Housing Units Permitted by County, 2020-2025

What follows is the transcript of this chat as State Planner generated it. It is here to show how the answer above was arrived at: every query State Planner ran, what each one returned, and where the data came from.

transcript.md

↑ Contents

State Planner chat · https://stateplanner.com

State: New Mexico (NM)
Started: November 5, 2026, 5:30 PM UTC
Runtime: 52 sec
Status: completed

Prompt

New Mexico Building Permits Issued by County: Where Is New Housing Going, 2020-2025?

Model: Anthropic: Claude Opus 5.5 (anthropic/claude-opus-5.5, via openrouter)

Response

Where New Mexico's new housing permits went, 2020–2025

Building permits authorized 45,435 new privately owned housing units in New Mexico from 2020 through 2025. These are Census Building Permits Survey estimates, which include the Bureau's figures for permit offices that didn't report. Four counties account for 78% of the units: Bernalillo, Sandoval, Doña Ana and Santa Fe.

Housing Units Permitted by County, 2020-2025

How to read the map: counties are shaded by the total units permitted from 2020 to 2025:

Dark navy: 5,000 or more
Blue: 1,000–4,999
Light blue: 300–999
Pale blue: 1–299
Grey: no permits in the Census county file

The four largest totals are labelled on the map. Click a county to see how its units split by building size and its units per 1,000 residents.

Housing units authorized by permit, by county, 2020–2025

County	Units	Share of state	Single-family	2–4 unit bldgs	5+ unit bldgs	Units per 1,000 residents*
Bernalillo	8,998	27.0%	5,206	96	3,696	13.5
Sandoval	6,984	21.0%	5,645	92	1,247	43.8
Doña Ana	6,210	18.7%	5,766	285	159	27.1
Santa Fe	3,672	11.0%	1,443	14	2,215	23.4
Valencia	1,445	4.3%	1,445	0	0	17.6
Eddy	1,375	4.1%	956	41	378	22.0
Lea	988	3.0%	988	0	0	13.2
Curry	863	2.6%	362	287	214	18.5
Taos	713	2.1%	701	4	8	20.6
San Juan	613	1.8%	592	0	21	5.1
Lincoln	513	1.5%	507	6	0	25.9
Chaves	486	1.5%	486	0	0	7.7
Los Alamos	165	0.5%	165	0	0	8.5
Roosevelt	134	0.4%	134	0	0	7.2
Luna, Otero, McKinley, Sierra, Grant, San Miguel	127 combined	0.3%	—	—	—	0.1–2.0

*Six-year units divided by the 2025 Census county population estimate.

What stands out

Sandoval County (Rio Rancho and the Albuquerque metro's west side) permitted the most housing relative to its size: about 44 units per 1,000 residents, roughly three times Bernalillo's rate. Only Bernalillo permitted more units in total.
Two counties got most of the apartment permits. Statewide, about 81% of units were single-family and 17.5% were in buildings of five or more units. Bernalillo (3,696) and Santa Fe (2,215) together hold three-quarters of those larger-building units. Santa Fe is the only county where buildings of five or more units outnumber single-family homes (60% of its units).
Doña Ana (Las Cruces) is almost entirely single-family: 93% of its 6,210 units.
Outside the metro areas, the largest totals are in Eddy and Lea (the Permian Basin oil region) and in Valencia, south of Albuquerque. Lincoln and Taos have high rates for their size, at 21–26 units per 1,000 residents.

Year by year

Housing units permitted per year, top counties

The state total went from 5,219 units in 2020 to a peak of 8,633 in 2022. It stayed near that in 2023 (8,511), then settled around 7,600–7,700 in 2024 and 2025. The 2021 jump came mostly from buildings of five or more units, which rose from 760 units to 2,203. Single-family permits peaked in 2022 at 7,155.

Caveats

Thirteen counties have no rows in the Census file for these years: Rio Arriba, Cibola, Socorro, Torrance, Colfax, Quay, Mora, Guadalupe, Union, Hidalgo, Catron, De Baca and Harding. Sierra appears in only 2 of the 6 years and Grant in 4. Missing rows most likely mean the area has few or no permit-issuing offices, so construction there isn't counted. They don't mean nothing was built.
Some housing isn't counted at all: manufactured (mobile) homes, public housing, additions and conversions are left out. That matters in rural New Mexico, where many new homes are manufactured homes.
A permit isn't a finished home: it authorizes construction, and some permitted units are never built.
About the dollar figures: construction values in the map pop-ups and the data are nominal dollars as written on the permits (for example about $1.8 billion in Bernalillo). They aren't adjusted for inflation and are rough for comparing counties.

If you want to look below the county level, I can break this down by city or permit office (Albuquerque, Rio Rancho, Las Cruces, Santa Fe and others).

Chart Queries

Housing units permitted per year, top counties

   SELECT YEAR AS "Year",
          CASE
                    WHEN geoid IN ('35001', '35043', '35013', '35049', '35061', '35015') THEN replace(replace(name, ' County', ''), 'Dona', 'Doña')
                    ELSE 'All other counties'
          END AS "County",
          sum(units_total) AS "Units permitted"
     FROM nm_bps_county_permits
    WHERE YEAR BETWEEN 2020 AND 2025
 GROUP BY 1,
          2
 ORDER BY 1,
          2

charts/turn-1-permits-by-year-county.csv

Picture: charts/turn-1-permits-by-year-county.png

Exploratory Queries

Query 1 — Check which years 2020-2025 exist and county coverage.

   SELECT YEAR,
          count(*),
          sum(units_total) u,
          sum(units_1) sf,
          sum(units_5plus) mf5
     FROM nm_bps_county_permits
    WHERE YEAR >= 2019
 GROUP BY YEAR
 ORDER BY YEAR

year	count	u	sf	mf5
2019	23	5020	4285	702
2020	21	5219	4418	760
2021	21	7753	5465	2203
2022	20	8633	7155	1427
2023	20	8511	6835	1529
2024	19	7572	6488	770
2025	19	7747	6296	1264

data/turn-1-query-1.csv

Query 2 — County totals 2020-2025 with share, mix and per-1,000-resident rate.

     WITH p AS (
             SELECT geoid,
                    sum(units_total) u,
                    sum(units_1) sf,
                    sum(units_2 + units_3_4) small,
                    sum(units_5plus) mf,
                    count(*) yrs,
                    sum(value_total) val
               FROM nm_bps_county_permits
              WHERE YEAR BETWEEN 2020 AND 2025
           GROUP BY geoid
          ),
          pop AS (
             SELECT geoid,
                    pop_estimate
               FROM nm_census_county_pop_est
              WHERE YEAR = (
                       SELECT max(YEAR)
                         FROM nm_census_county_pop_est
                    )
          )
   SELECT c.geoid,
          c.name,
          p.u,
          p.sf,
          p.small,
          p.mf,
          p.yrs,
          round(p.val / 1e6) val_m,
          pop.pop_estimate,
          round(1000.0 * p.u / pop.pop_estimate, 1) per1k,
          round(100.0 * p.u / sum(p.u) OVER (), 1) pct
     FROM nm_census_counties c
LEFT JOIN p ON p.geoid = c.geoid
LEFT JOIN pop ON pop.geoid = c.geoid
 ORDER BY p.u DESC NULLS LAST

geoid	name	u	sf	small	mf	yrs	val_m	pop_estimate	per1k	pct
35001	Bernalillo County	8998	5206	96	3696	6	1819	667601	13.5	27.0
35043	Sandoval County	6984	5645	92	1247	6	1375	159565	43.8	21.0
35013	Doña Ana County	6210	5766	285	159	6	1588	229091	27.1	18.7
35049	Santa Fe County	3672	1443	14	2215	6	804	156907	23.4	11.0
35061	Valencia County	1445	1445	0	0	6	416	82013	17.6	4.3
35015	Eddy County	1375	956	41	378	6	335	62509	22.0	4.1
35025	Lea County	988	988	0	0	6	275	74749	13.2	3.0
35009	Curry County	863	362	287	214	6	180	46655	18.5	2.6
35055	Taos County	713	701	4	8	6	191	34564	20.6	2.1
35045	San Juan County	613	592	0	21	6	148	120340	5.1	1.8
35027	Lincoln County	513	507	6	0	6	220	19844	25.9	1.5
35005	Chaves County	486	486	0	0	6	94	63364	7.7	1.5
35028	Los Alamos County	165	165	0	0	6	41	19407	8.5	0.5
35041	Roosevelt County	134	134	0	0	6	20	18615	7.2	0.4
35029	Luna County	38	38	0	0	6	8	25407	1.5	0.1
35035	Otero County	25	25	0	0	6	9	70368	0.4	0.1
35031	McKinley County	25	16	0	9	6	4	68119	0.4	0.1
35051	Sierra County	23	17	0	6	2	6	11395	2.0	0.1
35017	Grant County	13	13	0	0	4	2	27252	0.5	0.0
35047	San Miguel County	3	3	0	0	6	1	26260	0.1	0.0
35037	Quay County							8454
35039	Rio Arriba County							39832
35019	Guadalupe County							4343
35006	Cibola County							26807
35007	Colfax County							12184

25 out of 33 are listed here.

data/turn-1-query-2.csv

Query 3 — Confirm population estimate year used for per-capita rates.

   SELECT max(YEAR)
     FROM nm_census_county_pop_est

max
2025

data/turn-1-query-3.csv

Map Layer Queries

Units permitted 2020-2025 — colored by "cls": 5,000+ units (#08306b), 1,000-4,999 units (#2171b5), 300-999 units (#6baed6), 1-299 units (#c6dbef), No permits in Census file (#eeeeee)

     WITH p AS (
             SELECT geoid,
                    sum(units_total) u,
                    sum(units_1) sf,
                    sum(units_2 + units_3_4) small,
                    sum(units_5plus) mf
               FROM nm_bps_county_permits
              WHERE YEAR BETWEEN 2020 AND 2025
           GROUP BY geoid
          ),
          pop AS (
             SELECT geoid,
                    pop_estimate
               FROM nm_census_county_pop_est
              WHERE YEAR = 2025
          )
   SELECT c.geom,
          c.name AS "County",
          coalesce(p.u, 0) AS "Units permitted 2020-2025",
          p.sf AS "Single-family units",
          p.small AS "2-4 unit bldg units",
          p.mf AS "5+ unit bldg units",
          round(1000.0 * coalesce(p.u, 0) / pop.pop_estimate, 1) AS "Units per 1,000 residents(2025 pop)",
          CASE
                    WHEN p.u IS NULL THEN 'none'
                    WHEN p.u >= 5000 THEN 'a'
                    WHEN p.u >= 1000 THEN 'b'
                    WHEN p.u >= 300 THEN 'c'
                    WHEN p.u >= 1 THEN 'd'
          END AS cls
     FROM nm_census_counties c
LEFT JOIN p ON p.geoid = c.geoid
LEFT JOIN pop ON pop.geoid = c.geoid

Picture: maps/turn-1-county-permits-2020-2025.png

Used Data Sets

nm_bps_county_permits — last synced 2026-10-14 · 785 rows
nm_census_counties — last synced 2026-09-27 · 33 rows
nm_census_county_pop_est — last synced 2026-10-15 · 1,838 rows

data/turn-1-query-1.csv

↑ Contents

year	count	u	sf	mf5
2019	23	5020	4285	702
2020	21	5219	4418	760
2021	21	7753	5465	2203
2022	20	8633	7155	1427
2023	20	8511	6835	1529
2024	19	7572	6488	770
2025	19	7747	6296	1264

7 rows.

data/turn-1-query-2.csv

↑ Contents

geoid	name	u	sf	small	mf	yrs	val_m	pop_estimate	per1k	pct
35001	Bernalillo County	8998	5206	96	3696	6	1819	667601	13.5	27.0
35043	Sandoval County	6984	5645	92	1247	6	1375	159565	43.8	21.0
35013	Doña Ana County	6210	5766	285	159	6	1588	229091	27.1	18.7
35049	Santa Fe County	3672	1443	14	2215	6	804	156907	23.4	11.0
35061	Valencia County	1445	1445	0	0	6	416	82013	17.6	4.3
35015	Eddy County	1375	956	41	378	6	335	62509	22.0	4.1
35025	Lea County	988	988	0	0	6	275	74749	13.2	3.0
35009	Curry County	863	362	287	214	6	180	46655	18.5	2.6
35055	Taos County	713	701	4	8	6	191	34564	20.6	2.1
35045	San Juan County	613	592	0	21	6	148	120340	5.1	1.8
35027	Lincoln County	513	507	6	0	6	220	19844	25.9	1.5
35005	Chaves County	486	486	0	0	6	94	63364	7.7	1.5
35028	Los Alamos County	165	165	0	0	6	41	19407	8.5	0.5
35041	Roosevelt County	134	134	0	0	6	20	18615	7.2	0.4
35029	Luna County	38	38	0	0	6	8	25407	1.5	0.1
35035	Otero County	25	25	0	0	6	9	70368	0.4	0.1
35031	McKinley County	25	16	0	9	6	4	68119	0.4	0.1
35051	Sierra County	23	17	0	6	2	6	11395	2.0	0.1
35017	Grant County	13	13	0	0	4	2	27252	0.5	0.0
35047	San Miguel County	3	3	0	0	6	1	26260	0.1	0.0
35037	Quay County							8454
35039	Rio Arriba County							39832
35019	Guadalupe County							4343
35006	Cibola County							26807
35007	Colfax County							12184
35053	Socorro County							15838
35011	De Baca County							1647
35057	Torrance County							16100
35059	Union County							3844
35003	Catron County							3827
35023	Hidalgo County							3929
35033	Mora County							4051
35021	Harding County							617

33 rows.

data/turn-1-query-3.csv

↑ Contents

max
2025

1 row.

Year	County	Units permitted
2020	All other counties	1071
2020	Bernalillo	1084
2020	Doña Ana	1109
2020	Eddy	448
2020	Sandoval	834
2020	Santa Fe	577
2020	Valencia	96
2021	All other counties	1425
2021	Bernalillo	1924
2021	Doña Ana	1338
2021	Eddy	129
2021	Sandoval	1789
2021	Santa Fe	840
2021	Valencia	308
2022	All other counties	3949
2022	Bernalillo	1820
2022	Doña Ana	1105
2022	Eddy	137
2022	Sandoval	783
2022	Santa Fe	590
2022	Valencia	249
2023	All other counties	3626
2023	Bernalillo	1294
2023	Doña Ana	861
2023	Eddy	264
2023	Sandoval	1281
2023	Santa Fe	926
2023	Valencia	259
2024	All other counties	3332
2024	Bernalillo	1316
2024	Doña Ana	964
2024	Eddy	156
2024	Sandoval	1278
2024	Santa Fe	244
2024	Valencia	282
2025	All other counties	3348
2025	Bernalillo	1560
2025	Doña Ana	833
2025	Eddy	241
2025	Sandoval	1019
2025	Santa Fe	495
2025	Valencia	251

42 rows.

maps/turn-1-county-permits-2020-2025.png

↑ Contents

provenance/nm_bps_county_permits.md

↑ Contents

Provenance: `nm_bps_county_permits`

This file describes how the rows in PostgreSQL table nm_bps_county_permits were obtained from an external source: where they were fetched and how fields were renamed and typed. Field meanings are in the mapping table below.

Identity

Table: nm_bps_county_permits
Family: BPS
Display name: Building Permits by County
Source URL: https://www2.census.gov/econ/bps/County/ (co<year>a.txt, 1990 on)
Geometry (source / stored): none / none
Update cycle: Yearly (each May)

Pipeline

The Building Permits Survey publishes one national county file per year, the annual summary (a), comma-delimited text with a two-row header. Every year from 1990 is loaded; the newest year is the one before the current, once the Bureau has published it. Each year's file is downloaded whole and kept; the sync keeps New Mexico's rows (FIPS State 35) and checks each file's survey date names its year (2024, or 9099 in the 1990s files).

A county listed twice in one year (the 2014 and 2015 files repeat four Alaska boroughs and one Maryland county under a second name with the same estimates) keeps the row with more reported units; two rows with different estimates fail the sync. The monthly and year-to-date files are not loaded.

Changed from the source: cut to New Mexico; reshaped to one row per county per year; the three totals added; the region, division and survey-date columns dropped.

No licence terms: a work of the U.S. Government, credited to the U.S. Census Bureau.

Field → column mapping

Source field	Column	Type	Description
FIPS State, FIPS County	`geoid`	text	Two digits and three, joined
County Name	`name`	text	Trailing spaces trimmed
Survey Date	`year`	text → integer	Checked against the file's year
1-unit Bldgs, Units, Value	`bldgs_1`, `units_1`, `value_1`	integer	Estimates with imputation
2-units Bldgs, Units, Value	`bldgs_2`, `units_2`, `value_2`	integer
3-4 units Bldgs, Units, Value	`bldgs_3_4`, `units_3_4`, `value_3_4`	integer
5+ units Bldgs, Units, Value	`bldgs_5plus`, `units_5plus`, `value_5plus`	integer
1-unit rep … 5+ units rep	`rep_bldgs_1` … `rep_value_5plus`	integer	Reported only
(computed)	`bldgs_total`, `units_total`, `value_total`		The four estimate sizes added
Region Code, Division Code	(dropped)

provenance/nm_census_counties.md

↑ Contents

Provenance: `nm_census_counties`

This file describes how the rows in PostgreSQL table nm_census_counties were obtained from an external source: where they were fetched and how fields were renamed and typed. Field meanings are in the mapping table below.

Identity

Table: nm_census_counties
Family: Census
Display name: Census Counties (New Mexico)
Source URL: https://tigerweb.geo.census.gov/arcgis/rest/services/TIGERweb/tigerWMS_Current/MapServer/82
Also from: https://www2.census.gov/programs-surveys/cfs/technical-documentation/geographies/list2022.xlsx
Geometry (source / stored): polygon / geometry(MultiPolygon, 4326)
Update cycle: Yearly

Pipeline

Two Census Bureau sources, joined on the county's 5-digit FIPS code:

Boundaries: TIGERweb tigerWMS_Current layer 82 ("Counties"), where=STATE='35', GeoJSON, outSR=4326, keyset-paged on GEOID.
CFS areas: sheet CFSAreas of the 2022 Commodity Flow Survey county list, the New Mexico rows (ST = 35). Only the 2022 columns are kept.

The two must name the same counties, or the sync fails. FAF's nm_faf_county_xwalk is built from this table: each county's cfs_geoid is the CFS22_ID of one FAF6 zone.

Field → column mapping

Source field	Column	Type	Description
TIGER `GEOID`	`geoid`	varchar(5)	state + county FIPS
TIGER `GEOID`, last 3	`county_fips`	varchar(3)	county FIPS
TIGER `NAME`	`name`	varchar(60)	e.g. `Bernalillo County`
CFS `CFS22_AREA`	`cfs_area`	varchar(5)	CSA/MSA code, or `99999`
CFS `CFS22_TYPE`	`cfs_area_type`	varchar(1)	C, M or R
CFS `CFS22_NAME`, whitespace collapsed	`cfs_area_name`	varchar(120)	the workbook pads some names with doubled and trailing spaces
CFS `CFS22_GEOID`	`cfs_geoid`	varchar(19)	e.g. `E330000US3599999999`
TIGER `AREALAND`	`aland_m2`	bigint	square meters
TIGER `AREAWATER`	`awater_m2`	bigint	square meters
TIGER `INTPTLAT`	`intptlat`	double precision	internal point
TIGER `INTPTLON`	`intptlon`	double precision	internal point
geometry	`geom`	geometry(MultiPolygon, 4326)	`ST_Multi(ST_GeomFromGeoJSON)`

Geometry

SRID 4326 (WGS84). GIST index on geom.

Change detection

The Census sync's check compares the layer's vintage ("January 1, 2026 vintage") and New Mexico's row count, and the county list's Last-Modified and length. A change to either marks Census stale, and the next Census sync rebuilds this table, then FAF's crosswalk.

Attribution

US Census Bureau. TIGER products are free to use with acknowledgement of the U.S. Census Bureau as the source; the CFS county list states no terms.

provenance/nm_census_county_pop_est.md

↑ Contents

Provenance: `nm_census_county_pop_est`

This file describes how the rows in PostgreSQL table nm_census_county_pop_est were obtained from an external source: where they were fetched and how fields were renamed and typed. Field meanings are in the mapping table below.

Identity

Table: nm_census_county_pop_est
Family: Census
Display name: Annual Population Estimates by County
Source URL: https://www2.census.gov/programs-surveys/popest/datasets/2020-2025/counties/totals/co-est2025-alldata.csv (vintage 2025; each vintage <V> is 2020-<V>/counties/totals/co-est<V>-alldata.csv), for 2020 on
History (2000–2019): under the same datasets/ directory, 2000-2010/intercensal/county/co-est00int-tot.csv and 2010-2020/intercensal/county/asrh/cc-est2020int-agesex-all.csv (the revised population), 2010/2010-eval-estimates/co-est2010-alldata.csv and 2010-2020/counties/totals/co-est2020-alldata.csv (the components of change of the releases they revised)
History (1970–1999): under https://www2.census.gov/programs-surveys/popest/tables/, 1900-1980/counties/asrh/co-asr-7079.csv, 1980-1990/counties/asrh/pe-02-<year>.csv and 1990-2000/intercensal/st-co/stch-icen<year>.txt (the county age, sex and race estimates), and 1900-1980/counties/totals/e7079co.txt and 1980-1990/counties/totals/e8089co.txt (the decades' county totals tables, for their census counts)
Geometry (source / stored): none / none
Update cycle: Yearly (each March)

Pipeline

The Population Estimates Program's national county file, CSV, one row per county with a column per year for each figure. The sync keeps New Mexico's county rows (SUMLEV 050, STATE 35), dropping the state's own total (040), and turns each into one row per year, the years read off the file's POPESTIMATE<year> columns. Every county the file names must be one the county layer of TIGERweb draws, or the sync fails. A file that is not valid UTF-8 is read as Latin-1. These rows are series postcensal.

The years before 2020 are series intercensal, from four further national files that never change:

Population, 2000–2009: co-est00int-tot.csv, one row per county, codes written without leading zeros (SUMLEV "50", STATE "1"), which the sync pads. ESTIMATESBASE2000 and POPESTIMATE2000–2009; its POPESTIMATE2010 and CENSUS2010POP are not loaded. vintage 2010.
Population, 2010–2019: the Bureau's table of revised county totals for the decade is a spreadsheet that names counties without codes, so the sync reads the same figures from cc-est2020int-agesex-all.csv, one row per county and YEAR code, its POPESTIMATE the whole population: YEAR 1 is the April 1, 2010 base, 2–11 July 1, 2010–2019, and 12 (the April 1, 2020 census) is not loaded. vintage 2020. 2010 is taken from this file, not the 2000s one, which has its own figure for it.
Components: the original releases' files, as published, matched by county code: co-est2010-alldata.csv for 2000–2010 (2000 is April to June) and co-est2020-alldata.csv for 2011–2019. Their natural change column is NATURALINC<year>. A county with no row in the 2000s release (one coded anew in the 2010s) has empty 2010 components; any other missing county fails the sync.
pop_change is computed, not read: pop_estimate less the row before's (2010's less the 2000s file's 2009, 2000's less 1999's, or ESTIMATESBASE2000 for a county with no 1999 row). The releases' NPOPCHG_<year> describes their own population, not the revised one, and is not loaded.
1970–1999: each county's July 1 population is the sum of its age, sex and race estimates, which nm_census_county_pop_char loads from the same files (its provenance has their layouts): vintage 1980, 1990 and 2000 for the three decades. The Bureau's county totals tables for the 1970s and 1980s, e7079co.txt and e8089co.txt, are fixed-width text in blocks of five years, a long name wrapping onto a line of its own; only their first column, the decade's April 1 census, is loaded, into estimates_base on the 1970 and 1980 rows. Their July figures are not loaded: the 1970s table rounds counties to hundreds, and both are on the geography of their day as the age files are, but not always the same day's. No component of change is loaded before 2000. pop_change is computed as for the later decades, 1970's from the April 1970 count, and the 2000 row's is from July 1999.
A history row's county need not be one TIGER draws today (Connecticut's eight old counties, the 1970s' Alaskan census divisions and Virginia's Nansemond, and a handful elsewhere); the sync logs them.

Field → column mapping

Source field	Column	Type	Description
`STATE` + `COUNTY`	`geoid`	varchar(5)	county FIPS
the `<year>` suffix	`year`	smallint
the newest `<year>`	`vintage`	smallint	1980, 1990, 2000, 2010 or 2020 on history rows
	`series`	varchar(12)	`postcensal`, or `intercensal` on history rows
`POPESTIMATE<year>`	`pop_estimate`	integer
`ESTIMATESBASE2020`	`estimates_base`	integer	2020 row only (`ESTIMATESBASE2000`, `YEAR` 1 on the 2000 and 2010 rows; the decade tables' census column on the 1970 and 1980 rows)
`NPOPCHG<year>`	`pop_change`	integer	computed on history rows
`BIRTHS<year>`	`births`	integer
`DEATHS<year>`	`deaths`	integer
`NATURALCHG<year>`	`natural_change`	integer	`NATURALINC<year>` on history rows
`INTERNATIONALMIG<year>`	`international_mig`	integer
`DOMESTICMIG<year>`	`domestic_mig`	integer
`NETMIG<year>`	`net_mig`	integer
`RESIDUAL<year>`	`residual`	integer
`GQESTIMATES<year>`	`gq_estimate`	integer

The file's per-1,000 rate columns (R…) and GQESTIMATESBASE2020 are not loaded.

Geometry

No geometry column.

Change detection

The Census sync's check asks for the next vintage's file each time and moves to it once it is published; otherwise it compares the current file's Last-Modified and length, and each history file's. Any of them marks this table stale, and the next Census sync rebuilds it whole, every year from 2020 replaced by the new release.

Attribution

US Census Bureau, Population Estimates Program.

Housing Units Permitted by County, 2020-2025

Where New Mexico's new housing permits went, 2020–2025

What stands out

Year by year

Caveats

Layers

Data sets used

Ask your own question about New Mexico

Transcript

State Planner chat · https://stateplanner.com

Prompt

Response

Where New Mexico's new housing permits went, 2020–2025

What stands out

Year by year

Caveats

Chart Queries

Housing units permitted per year, top counties

Exploratory Queries

Query 1 — Check which years 2020-2025 exist and county coverage.

Query 2 — County totals 2020-2025 with share, mix and per-1,000-resident rate.

Query 3 — Confirm population estimate year used for per-capita rates.

Map Layer Queries

Units permitted 2020-2025 — colored by "cls": 5,000+ units (#08306b), 1,000-4,999 units (#2171b5), 300-999 units (#6baed6), 1-299 units (#c6dbef), No permits in Census file (#eeeeee)

Used Data Sets

Provenance: nm_bps_county_permits

Identity

Pipeline

Field → column mapping

Provenance: nm_census_counties

Identity

Pipeline

Field → column mapping

Geometry

Change detection

Attribution

Provenance: nm_census_county_pop_est

Identity

Pipeline

Field → column mapping

Geometry

Change detection

Attribution

Provenance: `nm_bps_county_permits`

Provenance: `nm_census_counties`

Provenance: `nm_census_county_pop_est`