Physician Shortage as a Distribution Problem

What the regulation actually asks

Appendix A to 42 C.F.R. Part 5 sets out the criteria for designating areas with shortages of primary medical care professionals. Part I addresses geographic areas and requires three criteria to be met.

First, the area is a rational area for the delivery of primary medical care services. This is a prior question and it does the most conceptual work. Before any counting happens, the regulation asks whether the boundary being drawn describes a place where people actually obtain care. A designation cannot be manufactured by drawing a line around a group of blocks with no physicians in them.

Second, one of two ratio conditions prevails. Either the area has a population to full-time-equivalent primary care physician ratio of at least 3,500:1, or it has a ratio of less than 3,500:1 but greater than 3,000:1 together with unusually high needs for primary care services or insufficient capacity of existing primary care providers.

Third, primary medical care in contiguous areas is overutilized, excessively distant, or inaccessible to the population of the area under consideration.

The third criterion is the one that settles the character of the whole test. An area with a poor internal ratio does not qualify if the physicians next door are reachable and have capacity. The regulation is not asking how many physicians exist within a boundary. It is asking whether this population can get to care — and treating the boundary as a convenience for measurement rather than as the thing being measured.

The deficit formula

The size of a shortage is computed rather than counted. For areas with unusually high need or insufficient capacity, the primary care physician shortage equals the area population divided by 3,000, less the number of full-time-equivalent primary care physicians.

That formula expresses a target ratio rather than an absolute standard of adequacy. It says nothing about whether 3,000 patients per physician is good care; it establishes a benchmark against which a local deficit can be quantified for the purpose of allocating federal resources. A physician reading a shortage figure should understand it as the distance between a place and a policy benchmark, not as a clinical judgment about workload.

Where the distribution logic is explicit

Part II of Appendix A designates population groups rather than places, and it is here that the regulation states the distribution point in terms.

A specific population group within a particular geographic area is designated where three criteria are met: the area in which they reside is rational for the delivery of primary medical care services; access barriers prevent the population group from use of the area’s primary medical care providers; and the ratio of persons in the population group to primary care physicians practising in the area and serving the population group is at least 3,000:1.

The second criterion presupposes that providers are in the area. The group is designated as underserved not because physicians are absent but because these particular people cannot use them. The regulation then enumerates what such barriers may be: economic, linguistic, cultural, or architectural — or they may involve refusal of some providers to accept certain types of patients or to accept Medicaid reimbursement.

That last clause deserves to be read twice. A federal regulation contemplates, as a recognised cause of designated shortage, that physicians who are present and licensed decline to see the patients in question. On the face of the rule, a shortage can exist in a place with adequate physician density because of who those physicians will accept.

The third criterion also uses a different denominator: physicians practising in the area and serving the population group. A physician who does not serve the group is not counted, however close their office. Effective supply, not nominal supply, is what the ratio measures.

Why “rational area” comes first

The ordering of the criteria is not incidental. Requiring a rational service area before any ratio is computed prevents the most obvious form of manipulation — drawing a boundary to produce a result — but it also encodes a substantive commitment about what a shortage is.

A ratio is a fraction, and a fraction can be made to say almost anything by choosing its boundary. Draw a tight enough circle around a residential district and the physician count inside falls to zero; draw a wide enough one around a metropolitan region and almost no area is underserved. Neither figure describes anyone’s experience of obtaining care.

By requiring first that the area be rational for the delivery of care, the regulation ties the measurement to a real catchment: the geography within which people in fact seek and receive primary care. Only then does it ask how many physicians serve that catchment. The consequence is that the denominator is a behavioural fact about how care is sought rather than an administrative fact about how lines are drawn — which is a distribution premise built into the first sentence of the test.

Different thresholds for different professions

Appendix C applies the same architecture to mental health and produces markedly different numbers. A geographic area is designated where it is a rational area for the delivery of mental health services; where one of several ratio conditions prevails; and where mental health professionals in contiguous areas are overutilized, excessively distant or inaccessible.

The ratio conditions are alternatives. An area qualifies with a population-to-core-mental-health-professional ratio of at least 6,000:1 together with a population-to-psychiatrist ratio of at least 20,000:1; or with a population-to-core-professional ratio of at least 9,000:1; or with a population-to-psychiatrist ratio of at least 30,000:1. Where the area has unusually high needs the thresholds drop to 4,500:1 with 15,000:1, or 6,000:1, or 20,000:1 respectively.

Two features are worth drawing out. The psychiatrist thresholds are far higher than the primary care threshold — 30,000:1 against 3,500:1 — which encodes an expectation about how psychiatric care is distributed relative to primary care. And the structure separates core mental health professionals from psychiatrists, allowing an area to qualify on either the general workforce or the physician workforce alone. The regulation is measuring two distinct kinds of scarcity that happen to occupy the same field.

Full-time equivalence, and what it hides

The ratios are expressed against full-time-equivalent physicians rather than headcount, and the choice matters more than it appears.

An area with ten physicians each devoting a fifth of their time to primary care in that area does not have ten physicians for the purposes of the test; it has two. Conversely, a single physician working across several locations is apportioned among them. The measure is therefore of clinical capacity delivered rather than of professionals present, which is again a distribution concept: the same individual can contribute to the supply of several areas or to none in particular.

The measure also has a blind spot worth naming. Full-time equivalence records how much time is devoted, not how much of that time is available to the people who need it. A physician practising full time in an area, entirely on a payer basis that excludes most of the local population, counts as a full FTE for the geographic test. The population-group test in Part II exists in part to catch precisely that case, which is why the two parts of the appendix are best read as a pair rather than as alternatives.

What follows for the national-supply argument

None of this establishes that aggregate physician supply is adequate, and this page does not argue that. The point is narrower and more useful: the instrument the federal government actually uses to identify underservice is not built to answer the aggregate question, and citing designation counts in an aggregate argument misuses them.

A count of designated areas measures how many local places fail a local test that has a reachability criterion and an access-barrier criterion built into it. That number can rise while national supply rises, if supply concentrates. It can fall without any new physicians entering the workforce, if existing ones redistribute or if payer acceptance changes. And it can remain stable while the composition of the shortage shifts entirely from absence to inaccessibility.

The practical consequence is that shortage figures and workforce-supply figures answer different questions and are not interchangeable. Increasing the number of physicians nationally does not, by the regulation’s own logic, resolve a designation that rests on contiguous-area inaccessibility or on refusal to accept Medicaid.

The three failures the criteria distinguish

Read together, the two parts of Appendix A separate three quite different ways an area can be underserved, and conflating them is the most common error in this field.

Absolute scarcity. Few or no physicians practise in the catchment, and the neighbouring catchments are too distant or too full to absorb the demand. This is the case the phrase “physician shortage” conjures, and it is the one the geographic test in Part I is built for.

Relative maldistribution. Physicians exist within reach but the local ratio is poor while contiguous areas are comfortable. Here the third geographic criterion becomes decisive: if the neighbouring supply is genuinely accessible, the area is not designated, because the population can obtain care. The regulation declines to treat an unevenness that patients can travel across as a shortage.

Structural inaccessibility. Physicians are present in adequate numbers and a defined population cannot use them. Part II designates the population rather than the place, and the ratio is computed only over physicians serving that group.

The three call for entirely different responses. The first is answered by getting clinicians to a place. The second is often answered by transport, telehealth or referral arrangements rather than by new clinicians at all. The third is answered by neither — it requires changing the reason the existing clinicians are unavailable to those patients, which is usually a reimbursement or language or accessibility question rather than a workforce one.

A policy instrument aimed at the first failure will not touch the third. That is the sharpest practical implication of reading the designation criteria as drafted, and it is why designation counts should never be cited without saying which part of the appendix produced them.

Retention, and why it is the harder half

If distribution rather than aggregate supply drives designation, then the operative question for any underserved area is not only whether physicians can be brought there but whether they remain. Recruitment and retention are frequently discussed together and behave quite differently.

Recruitment is a transaction. It responds to the instruments that policy is good at building — loan repayment, visa pathways, signing incentives, training placements — and its success is measurable at a point in time.

Retention is a condition. It depends on whether a practice is sustainable over years: whether call burden is survivable, whether specialty backup exists for the cases that need it, whether a spouse can find work and children a school, whether the practice is financially viable given the payer mix, and whether the physician has professional company. None of these is a payment, and none is resolved by the instrument that brought the physician there.

The two also interact perversely. A programme that places physicians into an area for a defined obligation period, without changing the conditions that made the area hard to staff, produces a population of clinicians whose departure date is known in advance. Continuity for patients is not established by the presence of a physician; it is established by the presence of the same physician over time, and a rotating series of obligated placements can satisfy a ratio while delivering very little of it.

That is not an argument against placement programmes, which plainly move physicians to places that would otherwise have none. It is an observation that the designation criteria measure a stock and the underlying problem is a flow.

Payer mix as a distribution mechanism

The regulation’s reference to refusal to accept Medicaid reimbursement identifies something that operates continuously and quietly on distribution.

A physician deciding where to practise is deciding, in substance, which payer mix to accept, because payer mix is a property of geography. An area whose population is predominantly covered by a low-reimbursement programme is an area where a given volume of clinical work produces less revenue. The physician has made no decision about any patient; the decision is about the viability of a practice.

The consequence is that access can fail without anyone refusing anyone. Physicians locate where practices are viable, viability tracks payer mix, and payer mix tracks the same geography as need. The regulation captures the endpoint — a population that cannot use providers who are present — without addressing the mechanism, because a designation rule identifies where the problem is rather than why.

Why the thresholds are ratios and not workloads

A recurring objection to the framework is that a population-to-physician ratio says nothing about how sick the population is or how much care each person needs. That objection is correct and the regulation partly anticipates it.

The second geographic criterion admits an area at a better ratio — between 3,000:1 and 3,500:1 — where it has unusually high needs for primary care services or insufficient capacity of existing primary care providers. Need and capacity therefore enter the test, but only as a tie-breaker within a defined band. Above 3,500:1 the ratio alone suffices; below 3,000:1 no amount of demonstrated need qualifies the area on this route.

That design accepts a known imprecision in exchange for administrability. A rule that required a full assessment of population morbidity before any designation could issue would be more accurate and would designate almost nothing, because the assessment would rarely be completed. A ratio with a need-sensitive band is a compromise between measuring the right thing and being able to measure at all.

The consequence for anyone reading a designation is that it establishes a place has crossed an administrative threshold. It does not establish that care there is inadequate, and the absence of designation does not establish that care is adequate. These are screening criteria for resource allocation, and they behave like screening criteria: calibrated to catch cases at acceptable cost, not to describe each one correctly.

The designation is a gateway, which shapes the incentives

One structural feature explains a good deal of the behaviour around these criteria: designation is not merely a description. It is the entry condition for federal programmes, and that gives communities, clinics and states a direct interest in the outcome of a measurement.

That interest is legitimate and it is also distorting. A criterion set that allocates resources will be applied by parties who want the resources, and the elements most open to argument are the ones with the least arithmetic in them. The ratio thresholds are hard numbers and difficult to contest. What is genuinely contestable is whether an area is rational for the delivery of care, whether contiguous supply is excessively distant or inaccessible, and whether a population faces access barriers of the enumerated kinds.

Those three are the criteria that carry the distribution logic, and they are also the ones decided by judgment rather than by division. The regulation could have been written with a bare ratio and no reachability test, which would have been simpler to administer and would have measured the wrong thing. Its drafters chose accuracy about distribution over ease of application, and the cost of that choice is that the elements doing the real work are the elements that can be argued about.

For a reader assessing any particular designation, this suggests where to look. The ratio will be right. The interesting question is how the area was bounded and how contiguous accessibility was assessed.

Reading a shortage claim

Four questions dispose of most assertions in this area, and each is answerable from the designation framework itself.

What was the unit? A rational service area, a county, a state, or the country? The regulation designates the first; policy arguments usually cite the last two, and the numbers are not comparable.

Which threshold? Primary care at 3,500:1 or 3,000:1 with high need, or one of the mental health combinations? A single word — shortage — covers thresholds that differ by an order of magnitude between professions.

Absence or inaccessibility? A geographic designation and a population-group designation describe different failures. The first says there are too few providers here. The second says there are providers here that these people cannot use.

Who counts as supply? The population-group ratio counts physicians practising in the area and serving the population group. A figure that counts all physicians in an area answers a different question from the one the regulation asks.

Rural practice, and the specific shape of its difficulty

Rural areas dominate geographic designations for reasons the criteria make legible, and the difficulty they present is more particular than distance.

The third criterion is nearly automatic. Contiguous-area supply that is overutilized, excessively distant or inaccessible is the ordinary condition of a rural catchment. Where the nearest alternative is an hour away, the criterion that defeats designation in a metropolitan setting is satisfied by geography alone.

Small denominators make the ratio unstable. In a catchment of nine thousand people, the difference between two physicians and three is the difference between 4,500:1 and 3,000:1 — between comfortable designation and none. A single retirement can create a shortage area and a single recruitment can end one, which makes the designation status of small communities genuinely volatile in a way that is invisible in aggregate statistics.

Full-time equivalence cuts hard. A physician covering three communities is a fraction in each. Rural coverage is frequently delivered exactly that way, which means the arrangement that keeps care available at all also registers as thin supply in every place it serves.

The scope demanded is wider. A physician practising where the nearest specialist is distant is managing presentations that would be referred elsewhere. That is a professional burden the ratios do not capture at all: the criteria count physicians, not the breadth of what each is required to do.

The last point connects back to retention. What makes a rural post hard to sustain is frequently not the volume of patients but the absence of anyone to share the difficult ones with, and no ratio expresses that.

Telehealth does not neatly resolve the geography

Remote care appears to dissolve the distance problem, and in part it does. It is worth being precise about which part.

Telehealth addresses the second failure identified above — relative maldistribution, where clinicians exist within a region but not within reach. It maps less well onto absolute scarcity, because a remote consultation still requires a clinician somewhere with capacity, and it does not by itself supply the procedural, obstetric or emergency care that a physically present physician provides.

Against structural inaccessibility it is close to neutral. Where a population cannot use local providers because of reimbursement, language or trust, delivering the same encounter over a connection does not remove the barrier and may add one. The population-group criteria list economic, linguistic, cultural and architectural barriers; only the last of these is straightforwardly eased by remote delivery.

This page states no position on telehealth policy and cites no authority on it. The observation is only that the three failures respond differently, and that an intervention aimed at distance addresses one of them.

What the framework implies for a physician choosing where to work

The criteria are administrative, but read carefully they describe the conditions a physician will actually encounter, and that makes them more useful to an individual than their bureaucratic form suggests.

A geographic designation resting on absolute scarcity tells a prospective physician that they will be one of very few, that contiguous support is distant, and that scope will be wide. A designation resting on population-group access barriers tells them something quite different: colleagues are present, and the practice they are joining will be defined by serving patients whom other local practices do not. Those are different working lives, and the same word covers both.

The full-time-equivalence point matters personally too. A post structured as coverage across several communities registers as a fraction in each, and the arrangement that makes the coverage possible is the same one that makes the workload unusual. A physician evaluating such a post is evaluating travel, discontinuity and the absence of a single clinical base — none of which appears in the ratio that justified the position.

None of this is legal advice about any post, and no data about any place is stated here. The point is that the designation criteria, read as drafted, encode a fairly precise description of what makes a place hard to staff — and a physician can read the criteria for that purpose rather than only as an eligibility rule.

What this page does not decide

First, it states the designation criteria in 42 C.F.R. Part 5 as read at source. It does not state current designation counts, current workforce projections, or any figure about the number of physicians practising in California or nationally. No such data was verified for this page.

Second, the regulation has been the subject of proposed revision over its life, and this page does not address the history of amendment or any pending change. The criteria stated here are those in the appendix text read for this page.

Third, it does not address the programmes that attach to designation — loan repayment, placement, reimbursement adjustments and the rest. Designation is a gateway to several federal programmes, and none of their eligibility rules is stated here.

Fourth, medically underserved areas and populations are designated under a separate framework not addressed on this page. The two are frequently conflated and are distinct instruments.

Fifth, the observations about retention and payer mix are analytical rather than legal, and no empirical source is cited for them. They follow from the structure of the designation criteria and should be read as reasoning, not as findings.

Absence of designation proves very little

A last caution follows from the gateway character of the criteria. Because designation is applied for rather than automatically conferred, and because the elements that carry the analysis require judgment, the absence of a designation is weak evidence about a place.

An area may be undesignated because it genuinely has adequate accessible supply. It may equally be undesignated because nobody applied, because the boundary drawn did not qualify as rational, because contiguous supply was assessed as reachable on a view a resident would dispute, or because the population most affected was not identified as a distinct group under Part II. The criteria are designed to identify areas that qualify, not to certify that unlisted areas are well served.

This is the mirror of the point made above about designation itself. Both directions of the inference are weaker than they look, and neither the presence nor the absence of a designation is a finding about the adequacy of care.

Sources

1. 42 C.F.R. Part 5, Appendix A — Criteria for Designation of Areas Having Shortages of Primary Medical Care Professional(s). Part I geographic criteria: rational area; ratio of at least 3,500:1, or between 3,000:1 and 3,500:1 with unusually high needs or insufficient capacity; contiguous-area overutilization, distance or inaccessibility. Part II population-group criteria: rational area; access barriers preventing use of the area’s providers, including economic, linguistic, cultural and architectural barriers and refusal to accept certain patients or Medicaid; ratio of at least 3,000:1 for physicians serving the group. Shortage size formula: area population / 3,000 − FTE primary care physicians.
2. 42 C.F.R. Part 5, Appendix C — Criteria for Designation of Areas Having Shortages of Mental Health Professionals: core professional and psychiatrist ratio combinations, and the reduced thresholds for areas with unusually high needs.
3. 42 C.F.R. Part 5 — Designation of Health Professional(s) Shortage Areas. Authority: Public Health Service Act §§ 215 and 332 (42 U.S.C. §§ 216, 254e). Source: 45 Fed. Reg. 76000 (17 November 1980).
4. 42 C.F.R. Part 5, Appendix A (2017 print edition) — govinfo official PDF, consulted to confirm the geographic criteria as printed.

Related analysis

• Correctional medicine and the boundaries of obligation
• Who controls clinical judgment
• Restrictive covenants in California physician employment
• Private equity and the medical practice