.github.io/msa-procurement-analysis/

Headline: 542 federal prime contract awards · $143.5M (2021–2026) · Fire/SCBA 73%, Detection 23% · 10 agencies · 26 states. Plus: $86.5M won in open competition vs rivals · $4.8M direct state/local payments across 19 jurisdictions · ~$1.8B national FEMA AFG demand pool · GAO bid-protest record vs 3M Scott & Draeger.

What's here

Path	Contents
ROADMAP.md	Project plan & progress checklist
scripts/	Re-runnable Python data collectors (USAspending.gov API)
data/raw/	Verbatim API pulls (never hand-edited)
data/processed/	Normalized CSV/JSON used by the site
site/	Static GitHub Pages site (the shareable deliverable)
docs/	Analysis notes & findings

Data sources

• USAspending.gov API — federal prime contract awards + competition fields (primary).
• Socrata government open-data checkbooks — automated sweep of state/county/city vendor-payment portals for payments booked to MSA (state & local footprint).
• FEMA Assistance to Firefighters Grants (AFG), Assistance Listing 97.044 — SCBA demand proxy.
• GAO bid-protest decisions — hand-verified head-to-head cases vs competitors.

Collectors (v3 — fire segment focus)

python3 scripts/collect_federal.py    # federal prime awards + competition (filtered to fire in build)
python3 scripts/collect_municipal.py  # municipal fire sweep (Socrata award/PO datasets, MSA+dealers vs competitors)
python3 scripts/collect_afg.py        # FEMA AFG grant volume by year (fire demand proxy)
python3 scripts/build_site.py         # -> fire-only site_data.json + site/data.js

See docs/municipal-landscape.md for the bidding-platform map, MSA dealer/co-op channel, and data-access reality.

Refresh / backup

All data is reproducible from the API — no manual entry. To refresh:

python3 scripts/collect_federal.py    # re-pulls awards into data/

Then rebuild/commit the site. Because everything regenerates from scripts, a full backup is just this git repo.

Note: MSA_Thesis.docx (the private thesis) is intentionally git-ignored and never published to this public repo.