The classic and modified paradigm theories

Associated mainly with the work of Peter Hall (Reference Hall1993), “policy paradigms” conceptualize institutionalized interpretive frameworks – a set of normative values and commitments, and causal beliefs and expectations – that shape policy decisions by cognitively constraining the range of policy options actors conceive as plausible and desirable. Like Keynesianism and monetarism (the original examples Hall used), these paradigms are often initially inspired by some scientific, expert ideas and then become politically taken for granted. Only exogenous economic and political crises that the prevailing paradigm seems unable to resolve can initiate debate on paradigms (Blyth Reference Blyth2002; Hay Reference Hay2018). These crises create a “battle of ideas” (Oliver and Pemberton Reference Oliver and Pemberton2004) between different factions of “idea-carrying” policymakers (Yee Reference Yee1996). Eventually, one paradigm – either the prevailing paradigm or a challenging framework – becomes dominant and beyond scrutiny.

This classic theory of paradigms has been criticized for two main limitations: the limited agency of actors and the assumed internal consistency and incommensurability of paradigms (Berman Reference Berman2013; Daigneault Reference Daigneault2014; Princen and ’t Hart 2014). In these theories, paradigms “trap” policy actors, making them unable to conceive alternatives (Carstensen Reference Carstensen2011, 148). Furthermore, paradigms have internal consistency (Daigneault Reference Daigneault2014), and competing paradigms are essentially mutually incompatible (incommensurable) because “the facts and evidence that are relevant in one paradigm are irrelevant in another” (Princen and Van Esch Reference Princen and Van Esch2016, p. 357). Thus, policy actors cannot use different paradigms at different times or for different policy purposes (Carstensen Reference Carstensen2011, p. 150) nor can the actors upholding different paradigms negotiate or compromise with each other (Princen and ’t Hart 2014). This prevents paradigm changes through the recalibration or combination of existing paradigms.

Some scholars have modified paradigm theory to overcome these problems (Carstensen Reference Carstensen2011; Wilder Reference Wilder2015; Princen and Van Esch Reference Princen and Van Esch2016; Alons Reference Alons2020). In these studies, like language or culture, paradigms structure the communication of ideas and the coordination of decisions by creating a collective interpretive framework, but they do not fully constrain them. Policy actors can act like “bricoleurs,” strategically drawing upon, reconfiguring, and rearranging the elements of existing paradigms to tackle with concrete problems pragmatically (Carstensen Reference Carstensen2011). They can also “mix” different paradigms to build coalitions with their adversaries, even though the interpretive frameworks they are mixing may look contradictory from an external, analytical point of view (Alons Reference Alons2020). Thus, paradigms can change “as elements within a belief system are reordered, added to, or partially replaced by other elements” (Princen and Van Esch Reference Princen and Van Esch2016, p. 359). By emphasizing “intra-paradigm ideational change” (Carstensen and Matthijs Reference Carstensen and Matthijs2018), the modified theories also eliminate the need for exogenous crises to enable significant policy changes.

While the classic theories could not explain the modifications or mixing of different policy ideas through exercise of agency, the modified theories similarly do not sufficiently explain the structural limitations on the flexibility of paradigms. For example, Carstensen acknowledges that new ideas must be “hooked onto” older ideas in order to “gain the acceptance of other actors,” and the bricolage of different ideas is successful when ideas are molded “in such a way that other actors – coalition partners as well as adversaries – come to accept it” (Carstensen Reference Carstensen2011, p. 158), but he does not explain why other actors continue to uphold some rigid and older criteria in their evaluation of new ideas and paradigm mixes, or which older ideas have the potential to assert this rigidity on new ideas. Similarly, Wilder (Reference Wilder2015) argues that paradigms have “relative commensurability” and are, therefore, irreconcilable only when the actors using them perceive them as irreconcilable, but leaves it unclear what perceiving irreconcilability entails.

Given that each approach is individually limited, scholars suggest that we must make strategic decisions in applying each approach to specific circumstances. One popular formulation suggests reserving the traditional paradigm theories for periods of exogenous crises, while using modified theories to understand institutional changes in normal times. In crises, paradigmatic clashes are more likely, as actors hold onto their ideas to reduce the uncertainty they face, while in periods of stability, they can more easily reflect on and revise their ideational positions (Daigneault Reference Daigneault2014; Wood Reference Wood2015). But this dependence on the identification of crisis times and normal times has created more debate and confusion in new paradigm studies, given that this periodization itself is a product of existing ideas (Blyth Reference Blyth2002; Hay Reference Hay2018). Some scholars have alternatively distinguished between paradigmatic and non-paradigmatic policy areas, arguing that the policies requiring “technical policy-making that are administered by a relatively stable policy community” have more paradigmatic rigidity, while those that have more political and day-to-day relevance have more paradigmatic flexibility (Princen and ’t Hart 2014, 272). However, this ignores the existing research on the possibility of paradigmatic flexibility even in highly technical areas like monetary policies (Carstensen and Matthijs Reference Carstensen and Matthijs2018).

There is yet another way to distinguish and select between these two approaches if we focus on policy change proposals instead of periods and policy areas. Both the traditional and modified approaches have prioritized analyzing materialized policy changes, i.e. policy change proposals that were eventually accepted and implemented by the policymakers (Schmidt Reference Schmidt2008, 308). At any given period, however, various policy proposals are concurrently being proposed and debated in a policy field, and some (if not most) fail to materialize. These failed proposals also occur during paradigmatic conflicts and changes. Therefore, rather than perceiving paradigm flexibility and rigidity as characteristics that develop outside of the policy-making process and shape this process by limiting the types of policy changes actors can propose and make, I look at them as characteristics that develop within the policy-making process as actors put forward and evaluate different policy proposals. This allows for the possibility that the same policy area in the same period can be paradigmatically rigid for one kind of policy proposal but paradigmatically flexible for another.

Paradigm modes

In this paper, I develop a theory on how and why the same policy actors can uphold flexible and rigid paradigm modes depending on the type of policy changes proposed and discussed. The elements of this approach are already present in the existing theories’ conceptualization of different types of policy changes, but their assumptions need some clarification and modification before they can be used to distinguish between different paradigm modes.

In classic theories, the first- and second-order policy changes, namely changes in the settings and instruments of policies, can take place through information updates and experiments that happen in “normal” times (so-called “Bayesian learning” (see Blyth Reference Blyth2013)). These theories assume that this is a functional updating of existing paradigmatic ideas which does not make any important changes to the core paradigmatic assumptions. Only the third-order policy changes in the policy goals result from a wholesale shift in core paradigmatic ideas during crises. I propose changing these two assumptions using the literature on policy instruments (Lascoumes and Le Galès Reference Lascoumes and Le Galès2007; Voß and Simons Reference Voß and Simons2014; Howlett Reference Howlett2019).Footnote ¹ Unlike Hall, this literature suggests that instrument choices are determined not solely by functional evaluations but also by “specific worldviews, ontologies, values” (Simons and Voß Reference Simons and Voß2018, p. 18). This means that the debates on policy instrument changes can also lead to substantial intra-paradigm changes, which the modified paradigm theories conceptualize. Furthermore, the instrument’s literature also suggests that the discussions over policy instruments do not necessarily occur in different periods than goal change discussion and can occur simultaneously.

In modified theories, meanwhile, the focus on intra-paradigm changes has replaced the attention on different kinds of policy changes. As paradigms can change through gradual reevaluation and mixing, no meaningful distinction is left between different policy changes: both the first- and second-order policy changes and the third-order changes in the policy goals can take place through gradual ideational changes. However, if we look more closely at their arguments, we can detect that they still make a distinction for policy goal changes. For example, Carstensen and Matthijs (Reference Carstensen and Matthijs2018) argue that the infusion of a new idea into a dominant paradigm typically “does not entail a shift in the goals that guide policy,” and it will take some time and a “potential successor down the line” to turn those gradual policy changes into “profound consequences for the government’s evolving policy agenda” (435). This tacitly acknowledges that gradual ideational changes in the form of paradigm mixtures and modifications are less likely to lead to policy goal reforms, at least not directly and immediately.

Given these clarifications and revisions, we must also revise the question these theories typically ask, i.e. how paradigms shape different kinds of policy changes, and instead question how different policy change proposals affect policy paradigms (their flexibility or rigidity). In other words, the question is not which policy changes are easier or more difficult given the theorized characteristics of paradigms; rather, it is how different policy change proposals shape these characteristics. To answer this, we must again refer to the literature on policy instruments. Based on this literature, I argue that the debates on instrument choices create the permissive structural conditions for policy actors to creatively modify their paradigmatic views, while the debates on policy goals make it more difficult for competing actors to shape and combine their paradigms.

The political discussions on policy instruments have three common characteristics that make them more amenable to paradigm flexibility. First, the beliefs that motivate policy instruments can be orthogonal to the paradigmatic theories that primarily motivate policy goals. Thus, the policymakers upholding different policy paradigms can support the same policy instruments if they share some common causal and normative beliefs on what instruments are desirable (Capano and Howlett Reference Capano and Howlett2020). For example, while Keynesians and monetarists may disagree on what should be the main goals of monetary policies, they may agree on creating independent central banks if they both value depoliticizing monetary policy-making and share the belief that central bank independence prevents politicization. Such agreements on instruments can draw distinct paradigms together.

Second, policy instruments are more amenable to institutional layering (Mahoney and Thelen Reference Mahoney and Thelen2009) than policy goals. Policymakers often propose new instrument layers (or “mixes”) without removing or changing the previous ones (Howlett Reference Howlett2019). Therefore, policy actors can modify or mix their paradigms when proposing a new policy instrument layer, without having to deal with the inconsistencies between instrument layers. By contrast, politicians may pay a higher political price or face more implementation problems when they layer different policy goals (Zittoun Reference Zittoun2014).

Lastly, policy instruments can more easily acquire multivalence, that is, policymakers can value policy instruments for different reasons without a common understanding of their benefits. This is slightly different from the first point on orthogonal motivations leading to an agreement on instruments. Here, actors support the same instrument for different reasons, like an electrician and a musician looking at a wire cord, one seeing the capacity to transmit electricity and the other seeing the capacity to produce acoustic waves. This suggests that instrument proposals can function as “coalition magnets” (Béland and Cox 2016), i.e. vehicles of collective action among actors that have essentially different views. By contrast, to facilitate coalitions between different paradigm-upholding groups, policy goals must be defined very broadly and abstractly, which is not always feasible (Béland and Cox 2018).

Importantly, the discussions on policy goals/instruments do not necessarily or automatically make policy paradigms more rigid/flexible; instead, for different policy proposals to shape paradigm characteristics, specific policy actors must “frame” (Schmidt Reference Schmidt2008) these different kinds of policy proposals in ways that harden or soften the paradigms. This framing involves “problem definitions” (Mehta Reference Mehta2010) or statements on “what has gone wrong” (Blyth Reference Blyth2007). Problem definitions differ from the broad theoretical arguments and explanatory assumptions about general societal problems that lie at the core of policy paradigms. While the broad theories of paradigms construct the need for policy reforms at a general level, problem definitions “highlight different aspects of a given situation,” drawing connections between specific policy proposals and particular pressing social and economic problems of the day (Mehta Reference Mehta2010, p. 39). Moreover, while the broad theories of paradigms are primarily formulated by experts and academics outside of government offices, problem definitions are strategically articulated by policy actors in decision-making roles inside government as they try to justify and make more appealing their policy change proposals to the public and to each other (Zittoun Reference Zittoun2014, p. 10).

Policy actors can use core paradigmatic theories while constructing their problem definitions, but they must use them within the opportunities created by the type of policy changes they discuss. I argue that because of the abovementioned special characteristics of policy instruments, policy actors can act like bricoleurs, constructing problem definitions by loosening (toning down) or intermixing core paradigmatic theories when they propose and discuss instrument changes. For example, because policy actors can perceive the benefits of the same policy instrument from different perspectives (multivalence), they can more freely – without objections from their own group or their adversaries – bend paradigmatic ideas to support instrument changes. This is similar to the concept of “instrument constituents” (Voß and Simons Reference Voß and Simons2014), but the framing strategies are different: while instrument constituents highlight the “merits” of policy instruments “regardless of their feasibility or political viability at any specific point in time” (Béland and Howlett Reference Béland and Howlett2016, p. 398), instrument bricoleurs instead highlight the desirability and viability of an instrument change at a specific moment. When policy actors propose goal changes, meanwhile, they tend to act like paradigm men and frame problems with a devoted, literal, and even exaggerated version of their core paradigmatic theories. For example, because layering is more difficult in policy goals, they would be forced to either directly align their problem definitions with the prevailing paradigm theories or openly challenge them by using alternative paradigmatic ideas. Both the paradigm men and bricoleur roles can be played by the same policy actors that are in central positions in government institutions (e.g. specific government agencies or specialized legislative bodies).

Table 1 summarizes these arguments. While policy change proposals create the permissive structural conditions for different paradigm modes, actors’ frames directly create these modes. When policy goals are being discussed and paradigm men harden core paradigmatic ideas to define the policy problems motivating their proposals, then paradigmatic constraints, coherence, and incommensurability are solidified for everyone. This rigid paradigm mode, as suggested by the classic paradigm theories, prevents most policy actors from imagining alternative policy ideas and options and makes political alliances crosscutting paradigm differences more difficult. When instead policy instruments are being discussed and bricoleurs can creatively reinterpret and mix core paradigmatic ideas while framing the social problems motivating policy changes, this relaxes paradigmatic restrictions, coherence, and incommensurability. As suggested by modified paradigm theories, this flexible mode allows actors to have more agency in pursuing policy ideas and changes conforming to their pragmatic needs.

Table 1. Different modes of policy paradigms

I demonstrate the utility of this categorical theory for understanding policy changes when a policy field is challenged by exogenous crises, policymakers are divided into rival camps upholding different paradigms, and each camp has balanced institutionalized power over policy decisions. The theory on paradigm modes suggests that both paradigm rigidity and flexibility are possible in this situation. In policy goal discussions, the paradigm modifications and combinations expected by the modified theories could not easily materialize and policy actors were locked in dispute, as expected by the classic theories. Nevertheless, significant changes to policy instruments could be achieved by the creative concessions and combinations between competing paradigms, as expected by the modified theories. Thus, when analyzing the effects of policy paradigms on policy changes during crises, we must pay attention to the nature of the policy change proposals and the framing of these proposals by strategic actors, which determine how rigidly or flexibly paradigms affect policy decisions. In this sense, this argument is aligned with the pragmatic theories of paradigms yet differs from them by suggesting that pragmatic needs can sometimes also dictate paradigmatic rigidity. When discussing policy goal change proposals, actors can find it more strategically appropriate to take hardline positions and refuse to make concessions.

Empirical strategy

The purpose of this paper is theory building, rather than theory testing (Eisenhardt and Graebner Reference Eisenhardt and Graebner2007). It therefore does not test alternative explanations based on an established theory of paradigms and their influence on policies but instead looks at the case to derive some descriptive and plausible hypotheses on how different paradigm modes emerge. For this purpose, I use both in-case comparison and process-tracing methods. With an in-case comparison of the proposals debated in the same policy field and period, I can identify the similarities and differences between different proposals and the arguments made to support them, which helps me derive plausible arguments on how the character and framing of proposals shape different paradigm modes. I complement these findings with the process-tracing method (Mahoney Reference Mahoney2016) by looking at how different policy proposals and problem definitions were temporally linked with the expert paradigmatic theories and the political support from or disputes with other policy actors. This helps to reveal the mechanisms connecting expert ideas with political problem definitions and then to connect these problem definitions with different paradigm modes. As Jacobs (Reference Jacobs2015) suggests, I also use process-tracing to contextualize the decisions of the main actors involved in policy debates and analyze how ideas interact with material and institutional conditions.

For this analysis, I selected two significant antitrust legislative proposals that eventually became law and two other antitrust legislative proposals that did not. The successful proposals coincided with those that revised antitrust policy instruments and the failed ones with those that would have revised antitrust policy goals. Table 2 summarizes these four proposals. There are three methodological advantages to using both successful and failed policy proposals. First, looking at what could have happened reveals the important components of uncertainty and unpredictability that would have been missed by looking only at successful proposals. Second, looking at policy proposals that existed prior to policy changes gives a broader overview of the battle of ideas before a policy decision. This way, as Mehta recommends, we can “identify a range of plausible possibilities at a given point in time, and then seek to isolate the reasons why some ideas were more successful than others” (Mehta Reference Mehta2010, p. 31). Last, this method mitigates the problem of deriving the ideas that existed before policy changes from the policies that were finally adopted (Daigneault Reference Daigneault2014, p. 461). After their acceptance, policy reforms can take on new meanings, which can be confused with the original meanings attributed to them by their proponents.

Table 2. Four major antitrust policy changes discussed in Congress in the 1970s

Data Availability Statement: This study does not employ statistical methods, and no replication materials are available.

As data resources, I compiled all the public hearing records and research reports published by two antitrust subcommittees in each chamber of Congress between 1968 and 1980: the “Subcommittee on Antitrust and Monopoly of the Committee on the Judiciary in Senate” and “Antitrust Subcommittee (Subcommittee No.5) of the Committee on the Judiciary in the House of Representatives.” After filtering out the documents unrelated to the selected policy change proposals and adding documents from other congressional subcommittees, antitrust agencies, and executive committees that discussed these proposals, I arrived at 42 documents. Their length varied between a few dozen to several hundred pages. These documents provided detailed qualitative evidence on the kinds of public arguments congressmen, administrators, experts, and the members of the executive made on antitrust policy problems and their solutions. I used a qualitative content analysis software to code the different frames, terms, and logics used by different paradigm-carrying groups, and the arguments made in favor of or in opposition to different policy change proposals. In addition, I used the works relevant to antitrust policy by well-known antitrust intellectuals of this period and the relevant sections of the “Legislative History of the Federal Antitrust Laws and Related Statutes” (Kintner Reference Kintner1978b) to give these policy debates intellectual and historical context.

Competing antitrust paradigms and the economic context

By the late 1960s and early 1970s, the USA antitrust policy field was divided into two competing paradigms. In the 1940s, Harvard economists and law professors (e.g. Carl Kaysen, Donald Turner) formulated the “structure-conduct-performance” (SCP) model of market competition, which became institutionalized as the dominant antitrust policy paradigm in the 1950s. This paradigm suggested that market concentration (i.e. low number of competitors in a product market) creates anticompetitive business behaviors that harm the efficient allocation of resources in the market (Kovacic Reference Kovacic1989, p. 1413). Thus, the main antitrust policy goal under this paradigm was to reduce market concentration by preventing concentration-increasing mergers and other business strategies to eliminate competitors. This paradigm reached its height in the 1960s, when the Warren Supreme Court fully embraced it to attack corporate actions that carried even the smallest risk of increasing market concentration.

Starting in the 1950s, a number of prominent scholars in the University of Chicago Law School (e.g. Robert Bork and Richard Posner) strongly criticized what they saw as “extensive confusion” in antitrust policies under the structuralist paradigm (Posner Reference Posner1987; Teles Reference Teles2008). They argued that the only goal of antitrust should be “consumer welfare,” that is, the “maximization of wealth and consumer want satisfaction” (Bork Reference Bork1967, Reference Bork1978; Posner Reference Posner1976). They suggested that the structuralist paradigm misconceived as anticompetitive conduct corporate strategies that helped increase productive efficiency and competitiveness (Bork and Bowman Reference Bork and Bowman1965; Bork Reference Bork1978). Most business strategies accused of eliminating competitors were in fact “either competitive tactics equally available to all firms or means of maximizing the returns from a market position already held” (Bork and Bowman Reference Bork and Bowman1965, p. 366). They proposed using neoclassical economic theory to examine the anticompetitive effects of corporate practices to correct such mistakes. By the late 1960s, this Chicago School perspective had gained substantial support within academia.

At the same time, the Chicago School ideas also started spreading inside the government. While the Nixon and Ford administrations did not commit to a clear antitrust policy, they were leaning toward the Chicago School approach (Peinert Reference Peinert2020). This was best evidenced by the judges they appointed to the Supreme Court and high-level federal courts, who used Chicago School ideas more often than their peers (Hovenkamp Reference Hovenkamp2018, p. 600). But more importantly, the Chicago School paradigm became increasingly dominant inside antitrust agencies, namely the Antitrust Division in the Department of Justice (DOJ), and the Federal Trade Commission (FTC), as they began employing more economists and economic analyses in their analysis of cases (Eisner Reference Eisner1991; Davies Reference Davies2010). Under the Assistant Attorney General (AAG) Thomas Kauper (1972–1976), the DOJ’s Antitrust Division sought “a greater capacity for economic analysis … both in terms of the development of specific cases … and in the development of an overall program that made economic sense” (Kauper quoted in Eisner Reference Eisner1991). Similar changes were occurring inside the FTC. As its chairman testified in 1974, “the Commission recently had had a renaissance, one might say, with respect to a reevaluation of what the competition is attempting to do” (Engman in (Congress 1974b)).

Although it was losing its hold over antitrust academia, antitrust agencies, and courts in the 1970s, the structuralist paradigm still resonated with congressional antitrust subcommittees. These subcommittees were composed of senior and high-ranking members of both political parties. For example, the chairman of the House Antitrust Subcommittee also chaired the House Committee of Judiciary. From the 1950s, these subcommittees regularly held oversight and investigation hearings to advocate for more antitrust law enforcement on large corporations and sponsored the amendments to expand antitrust laws. For example, the Celler-Kefauver Act of 1950, which created a stronger mergers and acquisitions (M&A) control system, was sponsored by the two chairmen of these subcommittees. Until the elections in 1977, there was remarkable continuity in these subcommittees, with some members holding their positions for decades. As a result, most of these congressmen continued to uphold structuralist ideas.

The late 1960s and early 1970s in the USA were also marked by a growing sense of political and economic crisis. Two situations were particularly relevant to antitrust policies. First, with the USA economy opening up to more foreign trade, American companies had suddenly become uncompetitive against those in Western Europe and Japan (Christophers Reference Christophers2016, p. 236–237). This led to a decline in exports and an increase in imports and caused business foreclosures, rising unemployment, and declining domestic investments. Second, mainly because of international oil price increases and the Vietnam War, domestic prices were rising, especially in food products, which impacted ordinary Americans. The Nixon government was forced to implement price controls, but these measures were unsuccessful, and the inflation rate reached double digits by the mid-1970s. Both stagnation and inflation crises could be connected to market structures and corporate size, which concerned antitrust policies.

In this policy field divided by competing paradigms and put under strain by exogenous crises, policymakers proposed four major reforms to the USA antitrust laws, only two of which succeeded.

Rigid modes of antitrust paradigms

The RP Repeal Act and the Industrial Reorganization (IR) Act aimed to contract or expand the types of business conduct deemed illegal by antitrust laws, thus reforming the official goals assigned to antitrust policies. However, these proposals never became law because of their failure to gain sufficient political support in a divided policy field. Their failure can be explained by the rigidity, internal consistency, and the incommensurability of the antitrust policy paradigms during the discussion of these two proposals. In these proposals, broad paradigmatic theories were used, highlighted, and exaggerated by the paradigm men articulating specific problem definitions in connection to the ongoing crises. Consequently, policy actors’ perception of the problems clashed, they were unable to make sense of each other’s problem definitions and solution proposals, and they perceived that the proposed policy changes would leave no room for different perspectives in policy decisions.

The RP Act repeal

The RP Act of 1936 prohibited price discrimination (selling the same product for different prices to different buyers) and below-cost pricing (also called “predatory pricing”) based on the argument that large businesses could easily eliminate smaller competitors by using these strategies and later harm consumers by increasing their prices. The repeal of this act came under serious consideration in 1969 when two expert committees, one led by Phil C. Neal (“Neal Report”) and the other by Chicago School economists George Stigler (“Stigler Report”), both recommended it. With the support of the Ford administration, the antitrust agencies took up these proposals and introduced two statutes to Congress, “The Predatory Practices Act” and “The Price Discrimination Act” in 1975.

The RP Act went against the Chicago School’s goals to protect consumer welfare and productive efficiencies, and its repeal would have been a significant shift in the USA antitrust policy goals. Chicago scholars saw the RP Act as the pinnacle of misdirected antitrust policies for protecting small companies. For example, they called a 1967 Supreme Court decisionFootnote ² that expanded the reach of the RP Act “the most anticompetitive antitrust decision of the decade” (Bowman Reference Bowman1967). They argued that price discrimination and below-cost pricing were in fact efficient business practices in the interests of consumers and should always be allowed. Posner argued, for example, that “A supplier might offer a discount or allowance to one distributor, but not to others, because the distributor did a better job of advertising. To object to such a discrimination would be tantamount to disapproving the payment of an extra bonus to a salesman who turns in an outstanding performance” (Posner Reference Posner1969, p. 56).

The policy actors supporting the repeal of this Act, mainly the antitrust agencies, grounded their proposal on these Chicago School ideas but went beyond the theoretical merits of this repeal. Solidifying and even exaggerating the core theoretical arguments of the Chicago School, they argued that repeal of the RP Act would help resolve inflation and stagnation. Policy actors argued that the RP Act created price inflexibility, preventing companies from producing goods more efficiently and reducing prices. For example, Neal testified in Congress “preventing discrimination is essentially a measure aimed at price cutting,” which Congress could not allow to continue, especially “when great efforts are being made to combat inflation and hold prices down” (Congress 1969). Similarly, an FTC commissioner argued: “If the FTC forces every chain grocery store and drug store in the land to pay the same price that is paid by the “mom and pop” grocer and the neighborhood pharmacist, the effect will inevitably be to eliminate a vast array of price discounts and thus raise the overall price in those important sectors of the economy” (Thompson in (Congress 1974c)). By connecting the RP Act with inflation and stagnation, the proponents sought to motivate Congress to repeal the RP act out of practical necessity.

However, the congressional antitrust subcommittees were firmly against repealing the RP Act based on a conflicting understanding of the problems. Against the Chicago School-based argument that the RP Act was harming consumer interests, they contended that “the interests of the consumers are best served by the preservation of small business,” as the RP Act intended (Congress 1969). Moreover, they questioned the accuracy of the agencies’ problem definition, arguing that it was “based upon an abstract economic model” and did not conform to the “realities of business” that they conceived (Congress 1969). They argued that the RP Act was in fact contributing to lowering inflation by forcing large businesses to give discounts indiscriminately. If the RP Act was repealed, the large businesses could impose higher prices everywhere. They even made a counterproposal to expand the RP Act with the “Antitrust Improvements Act” of 1975 (S. 1284). Furthermore, they perceived that the repeal of the RP Act would create a significant change in the totality of antitrust policy goals. This is evidenced by how they reprimanded the FTC for refusing to enforce the RP Act rules in the late 1960s as “picking and choosing” from the totality of the statutory antitrust goals recognized by Congress (Dingel in (Congress 1969)). With this strong resistance from congressional subcommittees, the proposal to repeal the RP Act was never realized.

The industrial reorganization act

The IR Act was one of the most debated antitrust proposals of the early 1970s. It aimed to order major companies in the concentrated sectors of the economy to break up (divest) their assets (see Jones Reference Jones1973). The Neal Report first recommended deconcentrating industries where the four largest firms held 70% or more of the market in 1969. Senator Philip Hart, who was the chairman of the Senate Subcommittee on Antitrust, introduced this proposal as three pieces of legislation in 1972 (S. 3832), 1973 (S. 1167), and 1975 (S. 1959). Hart’s proposal was even more ambitious, proposing to deconcentrate sectors with a four-firm market share of 50%. The Senate Antitrust Subcommittee organized a nine-part hearing series from 1973 to 1975 to discuss and plan breaking up select concentrated industries (Congress 1973).

The IR Act was in line with the structuralist school of the day. In the 1960s, some structuralist scholars had argued that the USA antitrust policies should fully embrace the structuralist goal to deconcentrate the economy by targeting not only “market concentration,” that is, concentration in specific product markets, but also “aggregate concentration,” that is, concentration over the overall economy (see Rill Reference Rill1966; Davidow Reference Davidow1968; Reilly Reference Reilly1968; Adams Reference Adams1968; Mueller Reference Mueller1970). This would particularly target conglomerate mergers, which escaped antitrust enforcement because they did not increase market concentration by combining companies operating in unrelated industries (Congress 1964). These ideas can be traced back to Galbraith (Reference Galbraith1967) influential book The New Industrial State, which argued that antitrust policy was a “charade” for failing to prevent the rising power of giant conglomerate companies.

The proponents of the IR Act, mainly the congressional subcommittees, based their proposal on the theoretical merits of a deconcentrated economy but also exaggerated the core structuralist arguments to link this proposal with inflation and stagnation. The antitrust subcommittees held extensive hearings on conglomerate mergers, emphasizing that these were “one of the most important and pressing economic, social and political problems of America’s recent history” (Representative McCulloch in (Congress 1970b). They argued that conglomerations were able to “administer” (control) prices in the economy and, therefore, inflation was not the result “of impersonal market forces but of conscious decisions by the firms involved” (Congress 1974a). This argument went above and beyond the structuralist scholars’ arguments.Footnote ³ Yet, many congressmen expressed their belief that “concentration of power is the structural cause of our simultaneous inflation and depression” (Horton Reference Horton2017, p. 206). They therefore contended that breaking up concentrated power would reduce consumer prices and create more productive growth. Senator Hart stated that his proposal “bears the seeds of a reform program that could produce an economy in which inflation and high unemployment would no longer be a way of life” (Hart Reference Hart1971). Similarly, the Joint Committee on Inflation in 1974 and several congressional reports supported the IR Act as a feasible solution to inflation (Congress 1974a, 1976b).

The antitrust agencies voiced strong objections to the IR Act, either directly, or indirectly by bringing in Chicago School-based scholars to testify in Congress. They argued that “a comprehensive deconcentration policy could do great harm by decreasing efficient performance and removing part of the incentive toward improving performance in a behaviorally competitive industry” (Congress 1973). They also raised doubts about the connections drawn between market concentration and inflation and stagnation (the problem definition). For example, AAG Kauper stated “My own feeling … is one of some skepticism…. We are dealing with a very generalized economic theory, which says and predicts that concentrated industries behave in a certain way … I don’t think that as of now the data in many industries is enough to tell us what these results are” (American Bar Association 1973). They argued that instead of helping inflation, deconcentration would “likely cause higher prices to consumers” (Wesley Liebeler in (Congress 1973)). The antirust agencies’ expression of skepticism was an important factor in creating a sense of unpredictability over the outcomes of the IR Act and leading to its ultimate defeat. Even a supporter of this act argued that “there is no guarantee that the actual consequences of the bill, if legislated, will be the intended ones” (Samuels Reference Samuels1975, p. 282).

Flexible modes of antitrust paradigms

By contrast, the Tunney Act and the Hart-Scott-Rodino (HSR) Amendment would not change the standards and goals of antitrust policies, but they would significantly alter the criminal or the regulatory instruments the authorities can use in implementing antitrust policies. They succeeded in attaining sufficient political support in a divided policy field because they were based on paradigmatic flexibility, concessions, and mixing. Bricoleurs played an important role in softening, soothing, and amending paradigmatic core theories while they articulated specific problem definitions supporting these two proposals. Thus, other actors could also align their own problem definitions with the proposed reforms, were able to make sense of each other’s problem definitions, and perceived that there was room for multiple policy goals inside the proposed policies.

Tunney act

The “most significant provision” of the Tunney Act was the increase in the maximum fines and prison sentences on criminal antitrust cases (Congress 1974c). Since the Sherman Act of 1890, antitrust fines had been revised only once, in 1955 (Congress 1970a). The Stigler Report recommended an increase in antitrust penalties in 1969 (Stigler Reference Stigler1968). Nixon’s Council of Economic Advisers brought this to Congress in 1970, proposing an increase in fines from $50,000 to $500,000 for corporations. Later, in 1974, Ford delivered a message to Congress proposing to further increase the maximum fines to $1 million (Congress 1974c). AAG Kauper also proposed increasing the maximum imprisonment to five years, thus upgrading antitrust violations from misdemeanors to felonies. The law was passed in 1974 with these last-minute changes.

The Tunney Act had a complicated relation with the competing paradigms. The increasing fines and prison sentences would mainly concern horizontal price-fixing arrangements between competing firms (also called a “cartel”). According to structuralist scholars, such arrangements would happen more regularly in a concentrated market economy by allowing the remaining competitors to collectively dominate a market without having to eliminate each other. Congress and the courts had thus reserved the biggest antitrust law sanctions and fees for price-fixing agreements.Footnote ⁴ The Chicago School had a more ambiguous stance on price-fixing. Nevertheless, most prominent Chicago scholars agreed that horizontal price-fixing was likely to cause substantial efficiency losses and condoned higher punishments for this conduct, since they advocated for adjusting criminal sentences to the seriousness of economic damage (Ghosal and Sokol 2014).

Like the other proposals, the Tunney Act had to be connected to the existing crises of the day to gain sufficient support, but these problem definitions used core paradigmatic theories loosely and with major adjustments. The antitrust agencies (particularly the DOJ), who were the main sponsors of this act, relied on the Chicago School idea that horizontal price-fixing was harmful to consumer welfare and deserved a punishment proportional to the economic harm, but they also made a connection between price-fixing and inflation, which contradicted core Chicago School theories (see Stigler Reference Stigler1962). Nevertheless, the antitrust agencies argued in Congress that strong enforcement on price-fixing could help resolve inflation. For example, when asked to comment on how his agency could play a role in battling inflation, Kauper testified in Congress that “It is the responsibility of the Antitrust Division to identify price rises which have not been compelled by increases in labor or material costs and to then investigate whether those unexplainable price rises are a product of collusion,” adding “Antitrust actions with the greatest short-run promise [to battle inflation] include a redoubling of efforts to detect and prosecute price-fixing conspiracies” (Congress 1974a). President Ford’s administration also used this problem definition, stating “heightened antitrust enforcement is a significant weapon in the current fight against double-digit inflation” (cited in Handler Reference Handler1975).

The proposal to increase antitrust sanctions in the Tunney Act had “no real opposition” (Kintner Reference Kintner1978a). The congressional antitrust subcommittees immediately supported the proposal. For example, in 1970, Senator Hart commended the antirust agencies and the Ford administration for offering this bill, for “it increases the effectiveness of our antitrust laws as a deterrent to harmful economic concentration” (Congress 1970a). As this statement suggests, the congressional subcommittees interpreted the merits of the Tunney Act based on their structuralist theories, although the supporters of the act did not use these theories. Congressmen were also very pleased that antitrust agencies acknowledged the connections between anticompetitive conduct and inflation, which they had refused to acknowledge in the previous discussions over the IR Act. This is evidenced by the fact that, when the agencies requested budget increases in 1975, arguing that tackling “price fixing means more criminal work, and thus more use of grand juries, also requiring additional resources” (Congress 1975a), Congress offered a 200% budget increase.Footnote ⁵ Therefore, as one historian of this law stated, Congress passed the Tunney Act based on its “frequently expressed belief that anticompetitive activities by business were in part responsible for the national economic difficulties existing at the time of enactment” (Kintner Reference Kintner1978b, p. 6537).

HSR amendment

The HSR Amendment created the pre-merger notification and clearance system that requires M&As deals above a certain size to be reported to the antitrust agencies 30 days before the deals are closed. Without a pre-merger notification system, antitrust agencies often missed the chance to intervene in not-yet-finalized mergers because the agencies could only rely on publicly available information on these mergers (Congress 1975b). In 1975, the DOJ requested more expansive Civil Investigative Demand (CID) powers to enable the collection of information on pre-consummated mergers.Footnote ⁶ The congressional subcommittees saw this as an opportunity to create a merger notification requirement and proposed the HSR Amendment.

Like the Tunney Act, the HSR Amendment had an unclear relationship with the competing antitrust policy paradigms. It expanded merger rules and made merger control more effective and was therefore in line with the structuralist emphasis on keeping markets deconcentrated by preventing concentration-increasing mergers. However, it could also be in line with the ideas of the Chicago School, which condoned merger control on the horizontal mergers (between competitors) creating very high levels of market concentration (Bork and Bowman Reference Bork and Bowman1965) and also criticized the use of divestments in merger control for creating inefficiencies and advocated for a more “efficient” merger control system based on the regulation of mergers (see Stigler Reference Stigler1968).

The proponents of the HSR Amendment promoted it by framing a problem definition based on a creative mixture of core paradigmatic ideas. The congressional subcommittees highlighted how the HSR Amendment would help prevent market concentration (Congress 1975b; see also Horton Reference Horton2017). However, they also acknowledged that the use of courts for prosecuting and divesting completed mergers could also be harming the economy. They argued that the divestment of merged companies was like trying to “unscramble the eggs,” which was wasteful, and achieved very little, very late (Congress 1975b). For example, Senator Hart, who supported the HSR Amendment after his own IR Act proposal failed, argued “Anyone who has looked at the problems in undoing a merger knows that, if the merger is not to be allowed, all the country and the companies would be much better off if it is never born.” He added that merging companies postponed and dragged litigation “so many years until the damage to the economy is irreparable and the ultimate vindication of the Government moot” (Congress 1975b).

The HSR Amendment generated more debate than the Tunney Act, but this debate contributed to furthering the flexibility and mixing of the competing paradigms. Since the beginning, the antitrust agencies supported the pre-merger control system in the HSR Amendment “in principle” (Congress 1975b). They argued that the use of economic analysis and data in merger control required substantial information from companies, which only the pre-merger notification system could provide (Congress 1975b). Furthermore, this proposal would help them deal with “midnight mergers,” that is, mergers planned and finalized very quickly to evade the antitrust agencies’ scrutiny. Nevertheless, the agencies also scrutinized the merits of some parts of the original proposal: an automatic halt for M&As if the agencies filed a complaint in court within 30 days of notification, and an “escrow provision,” which would separate and fix the value of the acquired stock and assets during this pause (Congress 1975b). They argued that these provisions would discourage companies from merging and “unless there is a recognizable harm, businessmen should be permitted to make and implement business decisions without the sort of disincentives this provision would create” (Kauper in Congress 1975b). They recommended “rather than mandating a court … to enter an order prohibiting consummation of a merger pending final judgment, the law should permit a court to require a showing by the government of probable illegality” (Congress 1975b). The antitrust subcommittees ultimately agreed to this demand and this concession furthered the mixing of paradigms by forcing the structuralists to acknowledge some efficiency limits over merger control. Thus, both paradigms agreed that divestment of merged companies was harmful to the economy. For example, in support of the revised bill, Senator Humphrey stated: “I believe that the Government ought to protect the American public from concentrations of economic resources which inhibit price and product competition, but I do not believe that the Government should act on whim, or simply cause people trouble, or act without a foundation of fact. This would be in the interest of neither economic stability nor social and economic justice” (Congress 1976a, p. 17, 704).

Discussion and conclusion

This article has argued that policy paradigms’ rigidity or flexibility develop endogenously within policy change proposals and discussions. The rigid paradigm mode emerges in policy goal change proposals and prevents most policy actors from imagining alternative policy ideas and options. When, instead, policy instruments are being discussed, a flexible paradigm mode emerges, allowing actors to have more agency in pursuing policy ideas and changes conforming to their pragmatic needs. I have shown the utility of this theory when analyzing policy changes in a paradigmatically divided and crisis-ridden policy field. If we do not assume that paradigms are always rigid or always flexible and we allow them to have rigid or flexible modes in the same policy field depending on what kinds of policy changes are being discussed, we can better understand why certain policy changes are ultimately rejected or accepted by policy actors when there are exogenous crises and competing paradigms. Both exogenous crises and paradigmatic competition play important roles in policy changes, but in different ways than the classic and modified paradigm theories assume. As different policy change proposals are made in connection to exogenous crises, the rejected policy changes can indicate growing paradigm rigidity in those specific debates, while the accepted policy changes can reflect paradigm flexibility, rather than a new paradigm superseding a previously dominant paradigm. Future research should test these ideas in different policy fields and different contexts, where different conditions of paradigmatic competition and exogenous crises can have a different impact on policy proposals.

Another important implication of this theory is that once paradigms become more flexible and mixed during instrumental policy change discussions, it may become more difficult to make them more rigid and firmly separated later. This may explain why the use of Chicago School ideas became increasingly commonplace in the USA antitrust policy decisions after the 1970s, although it never became institutionalized in antitrust law statutes. On the one hand, the successful legislation of the Tunney Act created an alignment between the structuralist and Chicago School perspectives on the destructive potential of price-fixing agreements. In the 1980s, the advocates of the Chicago School used this agreement to protect themselves from political scrutiny and ejection from the policy field. For example, during his confirmation hearings in Congress, Reagan’s AAG William Baxter was drilled on his assumed intentions to cut down on antitrust investigations, since his affiliation with the Chicago School was by then well-known. Baxter replied, “I would not be here unless I intended to enforce the antitrust laws very, very vigorously” (Baxter et al. Reference Baxter, Applebaum, Barnett, Holmes and Pollock.1981) and outlined his plans to increase the DOJ’s criminal prosecution of price-fixing cases. On the other hand, the HSR Amendment created an agreement between the structuralists and the Chicago School that divestment of completed M&As was economically harmful and should be avoided at all costs. Thus, the antitrust agencies were justified in creating a more regulatory system of merger control based on structural and behavioral remedies negotiated during the notification period and replacing the statutes and courts’ case law with their “merger guidelines.” This allowed, for example, the 1982 Merger Guidelines to introduce Chicago School-based evaluation criteria and priorities into the government’s official merger control policies.

These partial findings suggest that policy instrument revision proposals may be the gateways to paradigmatic change through the softening of competing paradigms. In classic paradigm theories, instrumental changes are assigned a secondary importance and no relevance for paradigm changes, while the modified paradigm theories entirely ignore the differences between different types of policy changes in terms of their paradigmatic effects. This study suggests that by succeeding to break paradigmatic restrictions over policy actors’ perceptions of alternatives and their reconciliation with opponents, instrumental changes can play a much bigger role in paradigm changes than previously assumed. Thus, future research should investigate not only when and how paradigm flexibility is possible but also the long-term consequences of this flexibility on paradigm and policy changes.