Medicine has never been cleanly one or the other. It has always been a hybrid, and the relative weight of the two poles has shifted with politics, technology, and the organization of care. This modern unease is not new confusion so much as a return of an old structural problem under new pressures (insurance, regulation, public health emergencies, and now AI).
The Historical Hybrid
Classical Greek medicine treated the physician primarily as a learned expert and skilled practitioner rather than a public official. The Hippocratic tradition emphasized technical competence, prognosis, and a limited fiduciary duty to the individual patient; the physician was usually an independent itinerant or household retainer, not an agent of the polis. There were occasional public physicians hired by cities, but this was the exception. Status derived from knowledge and results, not from office.
Medieval and early modern Europe layered a clearer dualism on top of this. University-trained physicians (the medici) were learned experts whose legitimacy came from texts, Latin, and university degrees; they treated the affluent and advised courts. Surgeons and barber-surgeons were craftsmen of lower status. At the same time, hospitals and charitable institutions—often monastic or municipal—delivered care as a public or religious duty to the poor. The physician’s expert role and the institutional public-service role were largely separated by class and function.
The modern professional form crystallized in the eighteenth and nineteenth centuries. States began granting organized medicine a monopoly on practice (licensing, medical boards, restriction of competition) in exchange for self-regulation, educational standards, and a public-facing ethical code. The Royal Colleges, the AMA, and continental systems all embodied this bargain: professional autonomy and status as learned experts were secured because the profession accepted certain public responsibilities. The Flexner Report (1910) in the United States is the clearest American expression of the expert side of that bargain—standardizing scientific education, closing weak schools, and elevating the physician’s epistemic authority—while the growth of public-health institutions, municipal hospitals, and later social insurance pulled in the opposite direction.
Twentieth-century welfare states and wartime medicine intensified the public-servant dimension. National health services (NHS, Canadian Medicare, etc.) explicitly framed physicians as providers of a social right, often salaried or under heavy regulation. In the United States the same pressure appeared later and more partially through Medicare, Medicaid, and EMTALA (the duty to provide emergency care regardless of ability to pay). Professional associations spent decades resisting full socialization precisely because they saw it as converting the expert into a state functionary. The result in most high-income countries is a mixed system in which physicians retain considerable clinical autonomy and high status while operating inside public financing, public quality metrics, and public-health duties.
Why the Two Roles Feel in Tension
The learned-expert role prioritizes specialized knowledge, judgment under uncertainty, and professional self-regulation. It assumes that the person who has mastered a difficult body of knowledge is best positioned to decide how it should be applied, and that external control will degrade both quality and recruitment. The public-servant role prioritizes universal access, accountability to taxpayers or the polity, and the subordination of individual clinical decisions to broader social goals (cost control, equity, population health). When these collide—on issues of scope of practice, defensive medicine, prior authorization, duty hours, or pandemic mandates—the physician experiences the contradiction directly.
Culturally the society has never fully resolved which pole should dominate. One persistent strand treats medicine as a calling or civic vocation (the “public servant” ideal, visible in wartime service, rural practice, and much of the language of medical oaths). Another strand treats it as a high-skill guild whose legitimacy rests on expertise and whose rewards should reflect scarcity and training investment (the classic professional ideal). Both appear in the same ethical documents and the same political debates. The result is institutional hybridity and periodic legitimacy crises whenever one side is felt to be overreaching.
Contemporary Expressions
Different systems weight the balance differently. In the NHS and similar single-payer arrangements, physicians are more clearly public employees or contractors delivering a state service; autonomy exists inside explicit public frameworks. In the United States the fee-for-service and private-practice tradition keeps the expert/entrepreneurial pole stronger, yet the same physicians are bound by public-payer rules, quality reporting, and the legal duty to treat. Hospital employment has grown rapidly, converting many into de facto organizational agents rather than independent experts. Public-health emergencies (HIV, COVID) temporarily enlarge the servant role; cost-containment and liability regimes enlarge the constraints on expert discretion.
The same tension appears inside training and practice. Residency is framed both as the acquisition of specialized expertise and as a period of service under institutional and public standards. Burnout and “moral injury” discourse often reflects the experience of being asked to act as an expert while being managed as a cost center or compliance node. Scope-of-practice fights with other professions, debates over physician-led versus protocol-driven care, and arguments about AI all reopen the same question: how much independent judgment does the public want to purchase, and under what conditions?
Implications and Continuities
Historically, societies have usually resolved the tension by granting the profession a privileged status (high barriers to entry, self-regulation, social prestige) conditional on visible public service and accountability. When either side of the bargain frays—when expertise is perceived as self-serving, or when public duties are perceived as eroding clinical judgment—the legitimacy of the arrangement declines. The current period is one of such fraying, accelerated by cost growth, administrative burden, and the prospect that some expert functions can be partially automated or standardized.
The classical and early-modern pattern is still visible: the physician’s claim to authority rests on a body of knowledge that is difficult to acquire and apply, while the claim to public resources and legal privileges rests on a duty that extends beyond any single paying patient. Neither claim has ever fully displaced the other. The practical question for any given era is how the two are weighted and how the resulting hybrid is governed.
A forward facing outlook
In the contemporary United States that question of governance is being answered less by deliberate redesign than by the cumulative force of payment rules, ownership structures, liability law, and technology. The hybrid is not dissolving; it is being re-sorted into more distinct tracks whose relative sizes will determine whether the learned-expert role remains broadly available or becomes a vestigial privilege.
One track is already dominant. Roughly 80 percent of U.S. physicians are now employed by hospitals or other corporate entities, up from just over half in 2018, and the share of practices under non-physician ownership has more than doubled in the same period. Employment converts the expert into an organizational agent. Clinical judgment is still exercised, but it is exercised inside productivity targets, quality dashboards, prior-authorization workflows, and capital-allocation decisions made by systems that must also satisfy bondholders or shareholders. Medicare’s fee schedule has further reinforced the shift. After inflation in practice costs, physician payment has fallen roughly one-third since 2001, while hospital outpatient departments continue to receive substantially higher rates for the same services. The resulting site-of-service differential has been a documented driver of consolidation. Site-neutral payment proposals circulating in Congress would narrow that differential and thereby slow the absorption of independent practices, but they face the familiar obstacle that hospital margins and rural access arguments are politically potent.
A second, smaller track is the deliberate reassertion of the learned expert pole. Direct primary care and concierge models have expanded rapidly; the number of such practice sites rose more than 80 percent between 2018 and 2023, and recent tax changes have made membership fees more attractive inside health-savings accounts. These arrangements exit most of the third-party payment machinery, restore panel size and visit length to levels compatible with continuous judgment, and treat the physician primarily as a learned professional contracting with patients or employers. Parallel experiments—physician-led accountable care organizations, independent practice associations that share contracting infrastructure without ceding ownership, and specialty “micro-practices”—attempt the same recovery of autonomy inside residual insurance frameworks. Their ceiling is set by the willingness of patients and employers to pay for unbundled expertise and by the continued legal enforceability of corporate-practice-of-medicine restrictions. Several Democratic proposals now seek a federal overlay on those state restrictions precisely because private equity and insurer-owned management-services organizations have used friendly-physician structures to circumvent them.
A third track is being written by non-physician clinicians and by software. State medical societies currently rank scope-of-practice expansion as their highest legislative priority, ahead even of Medicaid reimbursement; the American Medical Association reports having helped defeat more than 150 such bills in a single recent year. The counter-pressure is access in rural and underserved areas plus the simple arithmetic of physician supply. At the same time, clinical AI is moving from documentation assistance toward triage, chronic-disease management, and differential-diagnosis support. Organized medicine’s emerging policy is that AI tools must remain under physician direction, cannot themselves be licensed, and should not shift liability onto the physician for refusing an algorithmic recommendation. That stance preserves the formal learned expert role while leaving open the practical question of whether the “front door” of care migrates to payer—or system—controlled interfaces that then route residual complexity to employed physicians. If it does, the public-servant dimension expands by default: the physician becomes the accountable human backstop for a largely automated, population-level process whose parameters are set by regulators and payers.
Levers of Tomorrow
Governance of this hybrid reality will therefore activate a few concrete levers.
Medicare payment updates that at least track practice-cost inflation, combined with site-neutral rates, would slow the conversion of independent experts into employees; continued real declines will accelerate it.
Enforcement (or federal preemption) of corporate-practice rules will determine whether ownership can be separated from clinical control.
Scope-of-practice statutes and graduate-medical-education funding will set the relative scarcity of the physician license itself.
Liability doctrines and licensure rules for AI will decide whether algorithmic systems enlarge or erode the domain of professional judgment.
Because these levers are distributed across Congress, CMS, state legislatures, and courts, the most likely near-term outcome is not a single national settlement but further geographic and market segmentation: high-autonomy, often cash-based practices for patients who can exit insurance constraints, and more heavily protocolized, employed, and technology-mediated care for those who remain inside public or quasi-public financing. That pattern is historically familiar—the medieval separation of university physician and charitable hospital, restated in contemporary payment and ownership forms. Whether it remains stable depends on whether the public continues to underwrite the training and legal privileges of a high-status expert class while simultaneously demanding that the same class function, for most patients, as a regulated public rights delivery node.

