The uneven recovery of nursing-home census

Objective

National figures are the benchmark most decisions get measured against, and they are the wrong one for almost all of them. A building four points below the national index may be performing at or above its own county. A building matching it may be losing ground in a market that has fully recovered.

Michigan and New York show what the number hides. Both sit within a point of 95.4 and have almost nothing in common underneath it: Michigan's metros run 8.3 points ahead of its rural, New York's run 4.3 points behind, a spread of 12.6 points. One has a rural access problem. The other does not. The national figure reports them as the same state.

This study takes the CMS data those figures are built from and separates it by region, community type, facility type and facility size.

Regional variation

Indexed to pre-pandemic, the four census regions fan apart. All four fell 16.2% to 17.9% to a common Q3 2021 trough. What happened afterwards was not the same at all: the West is nearly back at 99, the South is at 98, while the Northeast sits at 93 and the Midwest at 92.

The urban-rural gradient

Sort every facility by the population of the county it sits in and the recovery separates into three lines that have not converged since. Large-metro homes are slightly above their pre-pandemic census index at 101. Rural homes sit at 96, about 3.7 points behind the metro average for two years.

Read those against each other, not against the national figure. The tier indices exclude 964 facilities that reported a census in 2020 but were never given a metropolitan flag, almost all of which left before CMS began populating that field. The exclusion is one-sided and lifts every tier: the national index is 95.4 across all facilities and 99.3 across the flagged ones. The gap between the tiers is unaffected. Their distance from the national number is not.

Outside work reaches the same gradient by a different route. McGarry BE, Wilcock AD, Gandhi AD, Grabowski DC, Geng F, Barnett ML. Changes in US Skilled Nursing Facility Capacity Following the COVID-19 Pandemic. JAMA Internal Medicine 2026;186(3):285-292. measured operating capacity, the highest occupancy a facility accommodated in the prior 60 days, rather than census. Over 2019 to 2024 they found licensed beds down 2.5% and operating capacity down 5.0% net, with 14% of rural counties losing a quarter or more of capacity against 6% of urban ones. Their 5.0% and the 4.6% here are different quantities on different windows, so the two do not replicate each other. What they share is direction and the rural tilt, from an independent measure.

The obvious explanation is that rural homes closed more often. Decomposing the change into existing homes, facilities newly certified and facilities leaving the panel shows that exits are the smallest term of the three:

	Existing homes	Newly certified	Exits	Net
Urban	−1.3%	+1.8%	−0.4%	+0.1%
Rural	−4.7%	+1.5%	−0.6%	−3.7%

The difference sits in the first column. Existing rural homes lost 4.7% against 1.3% in urban counties, a gap of 3.4 points. Newly certified facilities added back less in rural counties, 1.5% against 1.8%. Exits run higher in rural counties too, 0.6% against 0.4%, but that 0.2-point difference is a fraction of the same-store gap. Rural homes did not empty because more of them shut. They emptied while they were open.

What this cannot tell you is why the buildings went up where they did. The decomposition is an accounting identity, not a mechanism: same-store plus openings minus closures equals the net change by construction, whatever the reasoning of whoever built. Whether metro development responded to residents leaving existing homes, or to land, labor, payer mix or capital availability, is outside this data.

Nor can we say these are new buildings. What is observable is certification, not construction. Of the 331 facilities that begin reporting a census over the window, 81% hold a first Medicare approval date inside it. The remainder were certified earlier and simply were not reporting a census at the start, so they are re-entries rather than additions. And of the ones approved inside the window, nothing here separates a newly built facility from an existing building newly participating in Medicare. The column is labelled newly certified for that reason.

Robustness: state composition

This is the finding we stress-tested hardest, because a gradient this clean is often two or three states driving an average. This one is not. Rural trails urban in 39 of the 48 states with at least 150 residents in each tier. Remove the three worst and the gap narrows from 3.8 points to 3.5. Remove the five worst, 3.4. Remove the ten worst and rural still trails by 2.7 points. No small group of states accounts for it.

Robustness: facility type

Rural nursing homes are not a random sample of nursing homes. They are twice as likely to sit inside a hospital (6.2% against 3.3%), more than twice as likely to be government-owned (11.0% against 4.7%), and much less likely to be part of a continuing care retirement community (6.8% against 11.9%) or to be Medicare-only (0.5% against 4.9%). Any of those could be the real driver. A hospital swing-bed unit and a 240-bed suburban facility are different businesses.

So we removed them, one at a time and then all together:

Facilities included	Urban	Rural	Gap	Count
All (as published)	100.1	96.3	+3.8	14,727
Excluding CCRCs	100.7	96.6	+4.0	13,177
Excluding hospital-based units	100.1	96.3	+3.8	14,130
Excluding government-owned	100.4	96.7	+3.7	13,783
Excluding Medicare-only	100.0	96.2	+3.8	14,180
Excluding under 30 beds	100.2	96.3	+3.9	14,375
Excluding all five at once	100.8	97.3	+3.6	11,564

Dropping all five simultaneously removes 3,163 facilities, more than a fifth of the panel, and the gap moves from 3.8 points to 3.6. The gradient survives all of them. Free-standing, privately owned, dually certified nursing homes of ordinary size show the same gap the full panel does.

Ownership explains part of it

Ownership is the one test that does not come back clean. Rural counties hold fewer for-profit homes than urban ones, 67% against 77%, and more than twice the share of government-owned facilities, 11% against 5%. That matters because the three ownership classes did not recover alike, and the differences between them are larger than anything geography produces:

Ownership	Urban	Rural	Gap	Facilities
For profit	102.2	98.4	+3.8	10,901
Non profit	92.3	90.8	+1.5	2,882
Government	94.1	92.3	+1.9	944

A for-profit home in an urban county sits at 102. A non-profit in an urban county sits at 92. Ten index points separate them, against the 3.8 that separates urban from rural. On this evidence ownership is the stronger predictor of the two, which is not a finding this study set out to make and is the one most worth following.

The urban-rural gap does survive inside every class, at 1.5 to 3.8 points, so it is not ownership in disguise. But it narrows: standardizing the rural mix to the urban one takes the gap from 3.8 to about 3.3. Roughly an eighth of the raw gap is rural counties holding more of the ownership types that recovered worst. The remainder is geography.

Robustness: weighting

Every index here is a ratio of sums, so large facilities carry more of it than small ones. Comparing facilities present in both periods, the gap is 3.3 points weighted by census, 4.1 as an unweighted mean of facility-level change, and 3.0 at the median facility. It is not an artifact of the big buildings; the unweighted version is larger.

Robustness: facility size

Rural homes are smaller, with a median of 81 certified beds against 108 in urban counties. If small facilities recovered worse, “rural” could be standing in for “small” throughout this study.

The obvious objection is that this is the hospital units again, since a distinct-part unit inside a hospital is usually small and rural facilities carry twice as many of them. It is not. Excluding them, the medians are 109 urban against 83 rural, which is one bed away from the figures above, and the rural share of each size band moves by at most a point. Hospital units are smaller in both tiers, 77 beds urban and 50 rural, but at 9% and 13% of the under-60 band they are too few to move a median. Rural nursing homes are smaller as free-standing facilities.

Small facilities did recover worse. So did the largest ones:

Certified beds	Facilities	Census index	Occupancy then	Occupancy now	Beds
Under 60	2,692	97.3	81.5%	83.1%	−4.6%
60 to 99	4,628	100.8	81.0%	81.8%	−0.3%
100 to 149	5,016	100.4	80.6%	79.8%	+1.5%
150 or more	2,387	97.3	82.0%	79.5%	+0.4%

The pattern is an arch, not a slope. Both ends of the size range sit at 97 while the middle is back above 100, so there is no simple “bigger is better” effect for rural to be hiding behind. The occupancy columns say why the two ends got there differently. Small homes cut 4.6% of their beds and their occupancy rose to 83.1%, the highest of any band. The largest homes added beds and their occupancy fell 2.5 points to 79.5%. The two ends of the size range have different problems.

And the urban-rural gap holds inside every band:

Certified beds	Urban	Rural	Gap	Rural share of band
Under 60	99.5	94.1	+5.4	40%
60 to 99	101.9	98.5	+3.3	35%
100 to 149	101.7	95.6	+6.1	22%
150 or more	97.5	94.1	+3.4	10%

A rural home trails an urban home of the same size in all four bands, by between 3.3 and 6.1 points, and the gap is widest in the 100-to-149 band where rural facilities are only a fifth of the total. A 120-bed nursing home recovers differently depending on the county it sits in. This holds the business model roughly constant and varies only geography.

Holding everything constant at once

Everything above removes one thing at a time. That shows the gap survives; it cannot say how large the gap is once the other differences are accounted for together, because the differences overlap. A single specification can. The outcome is the change in a facility's own census between pre-pandemic 2020 and Q3 2026, for the 14,254 facilities reporting in both periods. The predictors are whether it is rural, its size, its type, its ownership, and which state it is in, carried as 53 fixed effects. Standard errors are clustered on state, since facilities in one state share whatever that state did.

Specification	Rural	Std. error
Rural alone	−3.80	0.63
Plus state fixed effects	−3.12	0.58
Plus state and facility controls	−4.33	0.64
The same, weighted by census	−4.16	0.65

Controlling for everything at once makes the rural gap bigger, not smaller, which is not what we expected. State fixed effects absorb about a fifth of the raw difference and ownership a further tenth. Size pushes the other way: rural facilities are smaller, and smaller facilities recovered better, so being small had been hiding part of the rural penalty. Holding state, size, type and ownership constant, a rural facility lost about 4.3% more of its census than a comparable urban facility in the same state.

How to read the table: the SE column is the uncertainty around each estimate. An estimate more than about twice its SE is one this data can separate from zero; one smaller than its SE is not.

What else the specification says

Term	Coefficient	Std. error	t
Rural	−4.33	0.64	−6.7
Size, per log unit of beds	−6.81	1.11	−6.1
Non-profit	−9.06	0.82	−11.0
Government-owned	−7.61	2.04	−3.7
Continuing care retirement community	−3.32	1.08	−3.1
Hospital-based	−2.52	1.47	−1.7
Medicare-only	−4.26	3.67	−1.2

Ownership is the largest term in the model. A non-profit facility lost 9 points more than an otherwise identical for-profit one in the same state, and a government-owned facility 7.6 points more. Both are roughly twice the rural effect and both are precisely estimated. The raw ownership split reported earlier was not a composition artifact; it survives controls for state, size and type.

The state effects are wider still. Their standard deviation is 5.2 points and they run from 27 points below the reference state to 4 above. Which state a facility operates in moves its census by more than whether that facility is rural, which is the same conclusion the divergence map reaches by a different route.

What this specification does not settle

It describes, it does not identify. Nothing here is an experiment, and every variable in it could stand in for something unmeasured: ownership for payer mix, state for Medicaid policy, rural for distance to a hospital. The R-squared is 0.070, so almost all of the variation between facilities is unexplained.

The sample is also restricted to facilities reporting in both periods. Entry and exit are 57% of the national shortfall and are absent from the model entirely, so this measures the gap among survivors and says nothing about which facilities left.

One control was dropped deliberately. Including how full a facility was in 2020 moves the rural coefficient from 4.33 to 6.28, which looks like a stronger result and is not one: baseline occupancy is census divided by beds, and the outcome is the change in census, so baseline census sits on both sides of the equation. A facility whose 2020 census is high through reporting noise will mechanically show both higher occupancy and a smaller increase. The 6.28 is partly arithmetic. The figure reported here excludes it.

Divergence and statewide decline

This distinction changes what you would do with the finding. Map each state by its urban-minus-rural gap rather than by its rural level, and two phenomena that a rural-only map treats as one pull apart:

• True divergence. Metros rebounded to or above par while rural fell behind. The widest gaps are Alaska +32, Arizona +24, Idaho +17, North Dakota +15, Vermont +15, Colorado +12. All 6 sit below the median rural base, and Alaska (309 rural residents in 2020) and Arizona (725 rural residents in 2020) head the list, so in every one of them a handful of facilities moves the number. The widest gaps that sit on a substantial rural base, at or above the median of 4,078 rural residents, are Arkansas +8 (7,998 rural), Michigan +8 (8,013 rural), Iowa +7 (11,434 rural), Kentucky +6 (10,374 rural), Kansas +6 (6,631 rural). That second list is the one to rely on.
• Whole-state decline. Both tiers fell together. Among states below the median index of 95.6 whose gap is inside two points, deepest first: Wisconsin 80, Minnesota 85, Wyoming 87, Nebraska 89, Pennsylvania 91, Louisiana 91.

Wisconsin and Minnesota look like the worst rural states in the country on a conventional map. They are not. Their metros fell just as far: Wisconsin's urban tier sits at 87 against a rural 85. Their rural decline is not rural-specific at all. It is a state-level story about Medicaid rates, workforce and bed supply, and the rural-only map gets that conclusion exactly wrong.

Across 48 states, 39 have urban above rural, with a median gap of 3.8 points. Divergence is the norm rather than the exception. It reaches 32 points in Alaska, where an urban tier at 132 sits against 309 rural residents statewide. The remaining 9 run the other way, with rural ahead of metro, by as much as 8.3 points in Washington.

36 of 48 states are still below their 2020 census. 20 of them shrank across both tiers; 16 shrank in rural while their metros held. Of the 12 that grew, 8 grew on their metros alone and 4 grew across both. Four groups, and it takes both dimensions to put a state in the right one.

West Virginia is the one state that grew overall with its rural ahead of its metros.

Archetype	States
Broad contraction, tiers together	20: WI, MN, MT, WY, ME, NE, PA, LA, NH, MS, WA, TN, IN, NY, IL, GA, AL, DE, MD, CA
Rural-led contraction, metros held	16: ND, SD, IA, KS, CT, MA, CO, MO, OH, OK, MI, HI, KY, TX, AR, VT
Metro-led growth, rural lags	8: AK, ID, NV, UT, AZ, VA, FL, OR
Broad growth, both tiers	4: NM, WV, SC, NC

Either axis on its own hides half the picture. Colorado and Washington both sit at 94 on total census, so a growth-only chart treats them as the same state, yet Colorado's metros run 11.6 points ahead of its rural while Washington's run 8.3 points behind. It runs the other way too: Wisconsin and New Mexico have almost the same gap, about 1.5 points, yet sit at 80 and 103 on level.

Contracting rows are listed worst first, growing rows strongest first. The two axes cannot share a panel: the gap needs a metropolitan flag, and that column is empty in the 2020 snapshot, so tier work reaches only facilities present in the latest snapshot. Growth is therefore measured on the true total with closures included, as everywhere else in this study. Measured on the flagged panel instead, 21 of 48 states would clear 100 rather than 12, because the homes a state lost are missing from its own baseline.

Agency staffing

Agency nursing tripled and then halved, and none of it lines up with the census gap. Contract hours went from 3.7% of direct-care hours in 2020Q2 to 10.7% at the 2022Q4 peak and back to 5.4% by 2026Q1, still 1.7 points above where they started.

It is worth testing because CMS reports contract hours separately from employee hours for every facility every quarter, and staffing is the mechanism McGarry et al. link to lost capacity. If rural homes leaned hardest on agency, and agency reliance cost census, that would join the staffing story to this one.

Rural did not lean hardest. Across 2022 rural facilities ran 8.7% agency against urban 10.8%. The tier lines did not cross until 2025Q2. The rural census gap opened while rural was using less agency than urban, not more.

A facility's own census path does not move its agency use either. Among the 13,817 facilities with staffing and census at both ends of the window from the 2022Q4 peak to 2026Q1:

Census over the window	Facilities	Census	Agency hours	Employee hours
Fell more than 2%	3,148	−11.2%	−49%	+2.2%
Roughly flat	1,982	+0.1%	−46%	+9.0%
Rose more than 2%	8,687	+18.8%	−44%	+21.1%

Agency came down by roughly the same amount whatever the census did. Employee hours are what track it. That points at a sector-wide unwind rather than facilities trimming agency as occupancy fell, which is the explanation worth ruling out, since it would have produced a census-agency correlation with nothing in it.

Put agency into the model that holds state, size, type and ownership constant, refit on the 14,185 facilities with staffing data, and the rural gap barely moves: from -4.36 without the agency term to -4.28 with it. That is 0.07 points, about 2% of the gap. Whatever is driving the rural shortfall, agency is not carrying it.

How much agency a facility used explains nothing on its own. Each extra percentage point of agency goes with +0.032 points of census, and the uncertainty around that estimate is 0.026, wider than the estimate itself. In plain terms the data cannot separate it from no effect at all.

Asked as a yes or no it looks different. Facilities that used any agency at all gained about 1.53 points more census than the 21% that used none, and that estimate (0.61 of uncertainty around it) is one the data can separate from zero. The direction is still the opposite of the hypothesis, and it still leaves the rural gap where it was, at -4.27. Using an agency at all is not the same thing as depending on one, and what separates the two groups is more likely whether a facility had residents it could not cover than anything about the agency itself.

A handful of extreme facilities do not drive any of this. Agency share runs 4.6% at the median and 52.2% at the 99th percentile; 175 facilities sit above 50% and 3 report agency for every direct-care hour. Capping the top 1% leaves the estimate at +0.034, and dropping every facility above 50% leaves it at +0.046 with the rural gap at -4.34. The aggregate table above is built from summed hours, not averaged ratios, so it is not exposed to them at all.

A first pass anchored on the 2022Q4 peak did appear to show a gradient, with the heaviest agency users gaining about three points more census. That window opens near the census trough, so it measures recovery from a low base, and the association disappears on the study's own 2020 baseline.

What this cannot say: the staffing series begins 2020Q2 and ends 2026Q1, so agency is never observed at either end of the census window; agency share is a ratio whose denominator moves with census; and agency use and census could both be responding to local demand, which is not measured here. The claim is narrow. Agency staffing is not the mechanism behind the urban-rural gap.

Demographic change

The intuitive explanation is that some places are aging faster than others. It does not survive the test. Sorting the 2,513 counties with at least 50 residents into quartiles by growth in the Medicare-eligible population:

County quartile	65+ growth	Census change
Slowest-aging quarter	+3.3%	−7.1%
Second	+8.8%	−6.1%
Third	+13.5%	−4.8%
Fastest-aging quarter	+22.3%	−2.5%

The tilt is real and it runs the right way, but it is weak. A correlation of +0.08 across the window, which is well under 1% of the variation between counties. The second column is the point: even the fastest-aging counties, with 22% more seniors, still lost census. Aging does not account for the differences between counties in this window. One likely reason is that most of the growth in the 65-and-over count sits at the younger end of it, well below the age at which people enter nursing homes, but this study does not measure age within that population and cannot test that.

Decomposition: demand and supply

Decomposing the national change shows where it went. Over the past year same-store census rose 1.39%, which is real, broad-based demand recovery. Exits took 0.39% off and new certifications added 0.27% back, leaving +1.27%. Over the full window since 2020 the shortfall is 43% occupancy and 57% capacity: existing homes account for −1.97% and net capacity change, exits less entries, for −2.65%.

Occupancy, nationally

Counting residents asks how many people are in nursing homes. Occupancy asks how full the nursing homes are. They are different questions and, nationally, they now give opposite impressions:

	Pre-pandemic	Latest	Change
Residents	1,318,308	1,257,244	−4.6%
Certified beds	1,633,807	1,563,603	−4.3%
Occupancy	80.7%	80.4%	−0.3 pts

The bed base shrank by almost exactly as much as the resident count did, so occupancy is effectively back to where it started. A 0.3-point difference is within noise. An operator asking “has occupancy come back?” gets a different answer from the census number, and both answers are true, because they are answers to different questions.

This is also the cleanest available statement of the supply finding. Census cannot fall 4.6% while occupancy holds unless capacity left at the same rate. It did: 70,000 certified beds, gone.

Occupancy by geography

Split by geography, that reassurance does not hold:

	Occupancy then	Occupancy now	Change	Beds
Urban	82.4%	82.0%	−0.4 pts	+0.6%
Rural	77.0%	74.8%	−2.2 pts	−0.9%

Urban homes are as full as they were. Rural homes lost more than two points of occupancy while barely reducing their bed count. Rural nursing homes have roughly the bed count they had before. Fewer of those beds are occupied. They started five points below urban occupancy and are now seven points below. Whatever national occupancy says about the sector, the gap it hides is wider than the census gap, not narrower.

Put the national figures together and the pattern is consistent. Demand is recovering: existing homes filled by 1.4% over the past year, and occupancy is back to 80.4% against 80.7% before the pandemic. Supply is what is shrinking, by 70,000 certified beds. The deepest contracting states show the same signature. Wisconsin and Minnesota each removed about 14% of their beds, roughly three times the national rate of 4.3%, while their populations over 65 grew faster than the nation's. The older population has grown. The bed count has not.

Which returns to the number at the top. 95.4 is arithmetic, not a market. It is the midpoint of a sector where Wisconsin sits at 80 and Alaska at 116, where metros are back at par and rural is 4 points short, and where the 11 states that land on it did so from opposite directions. Any operator, lender or agency measuring a building against the national figure is measuring it against a number that matches nowhere they could visit. The comparison worth making is to its own county, its own tier and its own state.