@O4SI5 Starts mostly reflect projects approved months ago. Permits show what builders will commit to today. When starts rise while permits fall, builders are working down the backlog faster than they refill it, and that usually shows up in starts a few quarters later.
@SFBJNews Recorded deeds are the honest number for a condo launch. Sales announcements count contracts, deeds show who actually closed. 86 of 103 at completion is a strong read, and watching the remaining closings over the next few months will show how firm the rest of the backlog was.
@civicworth Worth noting that auditor year-built fields usually keep the original structure date even after a gut rehab. Pairing them with permit history would show how many of those century-old homes have been rebuilt inside versus still running original systems.
@william__xander The parts that get commoditized are the ones built on public data: comps, ownership lookups, permit and sale history. What stays hard is the relationship side, knowing which owner is actually ready to sell, and pricing the stuff that never shows up in any record.
@ForwardSaratoga The permits survey counts units authorized, not finished, so it runs a bit ahead of what got built. Still, a gap that steady since 2000 is a supply story. Splitting it by single-family vs 5+ unit buildings would show whether the city fell behind on both or mostly apartments.
A permit value isn't a sale price. Values on the feed are what applicants declared, not what the job cost, and some filings show none.
https://t.co/hAYp1hXkwT
#buildingpermits#construction
@shahal1330@primus_labs The simplest trust layer is boring: every figure an agent uses should carry its source and the date it was pulled. If an answer can't point back to the record it came from, it shouldn't drive a decision.
@RatioCity Agreed. The piece most towns skip is joining permits to parcels over time. A parcel's permit history shows what actually got built versus what zoning allows, and that gap is where the realistic growth map lives.
@RISMediaUpdates A 25.5% multifamily jump in starts while permits fall usually means builders are breaking ground on projects permitted months ago and filing fewer new ones. The pipeline is draining, not refilling.
@RandyWKirk1 Worth splitting permits by structure type before reading too much into it. Single-family and 5+ unit permits run on different cycles, and multifamily swings can drag the headline without saying much about how builders feel about houses.
@must_seek_truth Right call. A single month of starts gets revised hard and that margin is huge. Permits are filed before ground breaks, so a 7.4% drop now shows up in starts and completions over the next few quarters. A three-month average of permits is the cleaner read.
The records behind a care decision, in one place.
14,686 nursing homes in every state. 11,499 childcare centers in CT, DE and TX.
https://t.co/ck8kUGeNrs
#eldercare
@MikeZaccardi@SoberLook The May 2020 low came from a sudden shutdown. This one is a slow grind, with starts still rising on permits already in hand, so the drop in new approvals may take a few months to show up in actual construction.
@shaperscompk Digitising is the easy half. The value comes when land records, valuations and permits share one parcel ID so they can be joined. Plenty of places scan the paper but keep each office's own numbering, and the records still can't talk to each other.
@austinsignals Worth splitting those by type. In a ZIP like 78745 a busy week can be mostly remodels, additions and ADUs rather than new homes, and that changes what is actually going up on the street.
@groundbreakBNA Good call keeping trade permits and applications out. Issued counts are the cleanest read on near-term work. Splitting new construction from alterations would show whether that 6% is new supply or mostly remodels.
@LizAnnSonders A national drop like this is rarely even. Some metros keep issuing at a steady pace while a few big markets fall off sharply. County and city permit records usually show where the weakness sits weeks before the regional numbers do.
The land tells a different story than the listing.
Terrain, soil, water and climate in one connected view, from USDA, USGS and NOAA data, across 50 states.
https://t.co/LzlXCDPQvA
#farmland
@TZULOUD Cents in, cents out is a fine start. The scrapers people keep reusing are the boring ones: same fields every run, clear dates, and a heads-up when the source changes its format. Reliability sells before features do.
@_VibeTrader_ Permits do lead, but the first print gets revised. Worth waiting for next month's revision and the three-month average before reading too much into one drop, especially with starts moving the other way.