II. FIRST ROUND: INFLATION-INDUCED VS. INFLATION-FEEDING BOTTLENECKS

Understanding policy requires assessing the theories that give it support. This section outlines the long-lived debate over inflation theory and price stabilization policy, the so-called structuralist–monetarist stabilization debates that took place in the 1950s and the 1960s (see Baer and Kerstenetzky Reference Baer and Kerstenetzky1964; Boianovsky Reference Boianovsky2012; Bianchi Reference Bianchi2016).^{Footnote 2}

The historical context of this debate highlights a State-led industrialization process that was set in motion in the postwar period, having as the main features a persistently high inflation, resilient balance of payments deficits, secular structural public deficits, and hard-to-overcome concentration of wealth (see Simonsen Reference Simonsen1995 for a brief historical overview of inflation in Brazil). Among these traits, inflation has become a dominant aspect of the region’s economic history, a phenomenon deeply embedded in its institutional fabric and mindset. It grew to become the forefront enemy and triggered the innovative streak of analytical breakthroughs that undergirded the successful stabilization plans in the early 1990s (see Bastos Reference Bastos2002).

The monetarist–structuralist controversy was staged in the postwar period and can be simply defined as “market competition vs. planning” as best promoters of economic development. The controversy gained momentum due to theoretical and political opposition to the anti-inflationary programs in Latin America. These were sponsored by the Bretton Woods institutions, such as the International Monetary Fund (IMF) and the Bank for International Development (BID), later renamed as World Bank (Hirschman Reference Hirschman1962, p. 82). Actively opposing these policies was the emerging Latin American lineage of economic thought, named structuralists. They proposed that inflation was the result of underlying structural imbalances in the economy, not of excessive monetary emission (Rangel Reference Rangel1963 and Reference Rangel1985).

The structuralist school advanced that the economy was irreparably inflexible due to the widespread existence of bottlenecks and obstacles that inhibited changes. The underdeveloped technological apparatus explained a chronically inelastic supply. State action was needed, as unassisted market processes would have limited power in promoting economic progress (Little Reference Little1982, pp. 20–21). Moreover, the price system was not believed to be an ideal resource-allocating device.^{Footnote 3} The market power of oligopolistic corporations, the specific inelasticity of the supply of agricultural products, and the secular effective demand problem due to the highly concentrated income distribution profile that marked Latin American economies explained the dynamics of inflation. For economists such as Raúl Prebisch, Celso Furtado, Juan Noyola Vásquez, and Osvaldo Sunkel,^{Footnote 4} inflation in underdeveloped countries was an outcome of a deformed economic structure. In order to adequately understand rising prices, the framework should take heed of the interaction between ‘basic inflationary pressures’—resulting in structural rigidities—and the ‘propagating mechanism’ of competing income claims accommodated by monetary expansion. Inflation was then sustained by excess competing claims of societal groups over the economy’s output, which was given by a rigid productive structure filled with bottlenecks and constraints.^{Footnote 5} The policy implications were that dealing solely with inflation attacked only the symptom, not the cause; tackling the inadequacies of the productive structure would render inflation control more effective.

The term “monetarist” was coined to denote this specific set of policies, which usually included tighter credit constraints, cuts in public expenditures, partial wage freezes, devaluation, and the repeal of various types of subsidies and direct controls; it was first used by Roberto Campos, one of the monetarist contenders, alongside Eugenio Gudin (Boianovsky Reference Boianovsky2012). In a nutshell, monetarists held that inflation was always and everywhere a monetary phenomenon (Friedman Reference Friedman1963; Gudin Reference Gudin and Hague1962), and, therefore, could only temporarily promote development (as “economic growth” was dubbed at the time). Lasting prosperity was, however, claimed to be incompatible with the inflation and should, therefore, be curbed as quickly as possible, before it degenerated into unbearable tensions (social and economic). The only effective method, according to monetarists, seemed to be the curbing of excess demand, through a prudent combination of monetary and fiscal policies supplemented by international financial assistance. As a response to structuralist claims, most of the alleged supply inelasticities and bottlenecks were not considered autonomous or structural, but rather consequences of price and exchange rate distortions generated during the course of the inflationary process itself (Campos Reference Campos and Hirschman1962, pp. 69–70). In sum, based on Latin American experience, Campos alleges that bottlenecks were originally inflation-induced, even though at a later stage they may become inflation-feeding.^{Footnote 6}

Against this intellectual melting pot, a few seminal contributions anticipated not only the inner workings of a persistently high inflationary process, but also the nearly insurmountable obstacles policy-makers in the region would face attempting to control it a few decades ahead. Our focus here lies on the “inertial inflation” hypothesis, which finds its origins at this first moment of the controversy, although its heyday was yet to come. We highlight the contributions made by Celso Furtado, Julio Olivera, and Mário Henrique Simonsen.

In his magnum opus, Formação Econômica do Brasil, Furtado ([1959] 1963, ch. 35) tackles the “two sides of the inflationary process”—that is, “the reason why prices rise on a persistent basis” and “the effects of such rise on the economic process.”^{Footnote 7} Inflation and distribution are thus two sides of a single process. Furtado starts with a seemingly fictitious and extreme case (at least at that time), that of a dynamic equilibrium concept of distributionally neutral inflation. Inflation can be “neutral” in its effects on the real economy when “all prices rise simultaneously and at the same pace.” A neutral inflation implies that rising price levels do not affect the underlying distributional profile.^{Footnote 8} Under this dynamic concept of neutrality (e.g., one-year-long circuit), stabilization would engender the very outcome the system is defending itself from, the redistribution of real income. At any point during this period, “there is bound to be a group ahead of others in the struggle for the redistribution of income.” Such a group would benefit enormously from the stabilization of the price level at that point (Furtado [Reference Furtado1959] 1963, p. 239). Dynamically speaking, even if one could find an “average pattern of income distribution in a one-year stretch” that stabilized prices around this pattern, through a series of price and wage adjustments, it was unlikely that every group would be entirely satisfied with the outcome. If all groups are equipped to defend themselves (such as through indexation of prices, wages, and contracts) and have a clear notion of the position they occupy in that moment, the “elevation of the price level goes on displacing the system from one unstable equilibrium position to the other” with no automatic countervailing process to restore the system to a stable equilibrium position (p. 239). To this effect, stabilization becomes quite troublesome, due to this staggered structure of price and wage behavior.

As a result of this process-based understanding of a chronic inflationary process, structuralists contended that an orthodox monetary policy tackled only the propagation factors of inflation, leaving its underlying causes unharmed. These latter comprised both supply rigidities (which generate chronic inflationary pressures) and the monopoly pricing and social tensions that disseminate these specific upward price movements. Hence, a finite elasticity of supply and demand for goods implies that any change in relative prices will encounter a downward inflexibility of monetary prices and wages. These non-monetary pressures then compel authorities into monetary accommodation. Expansion of the money supply becomes nearly inevitable, once the social costs entailed by recession-feeding monetary shortages are acknowledged.

Although notions of monetary endogeneity, lags, and expectations had already been mentioned by Sunkel (Reference Sunkel1958), Julio Olivera (Reference Olivera1960 and Reference Olivera1964) slightly shares the focus of inflation theory on the causes of such imbalances and maladjustments of prices along the chain of production with its propagating effects upon price behavior—hence the notion that structural inflation induced monetary accommodation. If money is a passively adjusting variable, rather than a fully controllable exogenous one, Olivera argued, a non-monetary theory of inflation is better suited to portray economies fraught with such institutional and structural rigidities, for it looks “beneath the monetary surface, into the underlying region of physical flows, real prices and sectional disequilibria” (Olivera Reference Olivera1964, p. 322).

In an attempt to accommodate the structuralist inflation theory in the emerging procedure of mathematical modeling, Olivera (Reference Olivera1967) formalized his argument in a two-sector model, in which differences in elasticity of demand and supply promote sustained structural pressures over prices, termed “structural inflation.” In one equation Olivera decomposes inflation into the “structural inflation multiplier” and “the structural inflation multiplicand” or the autonomous component of inflation. The multiplier depends inversely on the degree of flexibility of both markups and wages. The “multiplicand” is in turn determined by the relative elasticities of supply and demand that dictate how amplified are the pressures on prices arising from excess demand over the supply of the non-progressive sector.^{Footnote 9} Therefore, changes in relative prices prompt a rise of the price level. The magnified effects of both the multiplier and the multiplicand of structural inflation are, once again, due to the sluggish factor movements that render the structure of supply rigid.^{Footnote 10}

To sum up, structuralists emphasized that inflation resulted from deep-seated underlying causes due to uneven structural changes along a historical process of development. Such imbalances were triggered either by the government-sponsored import-substituting industrialization policy or by the tendency of Latin American economies to undergo balance-of-payments crises. These maladjustments then boost rising prices through either cost-push and/or demand-pull channels. Inflation was hence a manifestation of a deformed economy.^{Footnote 11}

On a methodological note, one should be aware that the models of first-generation CEPAL were mostly formalized in discursive form—that is, in a sea of words. As the mathematical revolution slowly settled in economics, Olivera’s framework synthesized all the institutional bells and whistles so dear to structuralists in the values of fixed parameters, thereby assuming rather than explaining these features. This type of mechanical modeling was widely criticized by economists at CEPAL, for they overlooked the historical and institutional particularities of countries (see Palma and Marcel Reference Palma and Marcel1989, pp. 256–257), thus limiting the range of phenomena the theory encompassed.

A second point of relevance regards the strict dependence of the autoregressive component on the price of the wage goods, which singled out wage dynamics as a mechanical cause of inflation, overlooking the role markup behavior plays in the process. Once the agricultural problem (that boosts food prices that feed into wages) is resolved, the inflation memory component would vanish. Consequently, the price level would depend entirely on the relative price variation and supply-demand discrepancies. However, this mechanism alone cannot explain the self-sustaining thrust of inflation on a systematic basis, unless it is simply assumed that both relative prices stability and demand-and-supply equality are seldom obtained.

Edmar Bacha (2003, p. 145) would later state, without explaining, that this understanding of structural inflation became, in the 1980s, the “inertial inflation hypothesis.” But such a connection is not trivial and requires a careful account of the changes in methods and communication of ideas. Olivera’s contribution reveals a tension under the structuralist banner that could have prompted such changes. The methodological instability it created was a factor in the transition to the second round of this debate in the late 1970s. Crucial to this shift is Mário Henrique Simonsen’s hard-to-frame intellectual track. Surrounded by a complex setting of intellectual and institutional influences, Simonsen’s work provides a connection between the 1960s development and stabilization debates and the later inflation stabilization controversy along the 1980s, followed by James Tobin’s take on the subject. Surprisingly, the relevance of Tobin’s input has been greatly downplayed in the historical accounts of the inertial inflation hypothesis. He is one of the first writers to use the expression “inertial inflation” and his writings are filled with the conflict-based understanding of inflation. Both dimensions are crucial to better comprehend the diversity of inflation models that arose in the early 1980s and, among these, the emergence of the inertial inflation hypothesis.

III. THE SIMONSEN AND TOBIN CONNECTIONS

Mário Henrique Simonsen (Reference Simonsen1970a, Reference Simonsen1970b) was the first to implement a formal feedback model of inflation in Brazil. As we have seen, Olivera (Reference Olivera1967) had already delineated a simple inflation model with wage indexation feeding into relative price changes that ultimately affected the rate of inflation. It is worthy of note that the parlance on inflation-related issues is mostly dominated by developmental issues, such as relative price changes, social aspirations, competing claims by income groups, and distributive aspects of economic growth.

In a regime of chronic inflation, Simonsen added, the real incomes undergo an oscillatory behavior that is virtually consistent at the aggregate level, for when some groups obtain their peak real income, others are at their “trough” real income, and others are midway, given the unsynchronized structure of income adjustments, later known as the “sawtooth model” of nominal incomes under inflation (Vera Reference Vera2013). In aggregate terms, Simonsen argued, “everything appears as if individuals, instead of living purchasing power cycles, remained in the comfortable stability of average income levels,” a point Furtado had made about the dynamic neutrality concept a decade before (see section above). The very asynchrony among the various income peak levels may turn inflation economically viable. Were all the peak levels to be simultaneously obtained, the result would be the acceleration of price rises. It is thus the “effect of previous peak levels on the aspirations and expectations that explain the difficulties faced by monetary stabilization.”^{Footnote 12} Simonsen (1964, p. 20) hence ascribed the central cause of such difficulties to the socio-political background, “the incapacity of economic policy to confront the political problem of distributional compatibility.”

Simonsen’s writings reflect the standard of theoretical discourse at that period, although it absorbs by proximity the development-laden vocabulary of the Latin American economic debate. Simonsen synthesizes form and content in such a masterful way as to create an analytical framework for the subsequent exchanges among economists on inflation in Brazil.^{Footnote 13} Departing from neoclassical reasoning, Simonsen moved in successive approximations closer to Brazilian reality. After a thorough analysis of the limitations faced by the quantity theory in explaining inflation in Brazil, in a later work (Simonsen Reference Simonsen1970b), he found it more useful to illustrate his claim by means of a model of feedback inflation,^{Footnote 14} as follows:

$$p = {a_t}{p_{ - 1}} + b{g_t} + {c_t}$$

Where c _t denotes the autonomous component of inflation—or the “cost-push factor,” or even the “supply shock component”—which reflects institutional changes such as minimum wage readjustments, tax increases, or public utilities prices, as well as accidental events, such as bad crops or energy price rises. The parameter a is the feedback coefficient—that is, the pass-through inflation from the previous period to the present—and has values within the range 0 < a < 1, except for the hyperinflationary periods, when its value may exceed the unit. At last, g _t represents the excess demand component; there is a growth rate of output that provokes no inflationary pressure, given by y _n, so that ${g_t} &#x003D; {y_t} - {y_n}$; y_t being the effective rate of growth of output at a given period t. By assuming the absence of adaptive expectations or automatic price adjustments regulated by law (indexation) according to past inflation ($a{p_{ - 1}} &#x003D; 0$), the model illustrates both the demand-pull (g _t) and the cost-push and relative price changes (c).

This simple expression conveys two important aspects of chronic inflations: autonomous shocks have a permanent—although gradually decreasing—impact on the level of inflation rate, which depends on the lagged inflation coefficient, given by the parameter b. Economies facing chronic inflation develop mechanisms (such as formal indexation) to reduce the harmful effects it imposes on the economic organization of households, firms, markets, and government.^{Footnote 15} Therefore, a more disseminated practice of indexation (higher value of parameter a) leads to a higher equilibrium rate of inflation.^{Footnote 16}

Note that focus lies on the self-feeding component that gradually takes over the behavior of inflation, as the process of adjusting prices according to past inflation spreads throughout the economy.^{Footnote 17} This aspect is important in fleshing out the differences in emphasis between his inflation-feedback model and Julio Olivera’s account. The latter underscores (and analytically derives) the “structural inflation multiplicand,” which has to do with the changing structure of production under the import-substituting industrialization process of Latin American countries, whereas the self-feeding component receives less attention. In Simonsen’s account, the structural inflation is dubbed the “autonomous component” (the “c _t” component), but is simply assumed—as a constant—rather than being derived in analytical terms, whereas the propagating mechanism (self-feeding component) assumes a pivotal role in his model.^{Footnote 18}

This apparently minor change in emphasis from Olivera to Simonsen reflects the very transition from the first to the second round of the monetarist–structuralist controversy, which is why Simonsen’s contribution became a cornerstone for subsequent analysis of inflation in Brazil, providing the framework for those discussions in the early 1970s—namely, his early distributional conflict perspective, the self-feeding inflation model, the relevance of unsynchronized price and wage adjustments, his “sawtooth” diagram illustrating the effects of inflation on nominal incomes, and the cautionary remarks concerning the difficulties involved in price freeze policies, also known as “incomes policy,” widely adopted in the 1980s in Argentina, Israel, and Brazil. Moreover, he made constant efforts to merge the analysis of the Brazilian case with the overall developments in economic formal theory building, especially those of a North American lineage.

Another contribution helped to connect the two rounds of the same decades-long controversy over the causes of inflation, while also bridging the gap between the North American and the Latin American structuralist lineages. In a largely overlooked paper, James Tobin (Reference Tobin, Flanders and Razin1981)^{Footnote 19} developed a taxonomy for the diagnostics for inflation that may very well be the 1980s inflation-centered analog of Prebisch’s (Reference Prebisch1949) Manifesto. The passive money hypothesis was recast under new terminology and the monetarist plea was quickly dismissed on such terms.^{Footnote 20} Tobin is most direct and succinct in his classification of the three main types of causes to inflation: inertial inflation, specific price inflation, and conflict inflation.

By “inertial inflation,” Tobin means “the self-replicating pattern of wage and price inflation,” which, “once it is built into the system in a consistent manner it continues very much on its own” (Tobin Reference Tobin, Flanders and Razin1981, p. 23). In such an inertial setting, some prices feed into economy-wide price indexes, and then propagate onto other prices and wages, thereby eliciting the “stubbornness of built-in patterns of wage and price inflation” (p. 24). However, there exists a viable equilibrium path of real outcomes, on which the rate of inflation offers no threat to economic and political equilibrium. As long as the economy has adapted its conventions, institutions (in which we may include monetary and fiscal accommodative policies), and expectations (especially in regard to relative prices), inflation poses no distortion to the economic process.

Finally, Tobin stresses “that inflation is the symptom of deep-rooted social and economic contradiction and conflict” (p. 28).^{Footnote 21} If players cannot agree to divide a growing pie, how will they concur with each other when dividing a smaller one? Moreover, Tobin draws attention to the aspiration gap that Robert Rowthorn (Reference Rowthorn1977) had explicitly modeled a few years before. An orthodox policy restricting demand would curb rates of growth of potential output, of real wages, and of real returns on capital investment, while not necessarily holding down the “standards of real income progress to which employed workers are accustomed or the profit rates that managers and shareowners expect” (p. 28). In this sense, an economic downturn may, through further conflict, widen the aspiration gap, boosting inflation.^{Footnote 22}

Tobin’s missing chapter is important, for it reflects the zeitgeist at the onset of the troubled 1980s, a period in which the economics of the Keynesian Neoclassical Synthesis was fast becoming a heterodoxy, as the perception of “disarray” overtook the economics profession by assault (Goodfriend and King Reference Goodfriend, King, Bernanke and Rotemberg1997). Moreover, it unearths a relevant network of ideas that provided a North American precursor to a seemingly Latin American issue.^{Footnote 23}

The next section looks into the second round of the controversy over price stabilization between the monetarist orthodoxy and the somewhat hard-to-define heterodoxy in Brazil (see Bacha Reference Bacha1982; Toye Reference Toye1987).^{Footnote 24} Structural causes of price rises were sensitively downplayed, as the focus of theory shifted towards the propagation mechanisms of inflation, in contrast with the early debates on the relation between economic development and inflation. From the institutional standpoint, the “Brazilian indexing style” (Fishlow Reference Fishlow1974) made up for the inflation-magnifying mechanism readily recognized as the chief distinction of the Brazilian experience with inflation—e.g., Pérsio Arida and André Resende ([1984] 1985, 1986)—instead of structural and historical features emphasized by the structuralist models of the first generation, as we have seen above.

IV. THE SECOND ROUND OF THE MONETARIST–STRUCTURALIST CONTROVERSY

The world inflation of the 1970s rekindled debates over inflationary processes (Harberger Reference Harberger1978). The 1973 OPEC-driven oil price rise generated a worldwide inflationary bulge that set the grounds for an orthodox reaction against the Keynesian-inherited policies governments led all along the postwar period. The classic divergence re-emerged with strength, eliciting a revival of academic and policy-oriented debates. As we have seen above, the long-lived inflations in Latin America provided the pretext for a further deepening of analysis on such matters. Some ideas were fished out from a not-so-distant past and reshaped under the guises of formal mathematical models, while new ones sprung out of the pressing economic landscape.

The second round of the monetarist–structuralist controversy was triggered by the mid-1970s. The intellectual battle between monetarist economists—such as Affonso Celso Pastore (Reference Pastore1969), Cláudio Contador (Reference Contador1974, Reference Contador1977), and Antonio Carlos Braga Lemgruber (Reference Lemgruber1977, Reference Lemgruber1978), among others stationed at Fundação Getúlio Vargas – Rio de Janeiro (FGV-RJ)—and structuralist-(neo)-Keynesian economists was thus fought, at first, on the empirical turf. The main issues regarded the endogeneity of money supply in Brazil (Lemgruber Reference Lemgruber1977; Cardoso Reference Cardoso1979, Reference Cardoso1981b), the effects of public deficits on the inflation via an accommodative monetary policy (e.g., Contador Reference Contador1974; Cardoso Reference Cardoso1981a, Reference Cardoso1983) and the effects of restrictive demand-management policies upon inflation rates by means of estimations of the Phillips curve for Brazil. The dispute became more contentious in the early 1980s, for intellectual and political reasons. Let us start with the latter aspect.

The 1980s were turbulent times, and the thin line separating academia from policy-making in Brazil evaporated in the midst of a political struggle against the exiting twenty-year-old military regime. In this sense, the writings of such scholars reflect this political atmosphere filled with a drive for change, intellectually fueled by the challenges history posed to Latin American societies. As the political transition became increasingly more secure and the dormant inflationary “dragon” arose from its twenty-year “harmless” sleep, the politically militant arguments gave place to more technical proposals. An important part of such changes has to do with their integration to public offices and their economic teams. Moreover, the economics profession began converging to a macroeconomic consensus at the international institutional level, where most of the scholars under our scrutiny held liaisons in the most prestigious academic and multilateral institutions.

The external debt crisis and the interest rates hikes that followed the oil shock in 1979 triggered a massive inflationary reaction in previously stable developed economies. But the tragedy that befell the Latin American economies was many times more intense. The turmoil was rapidly associated with the immense resource constraint these economies suffered, eliciting a profusion of theoretical analyses of this rare phenomenon of rapidly increasing rates of inflation, historically attached to economies disorganized by wars and the ensuing occupation of the defeated countries. The classic cases of hyperinflation in the first half of the twentieth century seemed, however, structurally different from the hardships Latin America endured, although the process had many similarities with those experiences, which required further investigation.^{Footnote 25}

In Brazil, the first half of the 1980s witnessed large-scale social and political transformations, as the military regime raised growing opposition in numerous factions of society. As it so happens, economists’ battling energy was harnessed towards actively trying to influence policy-making.^{Footnote 26} Abundant historical material on this period shows the prompt stream of short articles published in newspapers engaged in a sweeping controversy.^{Footnote 27} Contributions suggested ways of redirecting policy away from full-fledged orthodox measures launched by Minister of Finance Antonio Delfim Netto in 1980 in President Figueiredo’s term, the last president of the military regime that rose to power in 1964. These latter were perceived as a severe hindrance to economic growth at the very moment the country needed most a booming economy—that is, during a major political transition.

The neo-structuralist economists involved in the controversy found it quite striking that macroeconomic “fundamentals” (government budget and current accounts of the balance of payments) were in place, but left inflation unscathed. This raised doubts concerning the relationship between budget deficits and inflation, as well as the level of proportionality—and the true direction of causality—between monetary aggregates and prices. For instance, Arida and Resende (Reference Arida, Resende, Arida and Resende1986) were adamant in claiming the Brazilian budget deficit was close to zero and the current accounts of the balance of payments were on the positive side. And yet, inflation seemed to be stable at three-digit rates.^{Footnote 28} In this vein, it was argued that when contracts incorporate past inflation—by means of index linkages or backward-looking behavior—conventional anti-inflationary measures of aggregate-demand restriction become ineffective and end up exciting social unrest, once it increases unemployment and aggravates recession.

Therefore, a mix of political discontent and theoretical rebellion fueled a yearning for institutional change and reform both in policy and in academia. As a result, it was politics that eventually summoned seemingly irreconcilable economic approaches under the banner of heterodoxy. However, as we point out below, when this feeble economic consensus was challenged later on—as the economists behind it became policy-makers—compromise proved to be not only partial, but also too soft.

The literature alluding to this period in the intellectual history of Brazilian economics usually skips through the stage of the controversy in which academic exchanges happen, and moves quickly to assess how these ideas have influenced policy design.^{Footnote 29} Alternatively, we assess the analytical developments set forth by these scholars,^{Footnote 30} highlighting the tensions that emerged within the “inertial inflation paradigm” (Arida Reference Arida1992) from the late 1970s until the mid-1990s and quickly gained notoriety in the policy circles and academia.

The combination of internal and external forces unraveled a second-generation structuralist transformation problem, in reference to Carolyn Craven’s (1994) analysis of CEPAL’s self-contentious first moments. However, at some variance with the first-generation problem, we maintain that this change was not a complete rupture with their previous stance on inflation theories. It was rather a displacement of emphasis, allowing for a higher level of consistency with conventional readings of inflation. The stripping of inflation theory of its structural and institutional aspects smoothed the transition of the concept of inflation inertia in the first round—meaning a self-replicating pattern of rising prices—toward the meaning prescribed by New Keynesian economics. The following sections narrate how the internal analytical puzzles within the structuralist camp of the second generation facilitated this move.

V. THE SECOND-GENERATION STRUCTURALIST TRANSFORMATION PROBLEM

The self-titled heterodox opposition to the alleged monetarist status quo centered on the latter’s apparent oblivion to an “indubitable” fact (see Simonsen Reference Simonsen, Simonsen and Barbosa1989, p. 14): that the widespread use of indexing schemes under persistently high rates of inflation had led to a substitution of domestic currency as a unit of account of contracts for other referential assets, whose values are denoted in terms of the legal tender, which comes to define the terms at which contract obligations are met (Carvalho Reference Carvalho1994, p. 109). In a word, indexation tends to institutionalize the price-wage spiral (Simonsen Reference Simonsen1995, p. 5) and imparts a mechanical drive to it (see Heymann and Leijonhufvud Reference Heymann and Leijonhufvud1995, p. 31).

In the early 1980s, heterodox economists viewed inflation in Brazil as a fully inertial phenomena (Resende and Lopes Reference Resende and Lopes1982; Arida Reference Arida1983b, 1984; Resende Reference Resende1984; Lopes Reference Lopes1984; Bresser-Pereira and Nakano Reference Bresser-Pereira and Nakano1984). However, consensus is a rare phenomenon and often conceals profound differences among its participants. Despite the agreement upon the inertial character of inflation, five different models of inflation competed to explain inflation dynamics in the Brazilian economy during the 1980s. The rational expectations model (Sargent and Wallace Reference Sargent and Wallace1975; Sargent Reference Sargent1982) and the Friedman- and Phelps-type adaptive expectations models (e.g., Cagan Reference Cagan and Friedman1956) rationalize acute inflationary processes widely covered by the literature and shall not occupy us here. Alternatively, the institutional interpretation stated that in a highly indexed economy, such as Brazil, a great portion of basic macroeconomic prices (exchange rate, interest rate, wage, etc.) “rise by decree,” according to the previous period’s inflation rate.^{Footnote 31} Alternatively, we focus on two neo-structuralist explanations of a fully inertial inflation: the relative wages model and the distributional conflict model.

The relative wages model proposes that institutional, political, and cultural factors affecting the labor market engender workers’ resistance to (downward) changes in relative wages. In an inflationary environment and unsynchronized structure of readjustments, workers seek to periodically restore their nominal wages according to the peak real wage, thereby stabilizing both the average real wage effective in the previous period and, in the absence of exogenous shocks, the wage structure as well.^{Footnote 32} It consists of contributions by Silva (Reference Silva1981) and Bresser-Pereira and Nakano (Reference Bresser-Pereira and Nakano1984), as well as by economists at the PUC-Rio school, such as Resende (Reference Resende1981), Lopes (Reference Lopes and Sayad1979), and Resende and Lopes (1979).

The distributional conflict model is found mostly in Arida (Reference Arida1982, Reference Arida1992), Luiz Bresser-Pereira and Yoshiaki Nakano (1984), Eduardo Modiano (Reference Modiano1985, Reference Modiano1988), Arida and Resende (Reference Arida, Resende, Arida and Resende1986), Franco (Reference Franco1986), and less prominently in Modiano (Reference Modiano1988) and Lopes (Reference Lopes1985, Reference Lopes1989). In this setting, inflation results from the social impasse derived from an unresolved distributional imbalance between profits and wages—rents, taxes, and other forms of income—that may lead, given some specific hypotheses, to the inertial behavior of inflation. Entrepreneurs and workers have limited power in determining the others’ effective income. The problem relates to the dynamics of the “aspiration gap” (Rowthorn Reference Rowthorn1977) that triggers the wage-price escalation up to levels that simply reproduce the imbalances previously observed. Agents act to defend their peak real wage levels, and the relevant aspiration gap is thus given by the difference between peak and average real wage levels. This gap defines the degree of structural distributional incompatibility, which stems from a social impasse among the main income groups or classes, not from uncertainty or coordination failures in individual agents’ potentially consistent plans of action. This means that inflation inertia will result as long as the terms of conflict in distribution are stable. The level of inflation will be a function of the size of distributional imbalance and of the system of indexation, be it formal or informal.^{Footnote 33}

Both approaches are heterodox in their essence, for the strict refusal of aggregate demand restriction measures. They also share the concept of “inflationary equilibrium,” already present in Tobin’s (1980) assessment of stabilization policy, in which changes in the regime of nominal wages readjustments would only push prices upwards, leaving average real wages unaffected. However, the forces driving this equilibrium have slightly distinct properties. The main difference between these two inertialist models refers to assumptions regarding firms’ markup behavior and the ensuing matter of workers’ rational decision-making. This distinction reflects, above all, the more or less specified form assumed by “distributional conflict” in these models. For Bresser-Pereira and Nakano (Reference Bresser-Pereira and Nakano1984),^{Footnote 34} inflation is but a symptom of the inherently conflictive character of the economic system, whereas in most of the PUC-Rio models,^{Footnote 35} it is mostly due to the staggered structure of the government-sponsored wage-indexation scheme. In other words, this latter modeling strategy takes conflict as more of a placeholder—altogether with other types of exogenous shocks—than a substantial first-order determinant of inflation. It is really an assumption about labor market institutions, not of the economy-wide distributional dynamics. Workers’ aspirations are formally faced with an exogenous upper bound, which they must rationally look up to, so as to at least maintain their relative position within the structure of claims on output. This institutional detail of indexation is a recurrent theme in the models crafted on these static properties of workers’ behavior.

At a more abstract level, the relative wages models (un-)intentionally approximate structuralists to New Keynesian approaches—hence the relevance of Tobin’s input (see section III above).^{Footnote 36} Through this point of contact, doubts were raised that concerned the rationality premises inserted in the model. The first involved explaining why workers bothered claiming for nominal wage increases, if their efforts are self-defeating in the inertial inflation case. The second question involves providing analytical support to the notion that workers would choose to link wages to past inflation.^{Footnote 37} Finally, there was a decades-long puzzle regarding the policy implications of the inertial inflation dynamic: Is it possible to stabilize inflation without upsetting the distributional profile of the economy? In other words, is the distributional neutrality condition a binding constraint of a stabilization procedure? We turn to these aspects next.

VI. THE PUZZLES WITHIN THE INERTIAL INFLATION HYPOTHESIS

Two basic puzzles beset the inertial inflation “theory” as it was stated by Lopes (Reference Lopes1984, Reference Lopes1986), Arida (Reference Arida1982), and André Lara Resende (Reference Resende1981): Why would workers persist in bidding up nominal wages if these latter would always result in the same real wages as before? And, why, in doing so, would they link wage increases to past rates of inflation? Franklin Serrano (2010, pp. 407–409) has found no “rational justification” in their models for workers to behave in such a fashion. In fact, Lopes (Reference Lopes1984) argues clearly that such behavior is independent of the existence of indexation rules or overall index linkages of contracts, or even of adaptive expectations by economic agents; a straightforward defensive behavior or outright militancy by workers could guarantee successive rounds of wage negotiation aiming at restoring peak real wage levels.^{Footnote 38}

Wage Setting and Markup Behaviors: A Basic Framework

Following Bacha and Lopes (1983, p. 3), we begin by defining the real average wage in an indexed economic setting, which can be formally stated as

$$\bar{\omega } &#x003D; v - (\hat{p}/t)$$

where $\bar{\omega }$ stands for the natural log of average real wage (and is assumed to be always greater than zero), v for the peak real wage, $\hat{p}$ for the inflation rate, and t depicts the number of within-year revision of wages (t=1, for yearly revisions; and t=4 for quarterly wage adjustments).^{Footnote 39} If we insert the above relationship into a simple functional distribution profile of the whole economy

$$y &#x003D; m &#x002B; \omega$$

in which y denotes the natural log of real income and m, real markup rate, and $\omega$ for real wages.^{Footnote 40} If we assume $\omega &#x003D; \bar{\omega }$, we find that

$$m &#x003D; y - v &#x002B; (\hat{p}/t)$$

That is, real markup decreases in response to increases in peak real wage, and responds positively and proportionally to price inflation, other variables held constant. In taking time-derivatives of the log values of the variables (and assuming n=1, given its exogenous definition by policy), we get growth rates of each one, denoted by the “hat” sign above them:

$$\hat{m} &#x003D; \hat{y} - \hat{\omega }$$

or

$$\hat{m} &#x003D; \hat{y} - \hat{v} &#x002B; \hat{p}$$

From the above relation, we can delineate three cases. First, a real markup rate ($\hat{m} &#x003D; \hat{y}$) growing at par with real output implies immediately that $\hat{p} &#x003D; \hat{v}$. That is, changes in nominal wages have null effect on the distribution of income, and acceleration of inflation ($\hat{p}$) results entirely from attempts to raise real wage peak levels ($\hat{v} &#x0026;gt; 0$). Moreover, firms accrue every gain from economic growth. Second, $\hat{v} &#x0026;gt; \hat{p}$ would require frequent adjustments on the part of firms to restore their desired real markup rate. This is the stand taken by Bresser-Pereira and Nakano (Reference Bresser-Pereira and Nakano1984) and the PUC-Rio structuralists at their first moments in the debate (see above). The chief point is that incomplete power of groups over the behavior of competing groups implies that adaptation is never complete: the permanent gap separating desired and average real incomes inherently entices new rounds of dispute. Above all, desired levels of income were in no way deemed constant and could be affected by the very process of economic development. The systematically alternated attempts to achieve desired real incomes by each group of income claimants infused an internal thrust into inflation, keeping its rates stable at a given level if peak levels remained constant. Nominal markups could vary freely and faced no determinate upper bound; also, the non-labor costs of imported inputs are taken into consideration, together with interest as a cost-push factor (or the Patman effect) and the inflation-rising force of changing tax rates. Conflict-based inertial inflation is thus a potentially unstable phenomenon.

Alternatively, we can see the third possible case, whereby a constant real markup rate ($\hat{m} &#x003D; 0$) implies that

$$\hat{y} &#x002B; \hat{p} &#x003D; \hat{v}$$

This relationship means that the peak real wage varies with acceleration of price increases and real output growth. Solving for a speeding-up inflation, we get

$$\hat{p} &#x003D; \hat{v} - \hat{y}$$

which clearly holds workers’ income claims accountable for the increase in inflation rates. Were peak wage growth to be paired up with output growth ($\hat{v} &#x003D; \hat{y}$), inflation would be held constant ($\hat{p} &#x003D; 0$). Thus, conflict can be destabilizing solely because wages are pushing beyond the socially agreed-upon income share (which is frequently termed “underlying structure of income distribution”). Then, understanding and taming inflation require reconciling wage demands with other groups’ claims. Wage indexation is thus of paramount importance in this approach. We find such an assumption in the main contributions from the PUC-Rio school.^{Footnote 41}

Next, we look into what this stand on wage-setting behavior implies in terms of firm behavior.^{Footnote 42} If firms seek to maintain a constant real markup over costs, they will adjust nominal markups along the cycle of nominal wage claims. Firms are then logically understood as fully defensive decision-makers.^{Footnote 43} But, if firms simply react to a rise in wage bidding, what triggers workers’ demands? Conflict can arise for a myriad of reasons. In the structuralist tradition, discrepancy between real wage and workers’ targeted real wage stems from a sudden relative price change, such as a large-scale devaluation of the domestic currency or, in the case of Latin American structuralism, high food prices.^{Footnote 44} This aspect points to the question regarding the magnitude of the lagged-inflation pass-through coefficient.

To provide some support to their claims, economists at PUC-Rio marshaled empirical evidence on the assumptions of a=1 and $\hat{p} &#x003D; {\hat{p}_{ - 1}}$; that is, indexation was believed to impose, at that time, a full memory to the price system, generating the fully inertial component in the absence of shocks. However, the conflict dynamics was formally de-emphasized in their narrative by recognizing that incomes policy pursued by the government was the chief cause of the systematic cost-of-living wage adjustments. From this viewpoint, we could then formalize the conflicting claims component as a function of the difference between average and effective real wages as follows: $\left[ {c &#x003D; c\left( {{{\bar{\omega }} \over \omega }} \right)} \right];\,c' &#x0026;gt; 0$. Conflict would thus raise the level ($c &#x0026;gt; 0$), but could not explain the inherent momentum, of inflation (given by the parameter a).^{Footnote 45}

In sum, the principal assumptions behind the differences between the inertialist groups are in regard to firms’ markup behavior and workers’ aspirations. This distinction reflects above all the more or less specified form assumed by “distributional conflict” in these inflation models. As we have seen, both sides share the concept of “inflationary equilibrium,” in which changes in the regime of nominal wages readjustments would only stoke up prices, leaving average real wages unaffected. However, the forces driving this equilibrium have distinct properties.^{Footnote 46} Antagonism between groups collapses into an assumption about labor market institutions, bypassing the complexities of economy-wide distributional dynamics.^{Footnote 47} Workers’ aspirations are formally given an exogenous upper bound, which they must rationally look up to, so as to at least maintain their relative position within the structure of claims on output. No discussion is carried out as to why this peak real wage should remain constant. In its place, the institutional detail of indexation is a recurrent theme in the models crafted on these static properties of workers’ behavior. It fills in the blanks regarding workers’ rationality premises. The first involved explaining why workers bothered claiming for nominal wage increases, if their efforts are self-defeating in the inertial inflation case ($\hat{w} - \hat{p} &#x003D; 0$ or $\omega &#x003D; \bar{\omega }$). The second question implied providing analytical support to the notion that workers would choose to link wages to past inflation ($\hat{w} &#x003D; {\hat{p}_{ - 1}}$).^{Footnote 48}

It is rather telling that such features would pose any trouble to these scholars. If we recall the works by George Perry (Reference Perry1980), Charles Schultze (Reference Schultze1981), Arthur Okun (Reference Okun1981), and James Tobin (Reference Tobin, Flanders and Razin1981) on the self-sustaining thrust of inflation, they relied on conventions, norms, and general institutional details, which were deemed perfectly suitable factors accounting for such behavior. Given their academic background, it is yet more puzzling that the economists at PUC-Rio found it difficult to yield to an argument based on such ad hoc assumptions, which points to the self-imposed requirement of conforming methodologically, in some degree, to the emergent phenomenon of the rational expectations canon. At some point, the argument goes, forward-looking indicators would have to override workers’ behavior.

This requirement engendered enormous analytical trouble. A first solution consisted in assuming several exogenous unanticipated shocks coming from government-endorsed policies such as monetary correction of public taxes and of public utilities fees, maxi-devaluations of the domestic currency, increased frequency of wage adjustments, and so on. Under such circumstances, relative price changes would fool workers into “money illusion,” which would trigger wage bid-ups in the subsequent period. In the absence of such shocks, however, it became uncomfortable to base the self-sustaining behavior of inflation on workers’ adaptive behavior.

The second solution was found in the relative wage dynamics, a shortcoming that Arida (Reference Arida1982) had previously underlined without raising any alarm. A copious literature of the 1960s and 1970s had shown that discrepancies among wages in different industries could give rise to catching-up behavior of labor compensations that would create sustained cost-push pressures on inflation, rendering price indices inflexible downwards (see Laidler and Parkin Reference Laidler and Parkin1975). This inter-sector wage dispute solves the problem but leaves open the question of how workers obtain the information concerning wage adjustments in other sectors. Besides, why should workers disregard, in today’s wage bid, as in backward induction, future wage adjustments claimed by workers in other sectors as a response to the disequilibrium generated by today’s claim? What would prevent the explosive behavior of wage claims? The answer lies in one expression: wage indexation. Once propagation mechanisms have taken over the determination of price-setting behavior, a successful stabilization hinges necessarily on dismantling the widely indexed price and wage structures.

However differently each camp saw fit to classify the causes of inflation and its propagating mechanisms, the nature of policies required to deal with them was not in dispute; both sides of the structuralist camp disparaged any hope leveled at the efficacy of demand-management anti-inflationary measures. Alternatively, stabilization of prices should be achieved through an adequate management of wage policy coupled with a disciplining of markup rates. Conflict should be ameliorated somehow by way of a social pact. Uncertainty came down to one point: Would society adhere to it in a voluntary or compulsory fashion? The next subsection thus addresses the policy implications of the PUC-Rio approach: the requirement of distributional neutrality.

The Binding Neutrality Condition

A successful stabilization should see to three binding conditions. First, the structural distributional incompatibility had to be stabilized, which meant keeping steady the difference between negotiated and actual wages. Second, the wage indexation had to be suppressed, and, third, the system’s inflationary memory had to be wiped out. It should be sufficiently clear by now that these aspects are not independent from each other. For instance, transposing the distributional profile from a highly inflationary environment to one where prices rise only moderately amounts to reproducing the previous inflation-breeding distributive structure. Wiping out the inflationary memory would become, under such circumstances, almost impossible, and, therefore, the same could be said of eliminating wage-indexation schemes.

Following Tobin (Reference Tobin, Flanders and Razin1981), Bacha (Reference Bacha1988) pointed out that inertia could be understood as a coordination problem inside a soccer stadium, where the crowd stands in order to watch the game, when it would better suit the spectators to sit down. However, inertia would break only if everybody sat simultaneously. It would require the referee (meaning the government) to stop the game and threaten to proceed only after everyone has sat down. In economic terms, inflation is a result of unsynchronized behavior in price and wage adjustments. This diagnosis entailed a policy prescription that would bring all incomes to their average levels—and hence to mutual consistency—therefore discouraging agents to try to run ahead of inflation, or, metaphorically, standing up again on the bleachers in order to get a privileged view of the game.^{Footnote 49} This is termed the “distributional neutrality” condition for a sustainable stabilization, and it had two facets.^{Footnote 50} The first implied achieving the conditions for distributive neutrality at the analytical level. The second was sought after at the policy level, albeit conditioned by the former, to which we now turn in more detail.

Lopes (1986, ch. 6) provided a didactic exposition of the “inflationary process,” which is remarkably telling of the way economists at PUC-Rio pictured inflation. The starting point is the typical neo-structuralist national accounting identities, as in Taylor (Reference Taylor1983). After constructing both the indexed wage equation and the inflation equation, closing the model allows us to obtain the inflationary equilibrium:

$$\hat{p} &#x003D; {\hat{p}_{ - 1}} &#x002B; \delta {\hat{n}^&#x002A;}$$

in which ${\hat{n}^&#x002A;}$ denotes the change in nominal markup rates and $\delta &#x003D; {{n_{ - 1}^&#x002A;} \over {1 &#x002B; n_{ - 1}^&#x002A;}}$; as usual, “hats” over variables indicate proportional rates of change. In this simplified version, inflation drifts are caused by variations in markup rates. Should firms stick to the average profit rate of the economy, the inflationary equilibrium is guaranteed. But how do wages behave in this economy? Whenever inflation exceeds expectations, firms adjust their markups to maintain real profit rates. By contrast, when inflation rises, real wages are eroded and income distribution is shifted in favor of firms. To see that more clearly, we can define $\sigma &#x003D; {{wN} \over {pY}}$ as the wage share, which can also be written in terms of real wage ($\omega &#x003D; w/p$) and labor productivity ($Y/N$): $\sigma &#x003D; \omega /e$. Taking logs and then time-derivatives gives us the rate of change in the wage share: $\hat{\sigma } &#x003D; \hat{\omega } - \hat{e}$; the author also defines a target inflation rate (${\hat{p}^&#x002A;}$) workers take into account during the bargaining process. Lopes reached the conclusion that altering the inflation target affects the distribution of income by the following relationship:

$$\hat{\sigma } &#x003D; {\hat{p}^&#x002A;} - \hat{p}$$

This result unveils the daunting difficulties involved in any stabilization plan. Whatever outcome should hold that differs from workers’ fully achieving the targeted inflation rate implies a change in the distributional profile of the economy. If targets are underestimated, wage share is reduced from its inflationary equilibrium point. By contrast, if policy is more effective than predicted, distribution shifts in favor of wages. In the final analysis, inflation will be curbed by use of policy instruments to suppress claims from agents facing relative income losses. How should this be achieved?

The analytical limits of the model were disquietingly narrow, leaving out the solution to distributional effects across firms and across labor groups, as well as functional distribution categories. This hurdle pointed to one additional problem: the inter-sector lack of synchronism of wage and price adjustments. Eduardo Modiano (Reference Modiano1988) proposed one way around this obstacle, which consisted of gradually coordinating contract adjustments by means of a set of conversion rules that steered all contract values towards their average real values, measured in a stable unit of account in the post-disinflation period. Doing so required that no major redistributions of income and wealth were attempted. If these conditions were to hold, a disinflationary “neutral shock” could restore the patterns of income and wealth distribution prior to disinflation. Underlying this neutrality condition was the notion that a fully inertial inflation kept real incomes dynamically stable at their average levels; or, as Furtado (Reference Furtado1954) claimed back in the 1950s, it was distributionally and dynamically neutral.

In the model sketched out above, as well as in Arida’s (1982) model, a monetary contraction or a policy of price controls would necessarily imply a downward adjustment of the markup rate. Therefore, an adequate policy of wage control should accompany such disinflation measures, in order to maintain intact the pre-stabilization distributional profile. Needless to say, achieving such a result in the real world would prove much more complex than theory allowed one to grasp. Besides, a whole range of details were seen only within the scope of partial equilibrium analysis, such as exchange-rate management, the conversion of tax-collecting intervals and interest rates to the new state of low inflation, as well as the portfolio adjustments during the transition.

VII. CONCLUDING REMARKS

This paper aimed to highlight the analytical tensions within the inertial inflation framework, which help to understand why this theoretical endeavor did not manage to muddle through the emergence of a new neoclassical synthesis. By matching together normally unrelated threads of historical developments, we provide a narrative in which the legacy from the Latin American structuralists is gradually absorbed and processed into the newly emerging modeling mathematical formats of the new neoclassical synthesis, at least in its New Keynesian camp. We aimed at revealing the tensions and the selective process by which certain features are dropped in favor of more model-friendly assumptions. The emphasis and connection among the contributions by Furtado, Olivera, Simonsen, and Tobin constitute the main novelties of this paper. They have been duly noticed in early works in the history of economic ideas, but our narrative stands out in that we seek not to provide a rational reading of history, but rather an account of how difficult the construction of consensus can be within the community of economists, even on an abstract theoretical level.

Instability stemmed first and foremost from the choice of method. The mechanical models built by the neo-structuralists (unintendedly) de-emphasized institutions and structural features, which were denoted by specific “parameters” in ahistoric deductive models. Consequently, the second-generation models of inflation of this lineage tended to assume—rather than explain—long-term processes of structural change and their distributional imbalances. The contributions made by the group of economists from PUC-Rio, as well as in the contributions by Bresser-Pereira and Nakano (Reference Bresser-Pereira and Nakano1984), from the FGV-SP school of economics, underrated basic causes of inflation in favor of the importance of the propagation mechanisms embodied in indexation practices.

Particularly regarding the PUC-Rio economists, the role played by distributional conflict rapidly turned into a “supporting character” in the plot, remembered only when it suited the argument being made. Once the notion of conflict was no longer useful, the inertial inflation story was stripped from its structuralist substance, and its scope was massively narrowed down to the spontaneous non-synchronization of price and wage readjustments. As a result, the inertial inflation hypothesis became more congenial to the New Keynesian formulations, found in Guillermo Calvo (Reference Calvo1978), Jo Anna Gray (Reference Gray1976), and Stanley Fischer (Reference Fischer1977).

Of course, the decline of the inertial inflation hypothesis does not rest solely on these analytical problems and discourse format issues; yet, one cannot deny their relevance. Despite differences in workingout the causes and propagating mechanisms of inflation, the policies required to deal with them were not in dispute; both sides were highly skeptical of the efficacy of orthodox anti-inflationary measures. Stabilization of prices should be achieved, alternatively, through an adequate management of wage policy coupled with a disciplining of markup rates. Conflict should be ameliorated somehow by way of a social pact. The question was: Would society adhere to it in a voluntary or compulsory fashion? This points our discussion to the policy implications of these theories. An inertial inflation required the attainment of distributional neutrality at the policy level, which proved disquietingly hard to resolve in both the analytical and the policy dimensions. The analytical requirements of distributive neutrality, as clearly outlined by Lopes (1986, pp. 50–62), not only proved exceedingly unrealistic, but also entailed impractical—and, to some extent, untenable—policy requirements for achieving distributive neutrality in the real world. The diversity of approaches marching under the banner of the “inertial inflation hypothesis” provided little help in sorting out these policy puzzles. The theory failed to meet the required external consistency, as reality proved resistant to the policies derived from it. The ensuing evasion of the small group of PUC-Rio economists from the inertialist camp concluded the destabilizing process of the theory it was supposed to deliver.

Future material will further substantiate the instability of the inertial inflation hypothesis by fleshing out tensions within the academic community of the economics profession in Brazil. As we have hinted above, the works scholars at PUC-Rio published along the 1980s gradually evolved from a seemingly distributional conflict approach—incarnated in a relative wages model of inflation—to a fiscally dominated understanding of inflation. Hence, two further sets of problems must be tackled. The first is the need to differentiate the academic group, situated in the economics department of PUC-Rio, from the other “heterodox” proposals in the aftermath of the Cruzado Plan’s debacle. The development of the inflationary crisis provided further refinement in the understanding of the dynamics of inflation in Brazil, which may have defined the magnitude of political constraints faced by any stabilization policy in the context of an accelerating inflationary process. The second aspect of this change in both focus and emphasis concerns the fiscal element, which receives increasing importance towards the end of the 1980s and becomes central by 1994. These two aspects can reveal how these scholars-turned-policy-makers’ works smoothed the transition from this regional understanding of inflationary inertia to the concept upheld by the New Neoclassical Synthesis, thus accounting for the decline of the inertial inflation hypothesis in Latin America.