Practical men, who believe themselves to be quite exempt from any intellectual influences, are usually slaves of some defunct economist.—John Maynard Keynes.
What explains the sustained commitment to budget discipline among many Latin American governments notwithstanding the general level of regional disillusionment with Washington Consensus reforms?1 For example, the Bolivian government is well known for its rebuff of Western governance models, including President Evo Morales’s nationalization of the natural gas industry. In the realm of national fiscal affairs, however, its long-standing economy minister Luis Alberto Arce often sounded like a poster child for the International Monetary Fund (IMF) when touting the importance of “recurring budget surpluses, a prudent fiscal attitude,” and “defeating inflation” to the country’s economic model.2 Moreover, fiscal rectitude has endured beyond the global commodity correction, with the Bolivian government maintaining an average primary budget surplus of almost 1 percent of GDP over the last decade. What accounts for the persistence of macroeconomic discipline?
In this article, I develop a new theory about the political legacy of economic crises. I argue that macroeconomic policy choices are a product of crisis histories. Latin American politicians operate according to the standard political logic that assumes voters respond to current economic conditions, but their incentives change when their countries have experienced devastating economic crises. The political impetus to protect voters from negative income shocks can be as strong as the political incentive to pad their earnings.
In a region like Latin America that is characterized by considerable economic volatility, politicians are keenly aware of the political turnover induced by negative income shocks. They have coincided with the ouster of about two-fifths of the sixteen Latin American presidents that were removed early from office since 1978 (see Hochstetler and Samuels 2011). Hyperinflation, in particular, has yielded some of the most severe shocks historically, rupturing the domestic price system, depressing popular living standards, and erasing as much as 30 to 90 percent of earning power in inflation-ridden countries.3
I contend that presidents seek to minimize the risks of such traumatic crises reemerging during their administration by invoking historical analogies about the lessons of hyperinflation. I thus expect politicians from inflation-crisis countries to pursue fiscal discipline, based on the mainstream economic lesson derived from such crises that high fiscal deficits are inflationary (Sargent and Wallace 1981). International financial institutions (Thacker 1999; Vreeland 2003) and global financial markets (Frieden 1991; Mahon 1996; McNamara 1998; Mosley 2000; Wibbels 2006; Kaplan 2013) have long advocated for budgetary rectitude. For example, fiscal discipline topped Williamson’s (1989) original list of Washington Consensus reforms.
However, a nation’s understanding of its own crisis history is an important determinant of whether or not it heeds such advice. Hyperinflation helps explains Latin American governments’ more steadfast commitment to budget austerity relative to other reforms. Austerity, or the commitment to such fiscal discipline even during hard times, helped many countries successfully tame hyperinflation.4 By comparison, alternative heterodox interventions, featuring the combination of expansionary fiscal policies along with wage and price controls, were largely discredited after failing to control runaway inflation.
In the years following hyperinflation, Latin American governments have experienced other varieties of domestic crises, including protracted periods of unemployment that have temporarily raised the political saliency of heterodox interventions. Ultimately, however, their policy resilience has been limited by the roots of these crises, which have not disrupted the price system and the economy as gravely as inflation crises.
How are such differences in economic ideas channeled and sustained through the political system? I contend that there are two primary pathways: politicians and technocrats. First, presidents may prioritize fiscal sustainability within their overall economic agenda. In crisis-scarred countries, the persistence of inflation sensitivity among the electorate and businesses raises the political appeal of mainstream policy solutions that use fiscal discipline as a conduit for inflation-controlled growth (Kaplan 2013).
Presidents may also appoint technocrats or ministers with specialized training in economics to provide such governance. In fact, there has been a fivefold increase in technocrats serving as key members of Latin American presidential teams since 1970.5 Such technocrats first emerged in the wake of the 1980s debt crisis, when politicians hoped such expertise would help assuage foreign investors’ concerns about economic turmoil undercutting their profitability (Schneider 1998).
Given their powerful position in presidential cabinets, technocrats are an important transmission mechanism for explaining policy choices, but they have received less attention compared to such other channels as elections and partisanship. While a burgeoning literature on ministerial turnover (Amorim Neto 2006; Martínez-Gallardo 2012, 2014; Camerlo and Pérez-Liñán 2015a, 2015b) examines how institutions and critical events shape near-term cabinet changes, this study explores the extent to which memories of a crisis can yield sustained ideological shifts in ministerial composition.
In light of their status as noncareer politicians (Alexiadou 2015), technocrats’ professional training should allow them to adequately diagnose economic problems (Dargent 2014, 2017). However, in line with the Keynesian wisdom above, they are not exempt from ideological or political influences (Grindle 1977; Camp 1985; Domínguez 2006). In inflation-ridden countries, crises have moved politicians and technocrats closer to the economic center, prioritizing inflation control through fiscal discipline.
In testing these theoretical expectations, I first examine the effect of inflation crises on ministers’ ideological orientation, before exploring how crises affect politicians’ fiscal policy choices both directly, and indirectly through ministerial appointments. I focus on fiscal governance because a government’s priorities are reflected in its national budget, just as a firm’s or household’s preferences are conveyed through its balance sheet. Finally, I examine these claims in Latin America, a region that is ideally suited for this analysis because of its considerable variation in inflation crisis legacies: about half of its countries have experienced a severe inflation crisis (Table A.7 in the online appendix).
During the first stage of this analysis, I also build on research that shows that policy-makers’ education is a proxy for their policy preferences (Chwieroth 2007; Nelson 2014a, 2014b; and Alexiadou 2015). In order to operationalize the policy orientation of key members of presidential economic teams, I employ a unique, novel dataset, dubbed the Index of Economic Advisors (Kaplan 2017). This index characterizes the policy preferences of economic advisors (mainstream vs. heterodox) over the last half century, based on their professional background and education credentials. To my knowledge, it is the first index of its kind to incorporate Latin American universities, which are also classified by ideological orientation through a series of in-country surveys of Latin America economists from 2015 to 2016.
Employing cross-national data from sixteen Latin American countries from 1961 to 2011, I find that hyperinflation histories are often associated with governments that have more mainstream economic advisors and greater fiscal discipline than do their noncrisis counterparts. These empirical results may help explain Latin America’s well-known pattern of procyclical fiscal spending (Gavin and Perotti 1997; Pinto 2013), where downturns tend to coincide with sustained periods of budget austerity, which can limit government’s ability to respond effectively to business cycle fluctuations (Blyth 2013).
These findings also mark a notable departure from the developed-country literature on macroeconomic partisanship. In contrast to traditional models of the economy that expect a partisan split on inflation-control policies that favor businesses (Hibbs 1977; Bartels 2008), these findings show that transformative national events like economic shocks can blur traditional partisan differences, contributing to the lack of ideology in party systems (Roberts and Wibbels 1999; Lupu 2014). That said, this macroeconomic consensus is distinct from micro-level dimensions, such as privatization or public investment, where scholars have found a closer link between traditional partisanship and regulatory outcomes (Murillo 2009; Boix 1998).
This investigation also offers new insights for studies examining globalization, neoliberalism, and the Latin American left, which have found considerable variation in the extent of government intervention in developing economies. On one side of these debates, scholars contend that economic integration (Rudra 2002), global markets (Mahon 1996; Mosley 2000; Wibbels, 2006; Kaplan and Thomsson 2017), and international institutions (Thacker 1999; Vreeland 2003; Winters 2010; Dietrich 2013) have contributed to a retrenchment of Keynesian-style countercyclical fiscal policies and social safety nets. In support of this view, scholars find that a variety of factors, including a weak labor movement (Roberts 2002), party-brand dilution (Lupu 2014), strong business interests (Thacker 2000; Schneider 2004; Fairfield 2010), centrist and noneconomic voters (Baker and Greene 2011; Hellwig 2014), and reform-seeking politicians (Corrales 2000) helped facilitate a broad-based acceptance of this neoliberal consensus (Stokes 2001; Murillo 2002; Levitsky 2003). Other scholars, however, find that neoliberal reforms have not been uniform. Rather, many countries with legacies of import substitution industrialization (ISI) have crafted political bargains (Frieden 1991) that preserved supply-side interventions, including industrial promotion, public employment (Kurtz and Brooks 2008), labor protection (Carnes 2014), and social insurance (Wibbels and Ahlquist 2011). In the realm of macroeconomic policy-making, I contend that this policy variation reflects the nature of a country’s crisis history, with budget discipline tending to be more common in inflation-scarred countries.
Finally, these findings have significant implications for the study of policy-making beyond Latin America. Diffusion scholars suggest that the prevalence of neoliberalism reflects the spread of economic orthodoxy through Western diplomacy, an Americanized global economics profession (Hall 1993; Babb 2001; Montecinos and Markoff 2010), and multilateral institutions (Barnett and Finnemore 2004; Woods 2006; Simmons, Dobbin, and Garrett 2008).
Recent scholarship has found that the IMF operates as a diffusion mechanism for the spread of neoliberal economists and hence neoliberal ideas, which in turn increases the likelihood of preferential treatment for IMF loans (Nelson 2014a, 2014b). Diffusion scholars correctly identify an important global pattern: policy choices often reflect the ideological persuasion of key advisors. But to what extent do such choices have domestic roots? Some ideas, like the IMF’s austerity, often appear on countries’ menu of policy options but are not systematically adopted. I seek to explain this policy variation, claiming that states’ independent histories are pivotal to understanding when technocrats pursue austerity.
In the world of Darwin, genes propagate in advantageous environments. Similarly, in the policy world, economic ideas prosper under favorable conditions. At the end of the twenty-first century, an era of relative developed country stability, known as the Great Moderation, opened the door to a mainstream governance consensus that emphasized economic moderation and budget discipline. Following the world’s struggles with inflation during the 1970s, many political leaders in developed countries centered their economic agendas around the neoclassical synthesis in contemporary macroeconomics. This governing consensus was skeptical of the merits of fiscal policy interventions because of the perceived link between budget deficits and higher inflation (Sargent and Wallace 1981). Rather, it recommended employing an inflation-fighting independent central bank as the principal economic policy instrument.
These governance principles, which are often linked to accessing global financing, filtered down to the domestic agendas of developing-country politicians. But why would developing countries be willing to accept such austere principles?
I contend that transformative national events play a pivotal role in explaining the adoption of economic paradigms.6 Governing in such an uncertain environment, politicians are concerned about avoiding economic turbulence on their watch, particularly those crises that have exacted the heaviest domestic toll. They fear that crises could devastate popular living standards and catalyze political turnover. Indeed, economic crises have contributed to the ouster of many developing-country leaders, including Argentina’s Raúl Alfonsín and Fernando de la Rúa, Brazil’s Fernando Collor de Mello, Bulgaria’s Zhan Videnov, Bolivia’s Hernán Siles Zuazo, and Ukraine’s Leonid Kravchuk. Politicians thus react to their economic environments, choosing advisors whose economic training promises to avoid repeating the costs of the most devastating crises.
Employing such historical lessons is not unique to fighting economic wars. International relations scholars find that major historical events often guide decision makers (Jervis 1976, Khong 1992, and Levy 1994). In foreign policy, for instance, scholars have observed that historical analogies are often key elements of military and foreign affairs strategies. For example, many US presidents, including Harry Truman and Ronald Reagan, formulated their post–World War II military interventions using the historical analogy of Munich appeasement (Khong 1992). In economic policy-making, the relevant analogies reflect historical states of the economy, with crises often being the most salient events.
In Latin America, the region’s most salient historical crisis is hyperinflation. These traumatic domestic experiences helped align national economic agendas with those of international institutions and markets. In fact, previous scholarship has shown that political leaders often use international tools, such as IMF conditionality, to help achieve their own domestic policy priorities (Vreeland 2003).
I contend that the region’s understanding of its inflation crisis history raised the domestic political appeal of hiring mainstream economists who promised to contain inflation through fiscal discipline. In the wake of these crises, the importance of price stability gained political traction over heterodox policies that had struggled to rein in runaway prices. Heterodox programs typically used price controls rather than budgetary rectitude to curb inflation, overlooking the aforementioned link between large fiscal deficits and inflation. The political desire to use expansionary fiscal policies to redistribute income and spur development was a laudable goal in countries that had struggled with high income inequality. However, these programs were unable to break the vicious cycle of ballooning deficits, booming money supply, and unbridled inflation that haunted so many Latin American governments. When governments used the printing press (or central bank financing) to fund deficit spending, the money supply expanded at breakneck speeds, causing prices to soar without bounds (Dornbusch and Edwards 1991). Surging inflation further eroded real tax revenues and exacerbated budget shortfalls, with governments often responding by printing more money. Latin American hyperinflations soared thousands of percentage points higher annually, rupturing economic activity and popular incomes in a way that has been unparalleled by any business cycle fluctuations since that time. By comparison, the relative success of mainstream policies in eventually taming runaway inflation boosted their technocratic saliency, particularly the importance of avoiding unsustainable deficits.
I argue that the historical lessons associated with such traumatic crises are enduring, creating sustained shifts in postcrisis technocratic composition and national approaches to budgetary management. I anticipate that such historical memories are channeled into economic policy-making through two primary channels.
The first pathway occurs directly through political elites, where “a sufficiently acute crisis creates a consensus that the old order has failed and needs to be replaced” (see Drazen and Easterly 2001, 134). More specifically, in countries where runaway inflation has eroded middle- and working-class incomes, political leaders are more likely to internalize the lessons of budgetary overexpansion. In such inflation-scarred countries, job creation remains a political priority, but not if it entails using aggressive fiscal stimulus that spurs inflation (Kaplan 2013). A repeat of past inflation crises would carry prohibitive costs. Not only do they create grave income shocks that torpedo popular living standards, but scholars have also found that they have electoral repercussions. High inflation erodes electoral support for incumbent governments and discredits political parties (Remmer 1991; Stokes 2001; Murillo, Oliveros, and Vaishnav 2011).
Politicians therefore use sound fiscal governance to keep inflation under control and convey their managerial credentials to households, businesses, and investors. In fact, inflation control has also had a steady baseline of support since the end of hyperinflation. In its survey of sixteen Latin American countries, the Latinobarómetro found that more than one-quarter of the adult population believed “fighting inflation” was the most important issue for their country between 1995 and 2008.7 More recently, the Latin American Public Opinion Project’s (LAPOP) Americas Barometer shows that Latin Americans rank inflation, high prices, and “economic crises” among the top eight most serious problems facing their countries out of a list of almost forty different issues between 2008 and 2014.8
In addition to this direct effect, I also expect crisis memories to have a more indirect effect on economic policy-making through ministerial appointments. A burgeoning literature on executive-branch politics has found that institutional attributes and critical events often shape ministerial turnover (Amorim Neto 2006; Martínez-Gallardo 2012, 2014; Camerlo and Pérez-Liñán 2015a, 2015b). Building on a comparative politics scholarship that finds cabinet changes, party structures, and coalition-building can be a source of stability in presidential systems (Mainwaring and Shugart 1997; Corrales 2002; Cheibub, Przeworski, and Saiegh 2004), this literature examines how cabinet reshufflings are often aimed at mitigating uncertainty in both democratic and nondemocratic regimes (Chwieroth and Walter 2010) following negative popularity shocks (Camerlo and Pérez-Liñán 2015a), political unrest (Camerlo and Pérez-Liñán 2015b), and economic crises (Martínez-Gallardo 2014).
By comparison, this article also examines the effect of economic crises on presidential cabinets, but over a longer term. Rather than examining how critical events influence near-term cabinet changes, it explores the role of historical memory on sustained ideological shifts in presidential economic teams. Moreover, compared to the focus on interbranch negotiations (Negretto 2006) in the minister-retention literature, I study a single policy dimension, macroeconomic policy-making.
In this dimension, the executive branch directly formulates and implements fiscal policy, serving as the command and control of the budgetary process (Hallerberg, Scartascini, and Stein 2009; Bonvecchi and Scartascini 2011). Political factors that foster greater ideological diversity in the executive branch, such as left partisanship or coalitional cabinets, can certainly influence technocratic appointments and governance. However, given the executive branch’s considerable policy discretion in mapping a country’s course of fiscal policy, such intrabranch politics is likely to be more salient than interbranch politics. While Congress has a direct role in the budget approval stage, it often cannot change the broad counters of macroeconomic policy. In other words, legislators can bargain over line-item spending, but new expenditures are either capped or “offset” by additional financing sources (Hallerberg, Scartascini, and Stein 2009).
Past shocks may thus condition not only technocratic appointments but also the government’s fiscal policy stance. To avoid repeating their crisis histories, presidents appoint technocrats or ministers with specialized training in economics. While such technocrats are often expected to apply nonpartisan, professionalized approaches to decision-making, their policy preferences typically reflect a country’s understanding of its economic history. In other words, politicians tend to fight past economic wars.
In the shadow of inflation crises, presidents are more likely to appoint mainstream “inflation hawks” who view fiscal discipline as the pathway to inflation control, rather than more “dovish” heterodox advisors, who are willing to use fiscal intervention to catalyze the economy and remedy job losses. These heterodox advisors are most likely to emerge after unemployment shocks at the other end of the business cycle. When economic activity slows too markedly, unemployment can trigger social mobilization and public protests against mainstream governance. For example, Argentina, Ecuador, and Venezuela have all experienced unemployment shocks that raised the political saliency of job creation relative to inflation control. Unemployment reached historic highs in all three countries in the early 2000s, helping open the door to less conventional advisors that governed with greater budgetary intervention in the economy.
However, these shocks tend to be far less historically salient than inflation crises because they did not disrupt the price system, and hence the economy, as severely. Their more limited saliency helps explain some recent policy reversals. In Argentina, for instance, the return of moderate inflation under President Cristina Kirchner quickly stoked fears about the perils of runaway inflation, swinging the pendulum back toward a centrist president Mauricio Macri and a cadre of mainstream technocrats who pledged to control inflation with austerity. When a country like Argentina has suffered a dual-crisis legacy, the political lessons from unemployment shocks are often less enduring than inflationary crises, which mitigates their effect on ministerial appointments. By contrast, even years after an inflation crisis, mainstream economists tend to believe that fiscal discipline has its merits. It not only provides businesses and investors with a stable, long-run operational environment but also protects voters’ incomes.
Ironically, delivering such price stability may unintentionally create other risks. When appointments are driven by historical legacies more than by contemporaneous problems, leaders may not respond optimally to new challenges, impeding their national competitiveness. Austerity has undoubtedly helped keep inflation under control in Latin America, but it may have also stifled growth, production, and employment (Blyth 2013; Stiglitz and Greenwald 2014). It pays the most dividends at the peak of the economic cycle, when it can help curb high-growth spending sprees that might otherwise catalyze inflation or default. Without the proper economic diagnosis, however, the austerity prescription can inadvertently impair economic health.
In summary, in a region like Latin America that has historically exhibited considerable price volatility, I anticipate that macroeconomic governance is often a product of inflation crisis legacies. To evaluate this theoretical expectation systematically, I employ the following testable hypotheses. The first two hypotheses examine the link between economic crises and ministerial appointments, while the third hypothesis assesses how crises affect fiscal governance strategies both directly, and indirectly through these ministerial appointments:
H1a: Inflation crisis legacies tend to shape the professional composition of presidential cabinets. Following severe inflation shocks, economic teams are more likely to be comprised of mainstream advisors compared to those countries that have not experienced extreme inflation crises.
H1b: Heterodox advisors are more likely to emerge after unemployment shocks at the other end of the business cycle. These shocks tend to be less salient historically than inflation crises, which attenuates their relative effect on ministerial appointments in countries with dual-crisis legacies.
H2: Inflation crises affect fiscal governance both directly, through greater inflation-aversion among political leaders, and indirectly through ministerial appointments. Mainstream advisors are more likely than heterodox advisors to pursue fiscal austerity by improving budget balances.
Data and Methods
I test these hypotheses in Latin America, a region where the degree of economic volatility has been two to three times higher than developed countries (Maddison 2001). The region’s volatility in large part has reflected its inflation crisis history; about half of its countries have experienced an inflation crisis, making it a fitting environment to examine how crises affect governance. I anticipate that these crisis legacies are enduring, given the severity of inflationary trauma relative to other types of domestic volatility (i.e., unemployment shocks) in Latin America.
I expect inflation crisis memory to affect economic policy through two different channels: a direct political effect where crisis legacies create a political incentive to pursue fiscal discipline by widening the constituency that favors inflation-controlled growth, and an indirect effect where presidential appointments of technocratic advisors lead to more restrained fiscal governance.
Employing a panel of data covering sixteen democratic countries from 1961 to 2011 that includes a total of 225 economic ministers, the first-stage probit model explores the factors driving ministerial appointments. I find that crisis legacies are an important domestic channel that is distinct from more traditional factors influencing the appointment process, such as regional diffusion, partisan economic ideas, and coalitional cabinets. However, other political factors may not be as observable, creating potential bias. For example, mainstream economists working for business groups, media conglomerates, technocratic networks, and think-tanks interpret economic conditions and frame business sentiment, which can then influence ministerial appointments (Luna and Kaltwasser 2014).
I thus attempt to account for such endogeneity in the appointment process by using a Heckman-type correction to control for treatment selection (Heckman 1988; Przeworski and Vreeland 2000; Vreeland 2003; Chwieroth 2007). I calculate the inverse Mills ratio from the selection equation to use as a switching value to control for the nonrandom selection of economic advisors in the outcome equation. I can thus test for the independent effects of economic advisors on fiscal policy choices in the second stage of the model.9 This approach allows us to examine the indirect effect of crisis legacies operating through ministerial appointments, in addition to direct linkages between crisis legacies and fiscal governance (see the online appendix).
Data Description: Independent and Dependent Variables
Policy Orientation of Economic Ministers
In order to test the first-stage probit model, I use a dichotomous variable, Mainstreamit, that measures whether or not economic ministers, who are appointed by political leaders, have mainstream economics training. The dichotomous measure is based on the Index of Economic Advisors, an original cross-country dataset, which to my knowledge is the first of its kind to provide detailed information regarding both the educational credentials and professional background of Latin American economic advisors. In additional robustness checks using an ordered probit model, I also employ the full index, Mainstream_IEAit (see Table A.1 in the online appendix), rather than the binary measure as the dependent variable, finding that the results do not materially change (see model 2 in Table 1).
|(1) Probit||(2) Probit||(3) Probit||(4) Probit||(5) Probit||(6)Probit|
|Unemp. Crisis Legacy||–0.860||**|
|Hyper. *Unemp. Crisis||1.286||**|
|Hyper. (20yr memory)||0.956||**|
|Hyper. (yr since crisis)||0.077||*|
|Terms of Trade||–0.126||–0.068||–0.218||0.037||–0.036||–0.492|
|Ext. Public Debt||–0.021||***||–0.020||***||–0.017||***||–0.022||***||–0.015||**||–0.007|
|Age of Democracy||0.012||0.104||–0.319||–0.113|
In constructing these measures, I draw on an extensive sociology and political science literature showing that professional economics training often shapes policy preferences through socialization and diffusion. In other words, economists often diagnose problems and offer policy solutions through the “interpretive lens” of their professional training (Hall 1993; Dominguez et al. 1997; Babb 2001; Chwieroth 2007; Kogut and MacPherson 2011; Montecinos and Markoff 2010). Building on these findings, scholars have employed rich datasets on US economics training as a proxy for neoliberalism (Chwieroth 2007; Kogut and MacPherson 2008; Nelson 2014a, 2014b), based on the premise that neoliberal ideas diffuse from an Americanized global economics profession. Hallerberg and Wehner (2013) also employ similar indices to evaluate if OECD governments are more likely to appoint technocrats during financial crises; Alexiadou (2015) finds that left-leaning ministers are more likely to increase social welfare programs; while Flores, Lloyd, and Nooruddin (2016) gauge the effect of technocratic leadership on sovereign credit ratings.
Compared to these global indices, which often assume homogeneity, the Index of Economic Advisors allows for greater contextualization of educational backgrounds across a wider ideological spectrum. I begin with a similar premise, coding as mainstream those advisors that have trained at highly ranked economics departments outside of Latin America. However, I also code several Latin American universities, such as Pontificia Universidad Católica de Chile, Universidad Torcuato Di Tella in Argentina, and the Fundação Getúlio Vargas in Brazil, as mainstream economics departments (see Table A.6 for a full listing) because these universities embody similar approaches to those that are typically considered neoliberal in the United States.
Furthermore, to account for any economic departments that may diverge from mainstream economics both within and beyond Latin American borders, the index removes any universities whose economics departments are members of the Association for Heterodox Economics (AHE)’s International Directory for Heterodox Economists. I further ensure the robustness of the coding by corroborating this directory with an online survey conducted from 2015 to 2016 asking local scholars in sixteen Latin American countries to score their major national universities on a scale ranging from heterodox to orthodox.
Aggregating this information, I code professional educational training of finance ministers and central bank presidents as a binary variable according to the rule below. This coding rule yields an average of eighteen economic advisors per country, whose tenure averages almost three years.
In additional robustness checks, I expand the purview of the measure beyond this formal educational filter to include the professional background of key advisors, Mainstream_pit (see model 3 in Table 1). Given that preferences may change over time, these tests gauge the importance of work experiences and professional networks in domestic policy formation. This coding assumes that those advisors hailing from international financial institutions (e.g., the IMF or World Bank), global finance, or business tend to hold liberal economic beliefs that align with mainstream thinking (section A.1.1, Table A.1 in the online appendix).
Inflation Crisis Legacy
I measure the duration of the inflation crisis memory in several different ways. In the main text, I employ Drazen and Easterly’s (2001) definition of a hyperinflation crisis, specified as an observation two or more standard deviations above the mean. Given the expectation that past shocks condition current policy-making, the binary variable (Hyperinflation legacy) classifies a country as having such a legacy after inflation returns to the historical norm (i.e., within one standard deviation of the mean).
In additional robustness tests, I also employ a more encompassing measure of inflation crises (Inflation crisis legacy) to account for the region’s variation in inflation crisis experiences that include both hyperinflation and “very” high inflation crises (Tables A.3 and A.7 in the online appendix). Additionally, I develop a binary measure of crisis memory for unemployment shocks to observe if the appointment of heterodox advisors is more likely at the other end of the business cycle. Next, I create an interaction term for countries that have experienced both shocks to see if inflation crises are historically more salient, as expected. I also use a series of alternative crisis measures in further robustness checks, including Highest past inflation, Lagged inflation (t-20 and t-30 yrs), and Lagged unemployment (t-5). See section A.1.1 and Table A.3 in the online appendix.
Returning to the hyperinflation variable, I conduct further robustness tests to observe if the political saliency of past inflation crises fades with time, shortening the window of crisis memory to five, ten, and twenty years. Finally, I also create a measure that accounts for the total years since the end hyperinflation.
I control for a variety of global economic factors, domestic economic variables, and institutional factors that may affect ministerial appointments and national fiscal balances. I also use a slightly different set of controls for the ministerial appointments and fiscal policy regressions, as I expect different factors to be important for different outcomes. In the two-stage model, it is also generally recommended to include at least one extra explanatory variable that influences selection but not the subsequent outcome.
I incorporate several key political variables from the minister-retention literature, including coalitional cabinets, legislative minorities, and nondemocratic regimes (see Chwieroth and Walter 2010; Martínez-Gallardo 2012, 2014; Camerlo and Pérez-Liñán 2015a, 2015b), to help account for potential drivers of cabinet changes that operate independently of crisis memories. Additionally, given that the crux of the analysis focuses on the links between crisis memories, ministerial appointments, and fiscal governance, I also employ the standard control variables (including a lagged dependent variable) for fiscal policy regressions used in the political budget cycle literature (Brender and Drazen 2005; Keefer 2005; Barberia and Avelino 2011). See section A.1.2 in the online appendix.
To operationalize the hypotheses, I use the following specifications:
where Mainstreamit is equal to mainstream ministerial orientation; where Fiscalit = primary fiscal balance (as a percentage of GDP); where Hyperit = hyperinflation legacy; Unempit = unemployment crisis legacy. The index i = country and t = year. Xit = vector of control variables; Fiscali,t-1 = primary fiscal balance (one-year lag); λi = inverse Mills ratio derived from first equation of Heckman-type correction model, and used in the second equation to control for treatment selection; ηi = dummy capturing unobserved country effects; ɛit = error term. For more details, see section A.1.3–4 of the online appendix.
When Mainstreamit is the dependent variable, a positive coefficient for Hyperit would support the first hypothesis (H1) that countries with a hyperinflation legacy tend to have a higher share of mainstream economist in presidential cabinets, while a negative coefficient for Unempit expects that countries with an unemployment crisis legacy are more likely to appoint heterodox advisors (H2). In line with the second hypothesis, a positive coefficient on the interaction term, Hyperit *Unempit (representing countries with dual-crisis legacies) would also show that inflation crises tend to have a more pronounced effect on ministerial appointments than those associated with unemployment crises. Finally, when Fiscalit is the dependent variable, I expect to observe a direct effect where Hyperit yields greater budgetary restraint, and an indirect effect where independent of hyperinflation’s initial effect on ministerial appointments, Mainstreamit also yields greater budget discipline (H3).
The Effect of Past Inflation Crises on Technocratic Orientation
Do crisis legacies condition ministerial appointments? The first series of probit models display the effects of the independent variables on the professional training of Latin American economic ministers. The coefficient on hyperinflation legacies is positive and statistically significant (at the 99 percent confidence interval) across the regression models (see models 1–4 in Table 1 below). Employing these coefficients to derive the predicted probability of having a mainstream economics minister,10 I find that an inflation crisis history makes a government about 36 to 47 percent more likely to appoint advisors with mainstream economics credentials compared to noncrisis periods. These results lend considerable support to the first hypothesis (H1) that runaway inflation breeds fiscal conservatives schooled in mainstream economics. Moreover, these findings do not materially change when conducting an ordered probit regression with the ordinal variable (see model 2 in Table 1) instead of the probit models with the dichotomous measure.
I also expand the definition of a mainstream economist to include professional background (i.e., previous career experience), Mainstreamit, given the theoretical expectation that experience working for the private sector, global financial markets, or international institutions is more likely to align an advisor with a mainstream policy orientation. Employing such a more comprehensive measure (see Table A.1) did not yield any material changes in the direction or statistical significance of the coefficients. However, they do become greater in magnitude (see model 3 in Table 1), suggesting that transnational networks in global finance may also influence the extent to which policy makers adhere to the mainstream consensus.
Does the opposite shock, an unemployment crisis, temper hyperinflation lessons? Or do inflation crises tend to have more historical saliency than other types of crises, as expected? In the absence of an inflation crisis legacy, unemployment shocks do appear to have some saliency. The coefficient for unemployment crisis legacy is negative and statistically significant (model 4 in Table 1), suggesting that presidents are more likely to deviate from the mainstream consensus following such shocks. They tend to appoint heterodox advisors that believe heavy deficit spending can alleviate demand shortfalls and generate employment.
That said, the conditional regression model also shows that a dual-crisis legacy has a positive and statistically significant effect on the appointment of mainstream economists that value fiscal discipline and inflation control (see model 4 in Table 1). These findings suggest that historical lessons from inflation crises have a more pronounced effect on ministerial appointments than those associated with unemployment crises, lending support to the second hypothesis (H2).
These results hold when controlling for several institutional variables, including whether a country has an IMF program, a long democratic tenure, or a left-wing government. In models 5–6 in Table 1, we also add two more control variables to measure institutional constraints on executive power. Minority reflects whether or not the president has a legislative minority, while Coalition measures whether the cabinet includes members of nonpresidential parties. The negative and statistically significant coefficients for left partisanship and coalitional cabinets illustrate that greater ideological dispersion (through both left governments and coalition governments) tends to breed more heterodox advisors.
While the main findings show that historical memory is associated with sustained ideological shifts in ministerial composition, domestic political conditions also seem to play an important role. In the realm of macroeconomic policy, however, intrabranch politics appears to be more salient than interbranch politics. The minority coefficient is statistically insignificant, meaning that we cannot rule out the null hypothesis that legislative minorities have little influence over the appointments of finance ministers and central bank presidents. Given the considerable discretion that presidential economic teams have in formulating and implementing fiscal policy, perhaps legislative minorities are more likely to influence line-item spending debates rather than the overall budgetary framework.
In a series of robustness checks, I find that the correlation between past inflation crises and policy orientation is markedly resilient. First, I account for the possibility that the saliency of past crises fade over time. I modify the structure of the binary variable for inflation crisis legacy to track a shorter twenty-year window (see model 5 in Table 1). These robustness tests do not yield any material changes to the core findings, with the coefficient for inflation crisis legacies maintaining its precision for not only the twenty-year but also the five- and ten-year windows.11 Given that these windows begin once inflation returns to its historical norm (i.e., one standard deviation above the mean), it is unlikely that policy makers are simply reacting to a resurgence of crisis conditions. Rather, crisis memory appears to be an important factor.12 In a final robustness check, I find that the coefficient on years since crisis is also statistically significant, suggesting that crisis legacies appear to have an enduring effect on ministerial appointments (see model 6).
Notably, in the online appendix, I also conduct a series of tests using alternative crisis measures to allay potential concerns about the Hyperinflation legacy variable’s rigidity. None of these measures, including a more encompassing measure of Inflation crisis legacy, Highest past inflation, Lagged inflation (t-20 and t-30 yrs), and Lagged unemployment (t-5), change the primary results (section A.2.1; Table A.4).
The primary findings are also robust when controlling for shorter-term inflation and unemployment issues (models 1–6 in Table 1). This result supports the notion that crisis legacies are often enduring, even when accounting for business cycle fluctuations. If politicians from crisis-ridden countries are more likely to appoint advisors with inflation-fighting credentials, it also suggests that they may not be heeding sufficient attention to contemporaneous economic conditions.
For this reason, I also add a lagged dependent variable to control for the potential persistence of ideological minister types over time (see model 6 in Table 1). The coefficient for the lagged value of mainstream economists is positive and statistically significant, suggesting that economic ideology can be sticky. However, adding the lagged dependent variable to the specification does not materially change the main results.
I also control for the role of regional diffusion in explaining national ministerial appointments (model 4–6 in Table 1). The coefficient for regional diffusion is positive and statistically significant, providing support for regional proliferation of mainstream technocrats expected by the literature on diffusion. This trend may in part reflect the tendency for democratic governments to appoint technocrats to improve their access to financing (Beaulieu, Cox, and Saiegh 2012; Flores, Lloyd, and Nooruddin 2016).
While presidents may in part choose their economic officials based on regional trends, this pattern does not temper the domestic link between crisis legacies and ministerial appointments. The primary findings remain robust, lending support to the notion that there is a domestic channel for ideational change that is independent of the global dissemination of ideas. Moreover, we cannot rule out the possibility that individual crisis legacies are reinforcing the regional trend toward mainstream technocrats who value inflation control.
The Effect of Past Inflation Crises on Fiscal Policy Choices
The first stage of the selection model above (see Table 1) shows that inflationary crises provide a window of ideological opportunity, often conditioning the type of ministerial appointments. The model’s second stage (see Table 2 below) then shows that these economic officials, independently of the initial process leading to their appointment, tend to be more fiscally conservative than their counterparts without formal training in mainstream economics. In models 1–2 in Table 2, for instance, the coefficient on mainstream economists is positive and statistically significant, with average budget balances that are about 1 percentage point of GDP higher than their less-conventional peers. These findings hold when controlling for institutional variables, including the age of democracy, left partisanship, and IMF programs.
|(1) FE||(2) GMM||(3) FE||(4) GMM||(5) FE||(6) GMM|
|Terms of Trade||0.546||0.452||0.589||0.540||0.510||0.374|
|Ext. Public Debt (t-1)||0.002||0.001||–0.001||–0.001||0.001||0.000|
|Output Gap (t-1)||0.039||0.037||0.009||–0.001||0.029||0.022|
|Fiscal Balance (t-1)||0.545||***||0.461||***||0.598||***||0.563||***||0.553||***||0.473||***|
|Regional Fisc. Bal. (avg)||0.314||0.736||**||0.289||0.519||**||0.301||0.722||**|
|Age of Democracy||–0.659||**||–0.717||***||–0.668||**||–0.683||**||–0.712||**||–0.811||***|
Beyond this indirect effect operating through ministerial appointments, might there also be a direct effect of crises on fiscal governance? The coefficients on inflation crisis legacy are positive and statistically significant. This pattern is in line with the historic tendency of procyclical fiscal spending in Latin America (Gavin and Perotti 1997; Pinto 2013), where economic downturns tend to coincide with sustained periods of higher budgetary constraints. I find that the political commitment to fiscal discipline is prolonged by crisis memory, specifically the severity of past inflationary shocks. Average budget balances following inflation crises tend to be about 2–3 percentage points of GDP higher than during noncrisis periods (see Table 2).
For example, Brazil’s 2011 primary fiscal surplus was more than 3 percentage points of GDP greater than the negative primary budget balance recorded that year in Venezuela, a country that had never experienced an extreme inflation shock. However, the regression estimates above imply that after Venezuela’s annual inflation rate surpassed 500 percent in 2016 (or more than 2 standard deviations above the historical mean), such a severe crisis could eventually trigger a period of postcrisis fiscal consolidation in which budget balances are as much as 3 percentage points of GDP higher than their pre-shock levels.
Finally, the above pattern of crisis-induced austerity holds, even when controlling for contemporaneous economic conditions, including the lingering structural effects of postcrisis inflation, the business cycle, and unemployment. Results for other control variables are also consistent with expectations. The positive and statistically significant coefficient for global growth suggest that improved fiscal balances are often correlated with better global economic conditions. The coefficient for the lagged dependent variable, primary fiscal balances (t-1), also is positive and statistically significant, implying that a history of prudent fiscal governance makes budget discipline more likely today. Notably, I also find support for the political budget cycle literature (Brender and Drazen 2005; Barberia and Avelino 2011), with the negative coefficient for age of democracies suggesting that, all else equal, new democracies are more likely to have larger fiscal deficits.
A series of further tests show that the correlation between inflation crises and fiscal governance is robust. First, I use the more comprehensive measure of mainstream economist (i.e., educational background or previous work experience) to account for the informal training that advisors likely receive when working for international financial institutions, global finance, or international business. This robustness check did not yield any substantial changes in the direction of statistical significance of the coefficients for mainstream economists, or inflation crisis legacies (models 3 and 4 in Table 2). Additionally, employing the full Index of Economic Advisors, rather than the binary measure, as the independent variable does not materially change the primary results (models 5 and 6 in Table 2). Finally, these results remain robust after a series of tests using the Arellano-Bond GMM estimator (models 2, 4, and 6 in Table 2), which do not considerably alter their size, direction, or statistical significance. In short, these robustness tests provide considerable support for the theoretical framework, finding that crises affect fiscal governance both directly, and indirectly through the ideational beliefs of technocratic communities (see section A.2.2 of the online appendix).
To examine further the extent of influence that crisis legacies can have on ministerial appointments and governance, I extend the analysis by briefly discussing three country cases: Argentina, Ecuador, and Peru. These three countries are similar along economic and political indicators: they are presidential, high-middle-income South American countries, yet they maximize the variation in the main independent variable of interest—inflation crisis legacies. Peru suffered through hyperinflation in the early 1990s, Argentina lived through shocks at both ends of the business cycle (hyperinflation and an unemployment crisis) during the last three decades, while Ecuador has never experienced hyperinflation.
The Peruvian case perhaps best illustrates the saliency of crisis memories on policy making, given that a single political leader, Alan García, served presidential terms before and after the country’s inflation crisis. In 2006, García returned to the presidency sixteen years after being ousted for a hyperinflation episode that eroded wages and deepened poverty. During his second presidential life, he swapped his first-term interventionist policies—once deemed reckless by the IMF—for fiscal discipline and inflation-targeting that were praised by the same institution.13 To achieve these goals, he appointed Luis Carranza Ugarte, a former banker and one of Peru’s most orthodox economists, to the prominent post of minister of Economy and Finance. Notably, García’s successor and leftist political rival, Ollanta Humala, also used technocratic continuity to signal his commitment to low-inflation policies, retaining key García advisors with economic doctorates from Johns Hopkins and Brown University.
By comparison, Argentina has experienced considerably more within-country variation in its crisis history. Similarly to Peru, it incurred a hyperinflation shock that catalyzed macroeconomic reform. Liberal economists brandishing inflation-fighting credentials were central to the Argentina’s Convertibility Program, an economic model based on austerity that aimed to contain the runaway inflation of the 1980s. After being the cornerstone of its economic policy in the 1990s, a severe unemployment shock in the early 2000s help spark a policy reversal away from inflation-checking neoliberalism. For example, President Cristina Kirchner turned to heterodox advisors such as University of Buenos Aires–trained economist Axel Kicillof, who promised to use heavy government intervention to minimize the new political albatross of unemployment.
To what extent, however, do such unemployment crises dilute the tenets of inflation-controlled discipline? Similar to the statistical findings showing greater resiliency of inflation crisis memories (see model 4 in Table 1 and models 5–8 in Table A.4), the influence of unemployment shocks appears to fade with time relative to inflation crises. Under the Cristina Kirchner administration, a reemergence of inflation pressures during her second term triggered the public’s inflation sensitivity and punctured her popularity, eventually contributing to the election of centrist president Mauricio Macri and his technocratic team. While Argentina’s unemployment shock had temporarily raised the saliency of job creation in the mid-2000s, hyperinflation casts a longer policy-making shadow because of its devastation to both the price system and the overall economy. Even Cristina Kirchner had appointed a mainstream economic minister, Martín Lousteau, early in her term amid ongoing concerns about devaluation-induced inflation.
What happens to governance in countries without inflationary scars? For example, Ecuador’s inflation has not breached 100 percent per annum over the last half century. Without historical lessons to constrain deficit spending, policy is conditional on commodity booms and busts. During booms, Ecuador’s negative correlation coefficient (0.40) for terms of trade and mainstream advisors suggests that advisors tend to be heterodox, which is in line with the regional trend (see Table A.4). They also tend to oversee wider fiscal deficits. For example, during its boom years, the Rafael Correa administration posted an average primary fiscal deficit of almost 3 percent between 2007 and 2014. Later in the Correa administration, however, the oil market collapse prompted Correa to appoint as finance minister an MBA-toting Fausto Herrera, who slashed spending and hiked taxes. It took a commodity shock to propel Ecuador toward austerity, first under Correa and now under his successor Lenín Moreno. Without the sustained saliency of hyperinflation lessons, however, it is unlikely that the present austerity push would persist during an oil-price recovery.
The effect of past crises on policy-making communities is impressive. Employing an originally constructed data index, dubbed the Index of Economic Advisors, cross-national statistical tests in sixteen Latin American countries from 1960 to 2011 show that fiscal governance is conditioned by inflation shocks through both a direct and indirect effect. Presidents from countries with inflation-crisis legacies are more likely to prioritize fiscal sustainability within their overall agenda. They also appoint considerably more mainstream technocrats to presidential teams (see Figure 1), who then tend to govern with greater discipline than their noncrisis peers. These severe crises often cast a long shadow over policy-making, with budgetary restraint enduring even after the business cycle shifts toward slower growth and higher unemployment.
With such a shared professional training, it’s not uncommon for public discourse to center on fiscal prudence and inflation control, even in a country like Brazil where the Rousseff administration experimented with micro-level heterodoxy (i.e., using off-balance sheet financing from the Brazilian Development Bank to fund industrial expansions). Recent finance ministers, including Henrique Meirelles, Nelson Barbosa, and Joaquim Levy, have still emphasized long-term fiscal adjustment,14 fiscal stability,15 and inflation vigilance.16 In line with these policy goals, and despite some recent fiscal drift, the Brazilian government has averaged a primary budget surplus of about 1.5 percent of GDP over the last two decades.
However, such a sustained austerity focus suggests that Brazilian technocrats may be too influenced by their inflation crisis history. Given the extent of Brazil’s most recent recession, some economists argued that fiscal stimulus, not restraint, would have helped spark an earlier recovery. For example, according to Keynes, “the boom, not the slump, is the right time for austerity” (Blyth 2013). During the 2008 financial crisis, even the United States, the global champion of austerity, stimulated its own economy to exit the recession.
Similarly to Brazil, neighboring Argentina had also placed disproportionate weight on its past crises, relative to contemporary economic conditions through much of the 1990s and early 2000s. Its political leaders unleashed a team of technocrats, brandishing mainstream economic credentials from the University of Chicago to Harvard University, to wage its war on hyperinflation. They maintained a stringent commitment to fiscal discipline under the country’s convertibility law for more than a decade after the crisis had subsided. The enduring focus on inflation control neglected the country’s growing lack of competitiveness and helped sow the seeds for the 2001–2002 debt crisis. Indeed, the drawback of governing through a historic lens is that a lingering political focus on past crises can limit policy flexibility today.
In conclusion, the findings offer important new insights for the political economy literature, demonstrating the key role that both transformative national events and key economic advisors often have in shaping policy choices. This article also offers a new and innovative dataset that measures the policy orientation of Latin America’s key economic advisors, which can benefit many different types of future research endeavors that examine the effect of ideational factors on such national policy choices as privatization and the funding of social spending, military expenditures, and development. Finally, from a policy perspective, these findings also point to the potential risks of fighting past economic wars, which can leave governments captives of history and misguided ideological paradigms and thus unable to respond to new economic challenges.