One Minnesota County Has No Nursing Home. Seventeen Have Exactly One. The Health Department Already Ranks All 87 and Posts the Ranking as an Attachment to an Application Form.

Same rules in this section as always: a named public dataset for every claim, every disparity stated with its denominator, the innocent explanation weighed first and hardest, and a fix a real office could carry out.

This one nearly beat me. The obvious story — rural Minnesota abandoned, seniors driving an hour to find a bed — is not what the data says, and I am going to show you where it breaks before I tell you what survives.

The count

Two federal files and one state file agree on the headline number. Minnesota has 338 nursing homes certified for Medicare and Medicaid. CMS's Provider Information file, processing date August 1, 2026, puts them in 86 of Minnesota's 87 counties and counts 24,060 certified beds. CMS's iQIES Provider of Services file, processing date July 6, 2026, counts the same 338 facilities and 23,964 certified beds. The Minnesota Department of Health's own June 2026 tally counts 338 facilities and 24,018 licensed and certified beds.

The county with none is Red Lake County, whose 3,882 residents as of July 1, 2024 include 936 aged 65 and over.

Seventeen more counties have exactly one: Cass, Clearwater, Cook, Grant, Hubbard, Isanti, Kanabec, Lake of the Woods, Mahnomen, Marshall, Murray, Nicollet, Pine, Sibley, Stevens, Swift, and Wilkin. The median facility in that group has 46 beds. The smallest two, in Lake of the Woods and Mahnomen Counties, have 32 each.

Those eighteen counties hold 61,796 Minnesotans aged 65 and over — 5.8 percent of the state's 65-plus population — in places where one closure, one receivership, one owner deciding the arithmetic no longer works, takes the county to zero, or already has. Ten of the eighteen have already been through it once: eleven certifications terminated there since 2012, carrying 391 beds.

The CMS Provider Information file and the Health Department's June 2026 table were built from different bed definitions in different months, and they identify the same eighteen counties.

Why the denominator is residents 65 and over

Because the Legislature said so. Minnesota's bed moratorium at Minn. Stat. § 144A.071 has an exception for what the statute calls a hardship area, and the first criterion the commissioner of health must weigh is:

a low number of beds per thousand in a specified area using as a standard the beds per thousand people age 65 and older, in five year age groups, using data from the most recent census and population projections, weighted by each group's most recent nursing home utilization, of the county at the 20th percentile, as determined by the commissioner of human services;

Beds per thousand people 65 and older, measured against the county sitting at the twentieth percentile — which across 87 counties falls between the eighteenth- and nineteenth-lowest, not the twentieth. That is the state's own yardstick, and the denominator in it is the one I used.

My population denominator is the Census Bureau's Vintage 2024 county characteristics file for Minnesota — July 1, 2024 estimates, 1,057,147 residents aged 65 and over statewide and 119,959 aged 85 and over. The Health Department used the same Census release for its June 2026 table, and every county figure I checked against it matched to the person: 936 in Red Lake, 8,968 in Cass, 211,444 in Hennepin, 1,057,147 statewide.

Across the 87 counties, certified beds per 1,000 residents 65 and over run from 0.0 in Red Lake County to 90.3 in Traverse County. The median county is 26.5. The state as a whole is 22.7. (Those ratios are computed on the iQIES certified-bed counts, so one dataset carries both the capacity and the closure arithmetic below.)

The ends of that distribution do not line up with anyone's mental map of Minnesota:

Now the innocent explanation, and it is a serious one

Rural counties are losing people. Fewer people means fewer residents to serve, which means a closure can be demand contracting rather than a system walking away. If that is what happened, the beds went where the people went and there is no disparity here at all — only arithmetic.

So I tested it.

From the iQIES Provider of Services file I pulled every Minnesota nursing home record carrying a program-termination code. Sixty-eight of them. Sixty-four terminated on or after January 1, 2012, and I dropped two that are not closures: a 44-bed building in Wheaton whose certification ended January 1, 2024 and was reissued at the same street address that November, and a 65-bed certification in Red Wing that ended in 2016 at an address where another certified facility has operated since 1987. That leaves 62 nursing home certifications terminated across 40 counties since 2012, carrying 2,436 beds on the 47 records that report a bed count. Fifteen report none, so 2,436 is a floor.

Then I put each county's bed loss beside its population change, using single-vintage Census comparisons so I was not chaining two different estimate bases: 2012 to 2019 from the Vintage 2019 file, 2020 to 2024 from Vintage 2024.

Here is what came back.

In all 40 counties that lost at least one nursing home since 2012, the 65-and-over population was larger in 2024 than it was in 2020. In 38 of the 40, it was larger in 2019 than in 2012. In aggregate, the 65-plus population of the closure counties rose 25.8 percent over 2012–2019 and another 12.7 percent over 2020–2024.

And the 47 counties that lost nothing? Their 65-plus population rose 15.0 percent over 2020–2024, against 12.7 percent in the closure counties. Their total population rose 3.2 percent, against 0.5 percent. Directionally that is a point for the innocent explanation — the counties that lost buildings were growing a little slower — but it is a sliver, not an account.

At the county level the relationship nearly disappears. Correlate beds lost per 1,000 residents 65 and over against county growth across all 87 counties and Pearson's r comes back −0.15 against 65-plus growth for 2020–2024 and −0.11 against total population growth; over the longer 2012–2019 window, −0.27 and −0.19. Negative, which is the direction the innocent explanation predicts — counties that lost more beds per senior did grow a little slower. And small: population trend accounts for somewhere between one and seven percent of the variation in which counties lost beds. Tie-corrected rank correlations run weaker still, from −0.03 to −0.14, though with 57 of the 87 counties tied at zero recorded bed loss I would not lean on the rank version. Population trend is in this story. It is a minor term.

Two counties in the closure group did see their 65-plus population fall in one of the two windows: Norman, down 2.4 percent over 2012–2019, and Lincoln, down 0.3 percent. Both were up again by 2024. That is the entire demand-contraction case in a state of 87 counties.

The closures also did not fall evenly. Per 1,000 residents 65 and over, the 80 counties outside the seven-county Twin Cities metro lost 3.2 beds since 2012 against 1.4 inside it — better than twice the rate. Twelve of the 62 terminations were in the metro; 50 were not.

What actually separates the counties is replacement

Closures are a fact of any capital-intensive business. Buildings age out, owners retire, census falls below break-even. The question that decides whether a county keeps a nursing home is not whether one closes. It is whether one opens.

Since 2012, Minnesota gained 19 new nursing home certifications — facilities whose original Medicare and Medicaid participation date falls on or after January 1, 2012 and that are still certified. They carry 1,286 beds.

Twelve of the 19, and 864 of the 1,286 beds, are in the seven-county metro — seven in Hennepin County, two in Ramsey, two in Washington, one in Dakota.

Of the 422 new beds outside the metro, 224 — 53 percent — are three Minnesota Veterans Homes: Silver Bay in 2015, Luverne in 2016, Fergus Falls in 2018. A fourth is the Wheaton building recertified in November 2024, ten months after its certification ended, which is a reopening rather than an addition.

Which leaves, outside the Twin Cities metro, in fourteen years: a 14-bed facility in Chisago County in 2014, a 56-bed facility in Olmsted County in 2015, and an 84-bed facility in Freeborn County in 2018.

The Veterans Homes are not a coincidence. Section 144A.071, subdivision 3(e)(1) lets the commissioner of health

certify or license new beds in a new facility that is to be operated by the commissioner of veterans affairs or when the costs of constructing and operating the new beds are to be reimbursed by the commissioner of veterans affairs or the United States Veterans Administration;

That is an exemption the state wrote for itself, and outside the metro it is where most of Minnesota's new nursing home capacity in the last fourteen years came from.

Sixty-two out, nineteen in. That is the fourteen-year ledger, and it is the mechanism underneath the longer-run bed loss this series documented separately — a figure sourced there, not here.

The statute, read all the way through

The moratorium is not a ban. Overstating it in either direction wrecks the analysis, so here is the text. Subdivision 2(a) is a command to say no:

The commissioner of health, in coordination with the commissioner of human services, shall deny each request for new licensed or certified nursing home or certified boarding care beds except as provided in subdivision 3 or 4a, or section 144A.073.

Then come the exceptions, and they are real. Subdivision 3 is the hardship-area route. Subdivision 4a permits renovation, replacement, upgrading and relocation of existing homes. Section 144A.073 is the competitive exception process. Subdivision 4b lets a facility park licensed and certified beds on layaway, and it sets the outer limit in six words: "Beds may remain on layaway for up to ten years." Which means a county's licensed bed count is not the same thing as a bed someone can move into tonight.

But look at how the hardship route actually runs. Subdivision 3(c):

On August 15 of odd-numbered years, the commissioner, in cooperation with the commissioner of human services, may publish in the State Register a request for information in which interested parties, using the data provided under section 144A.351, along with any other relevant data, demonstrate that a specified area is a hardship area with regard to access to nursing facility services.

May publish. Nothing in the statute makes a county's condition trigger it. And subdivision 3(d) caps what the route can produce even when it runs: "Beginning July 1, 2011, the number of new beds approved must not exceed 200 beds statewide per biennium. After June 30, 2019, the number of new beds that may be approved in a biennium must not exceed 300 statewide."

Then subdivision 4d(d), which is the provision I would put on a poster:

The county or counties in which a facility or facilities are closed under this subdivision shall not be eligible for designation as a hardship area under subdivision 3 for five years from the date of the approval of the proposed consolidation.

A county that loses a building through an approved consolidation is barred for five years from the one remedy written for counties that have lost buildings.

The competitive process, meanwhile, is funded and running. The Health Department's moratorium-exception page, last updated September 2, 2026, states that for the biennium including carryover and through June 30, 2027, the appropriation available to fund the added medical assistance cost of approved § 144A.073 projects is $7,654,319. The same page lists every award since 2023: 23 awards totaling $4,779,709, spread across 13 counties. Metro counties took $2,460,090 of it — 51.5 percent — and 11 of the 23 awards. Two awards in four years went to a county with one nursing home or none: Mahnomen Health Center, $275,730 in 2026, and West Wind Village in Stevens County, $73,126. So the process does reach them. It reaches them about once every two years.

The honest complications, and the concession

Red Lake County is not the county in the most trouble. The Health Department's June 2026 table ranks each county twice — alone, and grouped with its contiguous neighbors — and its footnote is explicit that "Rank of 1 = LEAST beds". Red Lake ranks 1st of 87 on in-county beds per 1,000 residents 65 and over, meaning dead last, and 66th of 87 once you count the three counties that touch it: 328 beds against 9,711 residents 65 and over. I computed the same thing independently before I found the state's table — certified beds within 25 straight-line miles of each county's population-weighted center, from the Census Bureau's 2020 county mean-center file — and got 292 beds against the identical 9,711 residents. By the state's own contiguous measure, the counties at the bottom are Chisago, Isanti, Rice, Goodhue, Crow Wing, Pine, Washington, Kanabec, Sibley, Cass, Le Sueur and Morrison — a list that is mostly exurban, not remote.

Distance is not the crisis either. Using the same coordinates, the median Minnesota county's population-weighted center sits 4.4 straight-line miles from the nearest certified nursing home. The farthest is St. Louis County at 19.9 miles, which is a function of its size. Red Lake County's center is 15.8 miles from Thief River Falls. No county center in Minnesota is more than twenty miles from a certified bed. Anyone telling you seniors here drive an hour to the nearest nursing home is not working from this data.

And the non-metro counties have more beds per senior, not fewer. The seven-county metro holds 43.5 percent of Minnesota's certified beds and 50.1 percent of its residents 65 and over — 19.7 beds per 1,000 in the metro against 25.6 in the other 80 counties. Run my simplified version of the statute's 20th-percentile test and six of the counties it flags are Anoka, Carver, Dakota, Scott, Washington and Sherburne — suburbs whose population centers sit three to five miles from a certified bed, most of it across a county line. A metric drawn on county boundaries mismeasures a suburb, and the statutory test has that flaw baked in.

The concession. Operators close buildings because the arithmetic stops working, and a state that pays most of the bill owns part of that arithmetic. Minnesota sets nursing facility payment rates under chapter 256R. It sets the property-rate treatment of approved moratorium projects at § 256R.26. And § 256R.495, added in 2025 and effective only upon federal approval under the Revisor's note on that section, reimburses the wage floors the Nursing Home Workforce Standards Board adopted only for workers paid below the standard — the formula sums, for certified nursing assistants, "the difference between $22.50 and any hourly wage rate of less than $22.50 multiplied by the number of compensated hours at that wage rate", and nothing for the aide already above it. The cost side arrives in full; the reimbursement arrives with a floor, a cap, and an application due October 1. I worked through that sequence separately. When a fifty-bed building in a one-facility county decides it cannot make next year work, the state helped write that spreadsheet. That is not a defense of anything. It is an accounting of who is responsible, and some of it is us.

What we can do

1. Publish the table the state already makes. The Health Department computes licensed and certified beds per 1,000 residents 65 and over, and per 1,000 residents 85 and over, for every county and every contiguous county group, with ranks, dated June 2026. I found it as a PDF attachment on the application-materials page for moratorium exceptions. It belongs on a data page, as a machine-readable file, with a county-level closure history beside it, updated on a schedule. It costs nothing. It already exists.

2. File the assessment the statute has required since 1984. Section 144A.071, subdivision 4: "The commissioner of health shall submit to the legislature, no later than January 15, 1984, and annually thereafter, an assessment of the impact of the moratorium by geographic area, with particular attention to service deficits or problems and a corrective action plan." By geographic area. With a corrective action plan. And § 144A.351 already directs the commissioners of health and human services to compile the status of long-term care services "by county and region", and the list of what that compilation must include runs through a "summary of county and regional reports on long-term care gaps, surpluses, imbalances, and corrective action plans" and, at subdivision 1(a)(3)(iii), "comparative measures of long-term care services availability, including serving people in their home areas near family, and changes over time". Then subdivision 1(b), whole: "The commissioners of health and human services shall make the compiled data available on at least one of the department's websites." Two standing duties, both keyed to geography, neither one requiring a new statute.

3. Make the closure record public as it happens. Under § 144A.161 the state already learns of every closure before it happens. A licensee must notify the commissioner of health, the commissioner of human services, the county social services agency, both ombudsman offices and the relevant managed care organizations of an intent to close, reduce, or change operations, with the number of residents affected. At least 60 days before the closing date a written notice goes to every resident and to both commissioners, and subdivision 2(d) is one sentence long: "A closing facility is prohibited from admitting any new residents on or after the date of the written notice provided under subdivision 5a." The Legislature touched this section this year — Laws 2026 ch. 95, art. 3, § 5 struck a clause in subdivision 1a about housing-with-services units displaced by a bed coming off layaway — and left the core intact: the section applies wherever a facility is undertaking "a closure, reduction, or change in operations," and it binds both the facility and the county social services agency. Dated, county-identified notice of every closure in Minnesota already crosses two commissioners' desks. Publish it as a register.

4. Fix the trigger, not the cap. Whatever number the Legislature chooses for the biennial bed cap — 300 today — the hardship process should not wait on a discretionary "may publish" in odd-numbered Augusts. A county reaching one certified nursing home or zero is a fact the state already knows from the § 144A.161 notice and its own bed table. That fact should start the clock automatically, and the subdivision 4d(d) five-year hardship bar should not apply to a county that has been reduced to one facility or none. Nobody needs a new agency for this. Somebody needs to be the person whose job it is to watch the number.

Red Lake County's 936 residents over 65 are not stranded. They are sixteen miles from a bed in Pennington County, in a three-county group that ranks 66th of 87, and on any honest reading the state's nursing home geography is better than the headline suggests. What is not better is the trend underneath it: sixty-two certifications gone since 2012, nineteen replacements, twelve of those in the metro, three of the rural ones built by the state for veterans, and a hardship remedy that starts when a commissioner decides to start it. Seventeen counties are one building deep and one has none. The state ranks them every year and files the ranking behind an application form.

First the facts. Then the fix.

Sources

Facility counts, beds, and closures. Computed by us from two CMS files, not taken from any summary. CMS Provider Information (data.cms.gov/provider-data, dataset 4pq5-n9py, file NH_ProviderInfo_Aug2026.csv, processing date August 1, 2026): 338 Minnesota certified nursing homes, 24,060 certified beds, facilities present in 86 counties, and the latitude and longitude of every facility. CMS Provider of Services File — Internet Quality Improvement and Evaluation System (data.cms.gov, POS_iQIES.DATA.Q2_2026.csv, quarter beginning April 1, 2026, processing date July 6, 2026), filtered to state_cd = MN and prvdr_type_id = 20: 406 Minnesota nursing home records, of which 338 carry program-termination code 00 (certified, 23,964 certified beds) and 68 carry a termination code. Of the 68, 64 terminated on or after January 1, 2012; two of those 64 were excluded as same-address artifacts rather than closures — CCN 245585, Traverse Care Center, Wheaton, terminated January 1, 2024 and reissued as CCN 245638 at the same street address on November 13, 2024 with the same 44 beds; and CCN 245355, St. Brigid's at Hi-Park, Red Wing, terminated February 3, 2016 at 213 Pioneer Road, where CCN 245449, St. Crispin Living Community, has been certified since March 1, 1987. That leaves 62 terminations across 40 counties, 2,436 beds on the 47 records reporting a bed count and no bed count on 15. New certifications are the 19 currently certified facilities whose orgnl_prtcptn_dt falls on or after January 1, 2012, carrying 1,286 beds; 12 of the 19 and 864 of the beds are in Anoka, Carver, Dakota, Hennepin, Ramsey, Scott or Washington counties, and the three Minnesota Veterans Homes among the remaining seven (Silver Bay, October 20, 2015; Luverne, August 31, 2016; Fergus Falls, July 6, 2018) carry 224 of the 422 non-metro beds.

Population. U.S. Census Bureau, Population Estimates Program, Vintage 2024 county characteristics, cc-est2024-agesex-27.csv (April 1, 2020 base and July 1, 2020–2024 estimates), and Vintage 2019, cc-est2019-agesex-27.csv (2010–2019), downloaded from www2.census.gov. Minnesota, July 1, 2024: 1,057,147 residents aged 65 and over and 119,959 aged 85 and over. All growth comparisons are within a single vintage — 2012 to 2019 from Vintage 2019, 2020 to 2024 from Vintage 2024 — because the two series rest on different bases. County population-weighted mean centers are the Census Bureau's CenPop2020_Mean_CO.txt (2020 Census); distances are great-circle miles from that point to the nearest facility's CMS-reported coordinates, and are straight-line, not driving distance.

The correlations. Computed across all 87 counties, x = beds on terminated certifications since 2012 per 1,000 residents 65 and over (July 1, 2024), y = percent change in county population. Pearson's r: −0.147 against 65-plus growth 2020–2024, −0.114 against total population growth 2020–2024, −0.268 against 65-plus growth 2012–2019, −0.192 against total growth 2012–2019, giving r² between 0.013 and 0.072. Tie-corrected Spearman's rho for the same four pairs: −0.045, −0.025, −0.135, −0.039; 57 of the 87 counties have zero recorded bed loss, which is why the rank statistic is reported but not relied on.

The state's own table. Minnesota Department of Health, "Licensed and certified nursing facility beds per 1000 65+ and 85+ by counties and contiguous county groups, June 2026", PDF at health.state.mn.us/facilities/regulation/nursinghomes/docs/beds.pdf, linked from the department's "Exception to the Nursing Home Moratorium" application-materials page (last updated September 2, 2026). Its footnote: "Bpt: beds/population x 1000. Beds from MN Health Dept HC Directory Database, June 9 2026. Population estimates from US Census, released June 2025", and "Rank of 1 = LEAST beds, with darker red indicating fewer beds in the area. Contiguous Counties are defined as a cluster of counties sharing at least a partial border." That table reports 338 facilities and 24,018 licensed and certified beds statewide; Red Lake County at 0 facilities, 0 beds, rank 1 of 87 in-county and rank 66 of 87 contiguous with 328 beds against 9,711 residents 65 and over; and the same eighteen counties with one facility or none that the CMS files identify. The moratorium page is also the source for the $7,654,319 appropriation available through June 30, 2027 and for the 2023–2026 award lists totaling 23 awards and $4,779,709.

Statutes, all pulled raw from revisor.mn.gov and read in full. Minn. Stat. § 144A.071 — subd. 1 (findings, including the $1,000,000 construction threshold), subd. 2(a) (the denial command and its exceptions), subd. 3(b)(1) (the beds-per-thousand-65-and-older hardship criterion, quoted in full), subd. 3(c) (the discretionary August 15 request for information), subd. 3(d) (200 beds per biennium from July 1, 2011 and 300 after June 30, 2019), subd. 3(e)(1) (the veterans affairs exemption, quoted in full), subd. 4 (the annual assessment by geographic area with a corrective action plan), subd. 4a (replacement beds), subd. 4b (layaway, "Beds may remain on layaway for up to ten years"), and subd. 4d(d) (the five-year hardship bar after a consolidation closure). Its History line ends at 2025 c 38 art 1 s 2,3. Minn. Stat. § 144A.351, subd. 1 (the county-and-region long-term-care compilation and the duty to post it). Minn. Stat. § 144A.161 — subd. 1 (definitions, including "'Closure' or 'closing' means the cessation of operations of a facility"), subd. 2 (initial notice and the admissions bar), subd. 3 (planning), subd. 5 and 5a (the 60-day resident notice) and subd. 8 (county responsibilities); subds. 1a and 8 were amended in 2026 and were read in the session law, Laws 2026 ch. 95, art. 3, § 5 and art. 7, § 2, not in the posted 2025 edition. Minn. Stat. § 256R.495 (RATE ADJUSTMENT FOR NURSING HOME EMPLOYMENT STANDARDS, added per its History line by 1Sp2025 c 9 art 1 s 23 — Laws 2025, First Special Session, chapter 9 — and carrying the Revisor's note that the section "is effective upon federal approval"), including subd. 2's October 1 application deadline and the subd. 4(a) formula quoted here; § 256R.26 is identified as the property-rate authority for moratorium-exception and hardship-area beds by § 144A.071, subds. 2(g) and 3(d). That § 144A.071 was not touched by the 2026 regular session was confirmed against Table 2 of the 2026 session laws, Minnesota Statutes New, Amended, or Repealed, which lists nine other sections of chapter 144A and no entry for 144A.071 or 144A.351; and against the Revisor's own page banners, which flag § 144A.161 as affected by 2026 legislation and flag neither § 144A.071 nor § 144A.351.

This ran long — about 3,200 words against the house norm of 1,100 to 1,400 — because the innocent explanation took four paragraphs of data to test and the complications took four more, and cutting either would have made the piece read stronger than the evidence is. What I could not verify: whether any Minnesota area has ever been designated a hardship area under § 144A.071, subd. 3, or how many beds have actually been approved under it, because the State Register and mn.gov are not reachable with the tools used here; whether the subd. 4 annual assessment or the § 144A.351 compilation are currently published anywhere — the Legislative Reference Library's mandated-reports search returned 404 at every URL tried, and I make no claim that these documents do not exist; and the Minnesota Department of Health's own licensed-bed file, which I could not obtain directly and used only through the June 2026 PDF. Limits of what is above: a CMS program-termination code records the end of a Medicare and Medicaid certification, which is not always the closing of a building, and the iQIES file retains only four Minnesota nursing home termination records dated before 2012, so it cannot be read as a closure census before then; its 2025 and 2026 counts (one each) may reflect reporting lag rather than a slowdown. Certified beds are not staffed beds, and § 144A.071, subd. 4b layaway means licensed beds are not available beds either. The hardship calculation above is a simplified version of the statutory test — unweighted, not broken into five-year age groups or weighted by nursing home utilization — and is not the commissioner's determination. The seven-county metro is used as the metro definition throughout; CMS's own urban/rural indicator disagreed with it on several counties and was not used. MDH's June 2026 table reports some ratio columns as integers and others to two decimals, and its statewide ratio row appears to be an average of county ratios rather than a state total, so no statewide ratio was taken from it. The article and section numbers for the 2026 amendments to § 144A.161 come from the Revisor's own notice on that statute's page; my text extraction of the 2026 Table 2 misaligned its chapter, article and section columns, so Table 2 was relied on only for the absence of any § 144A.071 or § 144A.351 entry, which is a whole-document search and does not depend on column order. I did not separately confirm the effective date of Laws 2026 ch. 95, art. 3, § 5. MDH and CMS also differ on facility and bed counts in a handful of counties, Anoka most of all, because they count licensed-and-certified beds in June against CMS-certified beds in July and August. This piece describes public statutes and public facility data for a general audience. It is not legal advice, it is not advice about placing anyone in a nursing home, no client or case of mine is discussed, and reading it creates no attorney-client relationship. Corrections: campaign@madgettformn.com.