38.9 percent of US independent pharmacies closed between 2010 and 2021 against 21.9 percent of chain pharmacies, and the price of a prescription is set by a company that does not own the pharmacy

What is happening?

45 percent of United States counties are classified as pharmacy deserts, and 48.4 million Americans — about 1 in 7 — live in one. That is up from 41.2 million in 2021, an increase of 7.2 million people in four years. The shortage bites hardest where the population is thinnest, and the stores that vanish are disproportionately the independent ones.

Between 2010 and 2021, 38.9 percent of independent pharmacies closed against 21.9 percent of chain pharmacies. Retail pharmacy presence fell 5.9 percent in rural communities between 2018 and 2023 against 3.4 percent in urban ones. As of 2021 there were 101 noncore counties and 15 micropolitan counties with no retail pharmacy at all.

The reason the operators give is the price they are paid per prescription. In a survey of independent pharmacists released in January 2025, 40.8 percent said they were paid below the National Average Drug Acquisition Cost on more than 40 percent of the Medicare Part D prescriptions they filled, 96.5 percent said reimbursement threatened the viability of their business, and 30.3 percent said they were considering closing during 2025. The 2024 trade figures point the same way: a 10-year high in the cost of goods alongside a 10-year low in gross profits.

Whose problem is this?

The party that sets the price of a prescription and the party that goes out of business are not the same party. A PBM that reimburses below acquisition cost records no entry when the store closes, and the three largest PBMs control roughly 80 percent of the market, up from less than half in 2012.

Where does this problem end?

This is not only a rural problem, but the shortage is concentrated in the places with the fewest people. The counts in these sources sit on different units and that difference matters. One framing counts deserts, most of which turn out to be urban tracts, while the other counts people living in whole counties measured by driving time. Neither number is wrong and they are not interchangeable.

What is the state now, and what should it be?

Now

What should it be — the only written target anywhere in this material is a price floor, not an access level. Several states now require that a pharmacy be paid at least the National Average Drug Acquisition Cost plus a professional dispensing fee: Kentucky effective 2025-01-01, Nebraska at USD 10.38, Iowa for benefits beginning 2026-01-01, California under SB 41 signed in October 2025 at USD 10.05, and Alabama under SB 252.

Needs a new measurementthe access target is missing. No source we opened states a national target for how many counties should have a pharmacy, how far a person should have to travel, or what share of rural residents should be within reach of one. The state laws set what a pharmacy must be paid. They do not say how many pharmacies there should be.

How big is it?

The national figure holds up. 48.4 million people, 1 in 7 Americans, live in a pharmacy desert, which the underlying research defines as a place where the average driving time to the three closest pharmacies exceeds 15 minutes for more than half the population of the county. Two reports state 48.4 million and a third states 48 million, and the rise from 2021 appears in two of them, one giving 41.2 million and the other 41 million.

The definition does not travel with the number. One of the two reports carrying the 48.4 million count defines a pharmacy desert instead as an area more than 10 miles from the nearest pharmacy, which is a different measurement presented as though it were the same one. The source that does give the driving-time definition also applies it to a census tract in one clause and to a county population in the next, so even the stated rule is not internally clean.

Narrowing it to rural is where it breaks. The same research is reported as saying both that 90 percent of the affected population is rural and that 56.8 percent of all rural residents lack nearby access. Those two cannot both be right. Ninety percent of 48.4 million is 43.6 million, and if that were 56.8 percent of all rural residents then the rural population would be 76.7 million, which at the 20.0 percent rural share cited elsewhere in this dossier implies a national population of roughly 383 million. That is well above the actual United States population, so at least one of the two percentages is misstated in the reporting and we cannot tell which.

A second count points the other way again. One brief holds that over half of all pharmacy deserts are located in urban areas, which uses deserts rather than people as its unit, and a third measure counts 1 in 8 neighborhoods as pharmacy shortage areas. Three units, no bridge between them. The widest defensible statement is the national one: 48.4 million people, of whom the rural share is unknown.

Needs a new measurementa rural-scoped affected population could not be derived. The national count is firm, the rural narrowing is not, and the population file records `not-derivable` rather than a forced chain.

Under what conditions does it arise?

1. The price is set by someone who does not buy the drug. The pharmacy purchases the medicine at a wholesale price and is then reimbursed at an amount a third party decides. One owner reported paying USD 582 for a brand-name anticoagulant and being reimbursed USD 536.20, a loss of USD 45.80 on a single prescription. 2. Volume does not rescue a below-cost claim. The average independent store filled 67,601 prescriptions in 2024 against 59,644 in 2023, a record, in the same year the trade recorded a ten-year low in gross profits. A pharmacy that fills more prescriptions at a loss ends the year further behind than one that fills fewer. 3. Rural demand is thin and there is no second store to absorb it. Rural Medicare patients see a community pharmacist a median of 14 times a year against 5 visits to a primary care physician, so the closure removes the most frequent point of contact with the health system. Only 13.9 percent of pharmacists practice in rural areas while 20.0 percent of the population lives there. 4. The loss lands off the books of the party that caused it. Closure shows up as travel time and missed doses in a county, not as a line item at the payer.

What has been tried?

What was found?

The row on the net count of stores carries the tension in this dossier. The count of stores is nearly flat while the count of closures is large, which means openings are running close to closures and the churn is not visible in a net figure.

Why is it still unsolved?

Cost structure — the price of a prescription is set by a company that does not own the pharmacy and does not lose anything when it closes.

A pharmacy buys a drug at a wholesale price it can observe and is then paid an amount set by a pharmacy benefit manager under a contract it has little power to negotiate. When that amount falls below the purchase price, the transaction destroys value for the pharmacy every time it happens. That inverts the usual economics of a small retailer. Scale normally spreads fixed costs, but here each additional prescription can deepen the loss, which is why a record dispensing year and a ten-year low in gross profit appear together.

Rural geography makes the same arithmetic fatal faster. A thin patient base cannot offset a negative margin with volume, and there is no neighboring store to inherit the demand, so the marginal store in a rural county is the one that fails first. The three largest PBMs hold about 80 percent of the market, so leaving the contract is not a real option for a store that needs the patients covered by it.

The corrective actions taken so far sit on one side of the ledger. Moving DIR fees to the point of sale removed the retroactive clawback but did not raise the price. The state floors do raise the price, but they bind only the plans a state can regulate, and the newest of them began applying to benefits in January 2026, which is too recent for an effect to have been measured.

What observation would mean it is solved?

Candidates — (a) the share of counties classified as pharmacy deserts falls from 45 percent; (b) the number of pharmacy closures per year falls; (c) the share of Part D claims reimbursed below the National Average Drug Acquisition Cost falls toward zero.

Each fails on its own. A fall in the number of closures per year can also mean that everything closable has already closed. The county share in (a) moves with a driving-time threshold that a single new store in a large county can flip, so it is coarse in exactly the rural places that matter most. And (c) is measured today by asking pharmacists, which makes it a survey result rather than a claims-level fact; it would have to be observed in claims data before it could carry the weight of a solved verdict. Read together, (a) and (c) would be the pair worth watching, with (b) only as a check that the fall in (a) is not the result of exhaustion.

What is it connected to?

Fills with researchplausible neighbors include medication adherence and avoidable hospitalization, rural hospital finance, the Medicare drug price negotiation program, and vaccination access, since independent pharmacies report providing flu immunizations at 93 percent of locations. Relation type and evidence grade were not researched in this round.

What these sources do not say

• How many rural pharmacies closed in 2025 and 2026 specifically. The closure counts we confirmed are either national and undifferentiated, or they stop earlier — 2021 for the split between independents and chains, 2023 for the split between rural and urban. No source we opened gives a rural closure count for the two most recent years.
• How the net count and the gross count reconcile. Independent locations fell by 24 over thirteen months while a separate report counts 326 closures in ten weeks. Openings must be filling most of the gap, but no source we opened states the opening count next to the closure count.
• Whether the state reimbursement floors changed anything. Kentucky, Nebraska, Iowa, California and Alabama all have laws with dates. Not one source we opened reports a closure rate or a desert measurement taken after those laws took effect.
• How many of the affected people are rural. One source says over half of pharmacy deserts are urban; another says 90 percent of the affected people are rural; the same report also says 56.8 percent of rural residents are affected, and those last two cannot both hold, because chaining them implies a national population far above the actual one. No source we opened states a rural desert population directly.
• What DIR fees still cost. Sources confirm the 2024 shift to point of sale but none we opened quantifies what remains, or what the unresolved pre-2024 claim book is worth.
• Whether closures are reversible. No source we opened reports how often a closed rural pharmacy is replaced, or how long a county stays without one.

See the evidence

Every URL in this table was opened during this round and returned the figures attributed to it, and each was opened a second time during an independent verification pass. Three primary documents were not read directly: the two underlying research pages blocked the fetch with HTTP 403 on both attempts, and the rural health policy brief is a PDF whose text could not be extracted, so the figures from that brief reach this dossier through the federally funded rural health information service that restates and cites them. Two cross-confirmations are worth recording: the 45 percent county share appears independently in the health trade outlet and the national news weekly listed above, as does the rise since 2021, stated as 41.2 million to 48.4 million in the first and as 41 million to 48 million in the second, and the chain closure program of 500 stores in 2025 within 1,200 over three years appears in both the regional daily and the university research brief. Three conflicts remain unresolved and each is recorded rather than smoothed over. The first is the definition. The health trade outlet defines a pharmacy desert by average driving time to the three closest pharmacies exceeding 15 minutes for more than half the county population, while the regional daily carries the same 48.4 million count under a definition of more than 10 miles from the nearest pharmacy, and no source we opened reconciles the two. The second is that the count of independent pharmacies moves in opposite directions depending on the counting basis — the independent pharmacy trade association records 18,960 locations and falling, while the benefit manager trade association records 23,384 and rising over 2014-2024 — and no source we opened explains the difference. The third is the reason this dossier carries no population number: the 90 percent rural share and the 56.8 percent rural-resident share come from the same report, appear in no other source we opened, and cannot both be true, so neither was used. During this round a search summary also asserted a closure series of 1,764 in 2021 rising to 3,929 in 2025; no fetched page contained those numbers and they are excluded.

People affected

Estimated range Not derivable

The reason and what is missing are listed under “What is missing” below