← All posts · May 10, 2026 · 5 min read

What is Census BPS data, and why should trade contractors care

If you've ever read a real-estate or construction industry report that quotes "permits issued in the latest month," it almost always cites the US Census Bureau's Building Permits Survey — universally abbreviated as BPS. It's the authoritative source for new-construction permit activity in every US county, and it's free.

For trade contractors and suppliers, BPS data isn't just a stat to quote — it's a live signal of where new single-family construction is happening, at scale, in near-real-time. Understanding it is one of the cheapest ways to make better territory and account decisions.

What BPS is

The Building Permits Survey is a monthly survey conducted by the US Census Bureau. The agency contacts every "permit-issuing place" in the country — cities, towns, counties — and collects their permit totals for the month, broken into four categories:

• Single-family (1 unit per building)
• Duplex (2 units per building)
• Triplex/fourplex (3-4 units)
• Apartments and condos (5+ units)

For each category and each county, the survey publishes: number of buildings permitted, number of dwelling units across those buildings, and total declared construction value. The data goes back decades.

How it gets to you

The Census Bureau publishes the data in monthly CSV files at https://www2.census.gov/econ/bps/County/. Each file is named coYYMMc.txt (e.g. co2602c.txt = February 2026 county data). The file is ~400KB, plain CSV, no auth.

That means anyone with a script and a few hundred KB of bandwidth can pull the latest month for every US county the day it's published. Most analysts cite "the latest BPS release"; you can read the actual file directly.

What it doesn't tell you

BPS is aggregate. It tells you "Hillsborough County issued 305 single-family permits in February 2026." It does not tell you:

• Which builders pulled those permits
• The street addresses of the permitted lots
• Which subdivisions are involved
• Whether any specific lot has been sold

For per-permit detail, you need to look at the individual county's permit portal (Accela, ArcGIS, or whatever CMS they use). That's a separate, much messier data layer with different APIs per county. But BPS gives you the trustworthy aggregate to anchor everything else against.

Why this matters for trade operators

If you're an LBM operator, roofing contractor, HVAC company, plumber, or any other trade serving new single-family construction, BPS gives you four things you can't get easily anywhere else:

1. Comparable signal across every US county

Different counties use different permit portals, with different fields, different lag, and different quality. BPS normalizes that — the Census Bureau aggregates everything to a comparable monthly number. You can compare Hillsborough FL directly to Harris TX directly to Maricopa AZ without worrying about source differences.

2. Reliable trend detection

Per-permit data is messy month-to-month — a single big subdivision plat can throw off your numbers. BPS smoothes the noise because it aggregates every permit-issuing entity in the county. Trend analysis on BPS is more stable than trend analysis on raw permit data.

3. Free benchmarking

"Is Tampa hot right now?" "Is Phoenix cooling?" BPS gives you the data to answer these without subscribing to Zonda's $25k/yr feed.

4. Counties without per-permit data

Most US counties don't publish a permit portal you can scrape. BPS is the only systematic data source for those counties. If your service area includes 12 counties and 8 of them have no Accela/ArcGIS feed, BPS is your only signal.

BPS is the floor of your data stack, not the ceiling. Per-permit detail and builder identity add value on top — but if you don't have BPS, you're guessing at trends and dependence on which scrapers happened to work this month.

The gotchas

Lag. BPS publishes monthly with a ~45-day lag. February data comes out mid-April. That's fast enough for territory planning, slow enough that you can't react to individual permit events through it alone.

1-4 vs 5+. Single-family, duplex, triplex/fourplex, and apartments are all separate categories. If you sell into single-family only (most LBM, most trades), filter to the 1-4 unit categories. If you don't, you'll see apartment-tower permits inflate the count for urban counties.

Survey vs census. BPS is a survey — large permit-issuing places report monthly; smaller places report less frequently. The Census Bureau imputes missing data. For most counties this is fine; for very rural ones the numbers are estimated. Take the third decimal place with a grain of salt.

"Authorized" vs "issued." BPS measures permits authorized — meaning approved for issuance. Most permits get issued shortly after authorization, but a small percentage are authorized and then never issued. The numbers are very close but not identical to "issued" totals you'd see in a permit portal.

How we use it at PermitFocus

BPS is the floor of our data stack. We pull the latest monthly file every day, normalize it across every US county, and use it to:

• Score every county 0-100 against the trade focus you select
• Detect YoY momentum (current 12 months vs prior 12 months)
• Compare counties apples-to-apples for territory and acquisition decisions
• Forecast next 12 months using Holt-linear smoothing on 36 months of history

On top of BPS we layer per-permit detail (from county ArcGIS / Accela) where available, the FL DBPR licensee registry for builder identity, and BEA migration data for demographic context. But BPS is what makes the system work in every US county on day one — even the 2,400+ counties where per-permit data isn't publicly accessible.

You can browse all 3,022 US counties we track, each with its trailing-12mo BPS-derived volume. Or see a full sample county report with no signup.