Membership in the American Medical Association (AMA) has suffered a precipitous decline since the 1970s and, with it, a loss in revenue. The expansion of subsidized student memberships has bolstered its official membership number; only 12.6% of physicians who have completed their training now belong. The AMA strengthened its alliances with specialty societies and even tried to restructure the organization around organizational (rather than individual) membership: two hundred specialty societies are incentivized to encourage their members to join the AMA. Earlier federal policies supported by the AMA that gave it a central role in recommending Medicare reimbursement policies have established a membership pipeline for the AMA partly because specialty societies need to influence reimbursement policies. Commercial products have also helped subsidize AMA lobbying efforts that reinforce its historical position, despite a loss of members.
The American Medical Association (AMA) has had a long-standing influence over health policy, and its claim to be the leading physician organization in health policy has been unquestioned. Although today the organization is perceived as less powerful, in fact the AMA has retained power in specific policy arenas and domains, particularly physician reimbursement. Lobbyists working on Washington, DC's famed K Street ranked the AMA as the second most influential organization, after the American Association for Retired Persons, and staff in Congress also ranked it as the second most influential, behind the Pharmaceutical Research and Manufacturers of America (Heaney 2006: 901). The continued influence of the AMA over policies is surprising, however, given that its membership has been declining.
Historically, most physicians were members of the AMA: in the 1950s, 75% of doctors were members (Collier 2011). At the same time, for much of the twentieth century the AMA was able to rely on an economically coercive aspect of membership, which made physicians dependent on their membership in county medical societies to obtain specialty board examinations and ratings, for referrals, and to be included on “white lists” of approved doctors—those not on the list were denied access to hospital facilities (Campion 1984). In the hospital and graduate medical education arena, the AMA passed a resolution in the 1930s that required hospitals' accredited internship training staff to be composed solely of members of local societies.
Today, the AMA's total membership is apparently large, with 240,500 doctors belonging to the organization. At the same time, a seemingly large number of members belies a significant decline in AMA membership since the 1950s. Today, at most, just 18.7% of a total of 1.28 million US physicians are AMA members, which is not necessarily an embarrassing number, at least not until we understand that 107,984 of the AMA's members are students or residents, or around 45% of the AMA's 240,500 members and 46% of the 234,876 medical students and residents in the US. Of those who have completed their training, only 132,514 (12.6%) of the 1,048,601 physicians who have completed their training are AMA members; thus, physicians who have finished their education and training are much less likely to be AMA members.
Seen in a broader context, this declining trend may not be meaningful, because scholarship on social capital suggests declines in associational memberships are not unusual. The decline in membership would seem to confirm that “organized” medicine is in abeyance; Americans are less civically engaged than they were in the 1950s. A different interpretation might be given, however, if we consider that the AMA should have seen increases between the mid-1960s and the early 1990s, when “elite, well-educated Americans led the way in abandoning cross-class membership associations, meanwhile creating and supporting professionally managed organizations” (Skocpol 2004: 4).
The evolution of the AMA suggests the creation of a professionally managed organization that has reshaped itself in response to declining membership in ways that have counteracted its numerically smaller size. The first part of this article surveys trends in the membership of the AMA. I next describe how the AMA met the challenges of declining membership and competition from other organizations. The AMA addressed declining membership and competition by using a bigger-tent strategy that tried to formalize its coordinating role with specialty societies. But its real ability to dominate over other organizations was the result of policy feedback from its own earlier lobbying, which gave it a healthy revenue stream from coding and billing and essentially tithed the associations that had to have a certain number of AMA members.
Membership Characteristics and Theories of Organizational Maintenance
Classic work on interest groups posits that to have influence interest groups need members, and to get members they have to offer and continue to offer something that individuals or organizations feel they need (Hansen 1985). If the AMA is seemingly persistent and powerful in the face of a shrinking membership base, this is puzzling. Traditionally, for an organization to “speak for” or represent a group, we would expect it to have a larger share of the group. When membership is declining, as it has over time for the AMA, this suggests that it fails to provide benefits sufficient to address the cost of membership. Although some scholars believe the collective action problem is overstated, groups survive only if they resolve the collective action problem, and groups risk losing effectiveness and viability if membership falls. These theories of interest groups suggest organizational maintenance is almost impossible when membership is declining.
Although the AMA boasts almost a quarter of a million members, its membership is not as large as it would appear, given the overall number of physicians, which is well over a million doctors. Delving further, the balance between regular and student plus resident members shows the largest single constituency within the organization is medical students and residents (see table 1). Surprisingly, 107,984 (45.9%) of the AMA's 240,498 members are student members. Among those who have completed their education and training, 132,514 (12.6%) of 1,048,601 physicians who have finished their education and training are members of the AMA. Overall, only 240,498 physicians (18.7%) of 1,283,477 physicians in the US are members.
Comparisons of membership as a share of the total physician population can be complicated by the choice of the denominator, because the total number of physicians may not be consistent or comparable. With that caveat, while membership declined in the 1960s, membership declines accelerated in the 1970s, from 64.5% in 1970 to 46.9% in 1979 (AMA Special Collections cited in Ameringer 2008). In the 1980s and 1990s, membership continued to decline from 44.7% to 35.9% between 1985 and 1995 (Ameringer 2008). During that same period, state and county medical societies also saw declines: between 1977 and 1997, membership in state societies decreased from 73% to 47%, while membership in county medical societies decreased from 67% to 43% (Karlin 1998, cited in Ameringer 2008: 160). Different states had different requirements regarding AMA membership as part of the state or county membership (Ameringer 2008).
Historically the AMA did not depend on medical students or residents to bolster its membership numbers. In 1978, the organization had 207,000 members, of which 174,500 paid full dues. At that time, therefore, most AMA members were physicians who had completed their education and training: just 13.5% (28,000) were residents or students (Barclay 1978: 2082). In response to a decline in advertising revenue in the 1960s, the AMA increased membership dues in the 1970s (Campion 1984: 363–91). By 1985, the AMA president stressed the need for growing revenues that could support its activities while also acknowledging that the organization was limited in how much it could increase dues (Sammons 1985: 1585). Since 1994, the AMA has not raised membership dues (AMA Board of Trustees Report 2016:34). However, member dues largely seem to be spent on collective goods rather than individual and selective incentives. The AMA is not a 501c(3) but a 501(c)6 organization, which has to pay tax on a portion of its political expenses. The organization admits that most of the member dues go toward lobbying: “AMA dues are not deductible as a charitable contribution for federal income tax purposes, but may be partially deducted as a business expense. AMA estimates that 60% of your membership dues are allocable to lobbying activities of the AMA, and therefore are not deductible for income tax purposes” (AMA 2017).
Based on a comparison with the American College of Physicians (ACP), which has 152,000 members, the AMA spends significantly more on lobbying per member: $5 versus $86. The ACP earns less from royalties, which are only about 7% of its total revenue.
With the bulk of AMA membership dues subsidizing lobbying, new organizations have started that challenge the AMA's model. Doximity.com, a start-up social media organization targeted at physicians, is based on the premise that the professionalization of physician organizations has diminished the role of members and removed individuals' motivation for membership. Doximity claims that 70% of US doctors are verified members of the social network, “more than the American Medical Association”:
What you see is, people talk to each other, they help each other. The LinkedIns and the Facebooks, this kind of collaborative thinking is the kind of format that consumers prefer today. . . . Some groups have been wiped out in terms of retention rates, in terms of any other measure. . . . Associations are stuck in an “Industrial Age of thinking,” where they are more concerned with selling the products and programs they produce than they are with meeting member needs. (Doximity n.d.)
Doximity likely has a valid argument about interest groups that have become too large and commercialized; however, the purpose of organizations like Doximity is different, and they need to raise revenue from the sale of member data, which undermines the collaborative ethos they espouse. Start-ups like Doximity do not necessarily threaten the long-term viability of the AMA, however, but operate within a different niche. Some argue that, with changes in technology, the notion of membership itself is outdated and unimportant: “The very concept of organizational membership changed” (Karpf 2012: 25). Rather than rooted in popular membership, associations shifted “toward professionally managed organizations, many with no members or chapters at all” (Skocpol 2004: 6). While interest group theorists have long stressed the need for benefits to be provided in exchange for membership revenues, modern organizations could potentially require much less, and they have many different sources of revenue today that may not have been possible in past decades (discussed below). The AMA itself recognized that “a substantial number of physicians see no need to lend their personal support through membership. The Association will continue to strive to overcome this mistaken belief” (Sammons, 1985: 1584). This turns on its head the traditional idea of interest groups failing, because participants obtain solidary benefits from their participation on sites like Doximity, but these are relatively costless for the organization to provide. If the classic free-rider problem disincentivizes people from joining, since nonmembers can benefit without joining the organization, it seems that the pull for many members may be the act of participating itself.
Building the House for Medicine
Rosemary Stevens (2001) observed that physicians increasingly see their specialty society as their organizational home. Whereas the proliferation of specialty societies seems to be more controlled in other countries, in the United States there appears to be a low barrier to entry for new organizations, although most operate with some informal support from “parent” specialties such as surgery or radiology.
Feedback from payment policies may have also encouraged this proliferation, since under the resource-based relative value scale (RBRVS) specialties essentially have an interest in creating highly specific billing codes and associated reimbursement (Laugesen 2016). Despite the fact that the AMA has struggled in membership and that most physicians find their specialty organizations more attractive, it is also true that the AMA has carved a niche for itself as a “House of Medicine” and established itself as a place “where physician organizations come together.” The AMA describes its legislative body, the House of Delegates (HOD), as “medicine's most important crossroad” (AMA 2011: 5).
The AMA has shifted toward a professionalized association-of-associations model, which depends on voluntary members distributed geographically throughout the nation but is not as organized around geographic representation. For most of its history, the AMA was an organization of local and state societies. For example, association leaders described the purpose of the organization in 1978 as representing “our profession as a whole by acting as an umbrella organization for state medical associations and many specialty societies” (Barclay 1978: 2082).
The federation concept had its roots in its financial struggles that began in the 1970s. In 1996, the organization formalized the idea of the AMA as a national focal point for other organizations by changing its rules for representation in the HOD. After 1996, the rules put specialty societies on the same footing as state medical societies: the formula for allocating seats in the HOD was identical. Previously, the members of the HOD were local and state medical society representatives, for the most part. The AMA created a list of criteria for membership (see table 2), and societies are reviewed periodically to see if they meet those criteria.
The change in the formula for representation indicates how incentives to change institutional rules are central to the organizations and how the AMA's decision to build a “big tent” for the House of Medicine reflects, as always, a complex set of financial, reputational, and other incentives. The house concept was beneficial for the AMA because specialty societies need to have a proportion of their members join the AMA, meaning the AMA directly receives membership dues from the members of those societies who also join the AMA. The other main gain was reputational, because the AMA gained greater legitimacy through its claim of representation of physicians, even those who are not direct members but only belong to specialty societies. The AMA benefits because it can claim to speak for a larger number of physicians than simply those who are direct members.
Indeed, Michael T. Heaney's work on the identity of organizations is helpful in terms of how organizations position themselves in relation to legislators. The AMA can also claim to be the House of Medicine using two additional strategies. First, the AMA is seen to aggregate physician preferences or “speak for” organized medicine; rather than contact every specialty society, “a legislator can quickly know what their constituents (i.e., doctors) care about this year, so that these concerns do not materialize into a liability in future elections” (Heaney 2004: 617). Second, coalitions are an important way for the AMA to build its reputation as a crossroads for medicine. The AMA and two other organizations are perceived as among the top three bipartisan coalition leaders (Heaney 2006). The AMA is active in recruiting and creating letters that speak on behalf of many other organizations. Sometimes the coalitions have their own “brand,” but in many cases the AMA retains its strong identity and benefits from collaborations organized around specific issues.
In 2002 a Committee on Organization of Organizations was created, which comprised national medical specialty societies, to allow “an orderly transition of the AMA to an organization of organizations” (AMA Board of Trustees 2003). The committee said it would consider the kinds of services a new AMA would provide, such as a role in coordinating various services advocacy, standard setting, publications, and ethics of the profession).
Recognizing that it was facing membership losses, the AMA Board of Trustees initiated a task force to review membership structures around 2010. According to the Board, the goal was to address the growth and stability of its membership and its economic viability. Initially, the task force suggested that the organization create a hybrid membership structure that would have two kinds of memberships: direct membership and society membership (AMA Board of Trustees 2012: 42). This was also described as shifting to a structure with a “core organization,” under what was called a “unity membership” plan, to provide a “much-needed emotional and economic boost for the AMA,” as it was “struggling with declining membership and dwindling influence” (Romano 2001: 24).
Efforts to move beyond a symbolic role for the AMA as the home of the House of Medicine failed. Specifically, the AMA met resistance from specialty societies, which were especially opposed to the AMA levying fees on them for services “that they had been receiving for free” (AMA Board of Trustees 2012: 43). Specialty societies also told the AMA that the organization would not be able to claim specialty society members as direct AMA members. Under the current structure the AMA “can already credibly claim representation of almost all physicians through its House of Delegates,” but under a different structure allowing people to “opt-out at the society level could erode rather than validate representation” (43). The AMA assessed this effort and decided to adopt new strategies: “We have moved away from consideration of structural changes and have begun moving toward enhanced value and a greater focus on recruitment and retention of aggregated and individual memberships” (2012: 44).
Indeed, if organizations have resisted the AMA's proposals to further develop a more encompassing umbrella organization, there is also more direct competition from other organizations. For instance, the Alliance of Specialty Medicine speaks for around fifteen specialty and subspecialty organizations; compared to the membership of the AMA HOD, however, this is a very small group.
Diversified Revenue Sources and Rent-Seeking
Typical strains on organizations include a withdrawal of valued members, a decrease in the supply of available selective incentives, serious conflict over purposes, the challenge of rival organizations, excessive demands on time and energy of key personnel, and loss of morale or lack of a corporate identity (Wilson 1973).
A distinctive feature, perhaps, of the AMA is its long-standing organizational wealth and high levels of engagement, at least compared to other organizations representing professionals. As far back as 1933, the AMA had a net worth of $3,000,000 ($57,643,488 in today's dollars). Around 36% of its income was derived from membership dues and journal subscriptions. The remainder came from advertising in nine of its publications (in addition to JAMA), as well as a mail-order department, a directory of physicians, and an indexing service of twelve hundred medical periodicals.
For 1932, the organization had 98,041 members through county and state societies (who automatically were counted as AMA members), and 60,000 AMA “fellow” members who had direct membership in the AMA but paid small dues. Although perhaps these numbers showed some overlap, the total number of doctors in the United States at the time was 161,300. Other associations had less favorable numbers, with just 49% of dentists belonging to the American Dental Association, 24% of engineers belonging to one of five organizations, and just 16% of all lawyers joining the American Bar Association in 1934 (Shafroth 1935).
The revenue from publications and products stands in stark contrast to the typical interest group, which would be expected to receive a “modest” proportion of its income from publications and sales (Berry 1996: 82). While the financial base of the AMA in the twentieth century was largely driven by its membership income, it benefited from revenues generated by its flagship journal and especially pharmaceutical advertising (Ameringer 2008: 44). AMA revenue from advertising in the 1960s accounted for about 43% of total income (Campion 1984). In recent decades, the organization has become even more business focused through internal groups within the organization, such as the Business Products and Services team, which delivers “impactful and profitable physician solutions and services which fund AMA's mission-focused activities, operations, and administration” (AMA Board of Trustees 2012: 112).
Many organizations have used business income to subsidize their services, meaning membership revenue has become much less important. As a share of its total revenue, the AMA has depended less and less on its membership revenue, from 51% in 1972, with advertising (26%) and subscriptions (10%) the second and third largest revenue streams (see appendix), to just 12% from membership dues in 2016 (AMA 2016). Today, income from books, insurance, publishing and royalties accounts for 83% of AMA's revenue.
Resources are not the only requirement for endurance, however. Even with its sprawling organizational structure and two hundred constituent organizations in the HOD, the AMA has been remarkably cohesive and enduring. It has had some financial challenges, and the loss of members meant the AMA struggled in the 1970s and mid-1980s. By 1985 the president recognized that membership was unlikely to provide a steady revenue stream and noted that “other sources of revenue must be identified and developed” (Sammons 1985: 1585).
While table 3 aggregates revenues in the publishing and health solutions categories, the contribution of royalties, reported only for certain years, suggests that this is at least one-quarter of its revenue stream. Although there has been a drop in sales of AMA coding books, its revenue from coding books and electronic products is important, since the growth in the sale of “books and Digital Content” is “largely on the strength of continued growth in CPT royalties” (AMA 2016: 15). Indeed, in its 2015 annual report the AMA says:
AMA-published books and coding products, such as CPT books, workshops and licensed data files, make up the Book and Products unit. Revenues in this unit increased by $18.3 million. CPT royalties drove this increase, reporting $15.6 million increased revenues in 2015 over the prior year, up 37%, as the market for electronic use of CPT codes continues to expand. Coding book sales also increased in 2015, with overall book sales up $2.8 million in 2015. (AMA 2015: 16)
In 2016, similar results were reported:
AMA-published books and coding products, such as CPT books, workshops and licensed data sales, make up the Books and Digital Content unit. Revenues in this unit increased by $13 million. Royalties drove this increase, reporting a 33% increase in revenues in 2016, as the market for electronic use of CPT codes continues to expand. Coding book sales declined substantially in 2016, with overall book sales down $6.8 million in 2016. The move from print books to electronic data continues to impact these sales, as well as major changes in coding standards, such as required adoption of ICD-10, which drove a large increase in book sales during 2015. (AMA 2016: 14)
Many nonprofit organizations have had to expand the sources of their revenue. In many cases, that revenue may take the form of subsidies from entrepreneurs, other groups, or governments (Hansen 1985).
The AMA may not be unusual in its strategies of developing an increasingly diversified revenue base, but it has been successful in extracting rents from the political process that benefit the organization. If the efforts to formally draw together specialties under one association failed, other mechanisms are in place that do encourage specialty societies to seek membership in the HOD, in order to access key policy committees in the AMA. To address the collective action problem, the AMA has used some possibly unique strategies via quasi-monopolistic markets, some of which it dominates based on prior rent-seeking strategies, and tithing (essentially) organizations. As Robert Salisbury (1969) recognized, public policy decisions create selective benefits that then sustain the exchange process within interest groups when, for example, association membership is mandatory. The AMA benefits heavily from this strategy. Sampling the membership audits undertaken by the AMA over the years reveals around one hundred associations that have been counted as part of these audits, with a total membership across these years of 122,000, including over 50,000 members from the ACP. Specialty societies benefit if many of their members belong to the AMA. This is not to say that all are members of the AMA because they are trying to support their specialty societies, but the fact that so many have dual memberships (close to half of all AMA members) raises interesting questions about the extent to which the AMA has primary or secondary importance (see table 4).
The AMA's use of this strategy is somewhat consistent, given that for much of the its history it benefited from restraints on trade that allowed it to make membership in county medical societies a requirement for such benefits as hospital privileges. These were found to be illegal (see Laugesen 2016). Today, the AMA has encouraged organizational memberships that also deliver a pipeline of members by providing specialty society organizations with incentives to encourage their participation and their members to join the AMA.
On the face of it, the income generated from coding systems would seem like any other organization selling products that are useful to members. But the salable products are only part of the revenue base and do not reflect its role in designing and administering billing codes and the payment system as a whole.
As computerized claims processing developed in the 1960s, common reimbursement system standards were developed to link private providers and payers. The AMA took an early role in developing a proprietary billing and coding system, and the AMA's Current Procedural Terminology (CPT) gradually became more prominent. The CPT is composed of terms defined for medical procedures and developed by each specialty (Kirby 1979: 743) and a hierarchy of physician services organized under the major organ and body systems. In 1983 Medicare made an agreement with the AMA to use the CPT codes, and after that “many private insurers and most other government” programs subsequently began to convert their systems exclusively to CPT; by the late 1980s, CPT had become the uniform code set used for reporting physician services (Dickey 1998: 134).
As said, public policy decisions create selective benefits that then sustain the exchange process (Salisbury 1969); for example, the Farm Bureau won subsidies during the New Deal that then further enhanced wealth (Hansen 1985). This occurred for the AMA when legislation locked in the use of CPT in Medicare and Medicaid in the early 1980s. Later, the Health Insurance Portability and Accountability Act required use of a common coding system in the private and public sectors, and CPT was chosen along with the Healthcare Common Procedural Coding System. These codes include payment codes for supplies, ambulance services, durable medical equipment, prosthetics, orthotics, and supplies.
The Healthcare Common Procedural Coding System, which is managed by the Centers for Medicare and Medicaid Services (CMS), incorporates the CPT system, but CPT is managed and owned by the AMA. Creating new billing codes is not easy, because payment depends on approval from the CPT editorial committee. CPT has allowed the AMA to create a protective and self-reinforcing revenue stream that helps fund its lobbying activities, thereby resolving a classic collective action problem identified by interest group scholars. The AMA not only successfully protected its intellectual property but also used lobbying finesse to lock that system in place.
The AMA has a reliable membership base of medical specialty organizations because the organizations gain selective incentives through participation in two areas of rent-seeking in the policy process: the definition of billing codes, and physician payment policies that are highly specialty differentiated. The AMA's specialty-based membership relies on the AMA's critical reimbursement advisory role, through the RUC. Participation in this reimbursement policy committee of the AMA requires meeting AMA membership quotas for the organizations' members. Without access to the reimbursement process, it is questionable whether those societies could influence reimbursement policy. Thus, specialty organizations need their members to join the AMA. The American College of Radiology, for example, urges its members to join the AMA not just because they should be represented in the HOD but because there are collective economic benefits physicians can gain:
Aside from the HOD, there are multiple areas in which radiologists' input is essential to the success of our specialty. For example, the Relative Value Update Committee, a congressionally mandated committee that determines the relative value of all services in medicine, including imaging services, and the Current Procedural Terminology® Editorial Panel, which ultimately determines which services get reimbursed and at what amount, need essential input from the radiology community. (Ding and Sharp 2009: 727)
The AMA has used past efforts to engage in rent-seeking to increase membership and prestige. It also pursued a novel strategy of pursuing intellectual property rights through its creation of copyrighted billing codes. Its political power and policy-making role have reinforced organizational persistence. It has protected these codes using intellectual property law, and the AMA has successfully fought opponents who want to use their coding system without paying a licensing fee. The AMA's goals as an organization meshed and were protected by its lobbying arm, through changes in government policies that generate rents. The AMA resolves the collective action problem and the mismatch between its political stances and those of most physicians through what may be a unique or unusual strategy of intellectual property law.
A decline in AMA membership likely reflects larger trends in the membership of associations, although no corresponding equivalent decline in associations' influence is apparent. Indeed, the AMA in particular has not lost its influence in key areas (Laugesen 2016). The AMA's persistence and steady policy influence in the face of membership declines may not be unique, and other interest groups may also retain political influence even when their membership numbers are shrinking. Thus, it is possible that the AMA is like many organizations that may have lost members but retained or increased their influence.
This article does not address the question of whether the AMA is unique or unusual in losing or retaining influence and membership over time relative to all other interest groups, although there is likely evidence that might shed light on this, which could be instructive. Comparisons with other physician organizations would put some aspects of the AMA in perspective. While I make reference to understanding the underlying causes of the decline, this article seeks not to explain the reasons for the decline in membership but to document the extent of the decline and to understand how the organization responded.
The AMA used various strategies in the decades after membership began to decline, including trying to get specialty societies to formalize and become organizational members. That effort failed, but the AMA continues to have the support of almost two hundred specialty societies due to its successful efforts to lock in a billing and coding system (CPT) and a committee that advises the government on relative values (RUC) that ensures specialties join the HOD, which therefore ensures a pipeline of AMA members (a condition of participating in the CPT and RUC processes). The feedback effects of earlier rent-seeking efforts on the government side of its lobbying have paid dividends to the organization, which might have been otherwise challenged by a decline in revenue. The AMA has deep pockets for lobbying, which reinforces the status quo for using these two systems, which ultimately ensure the specialty societies are essentially “tithed” to the AMA.
The assistance of Sophia Leenay is gratefully acknowledged, along with the comments of Jeffrey Berry, reviewers, and participants of the conference.