Abstract
The global financial crisis and ensuing weak growth have increased interest in macroeconomic issues within comparative political economy (CPE). CPE, particularly the dominant Varieties of Capitalism approach, has based its analyses on mainstream economics, which limits analysis of the relation between distribution and growth and neglects the role finance plays in modern economies. It overstates the stability of the capitalist growth process and understates the potential effectiveness of government interventions. Baccaro and Pontusson have suggested a post-Keynesian (PK) theory of distribution and growth as an alternative. This article generalizes their point. PK theory highlights the instability of the growth process and lends itself to an analysis of income distribution and power relations. The article identifies the analysis of financialization and financial cycles, the understanding of neoliberal growth models, and the political economy of central banks as areas where PK economics provides specific insights for CPE. It also highlights that these arguments have important implications for government policy in an era of secular stagnation with ongoing social, distributional, and economic crises.
Political economy approaches assert that economic, social, and political factors have to be analyzed in conjunction. Comparative political economy (CPE) is the field that studies differences in institutions, policies, and economic outcomes across countries. It seeks to determine why some countries have higher incomes or economic growth than others, why there are different degrees of inequality, and how these relate to differences in the institutions structuring industrial relations, financial systems, and welfare regimes. CPE therefore needs a theory of institutions and politics as well as a theory of how the economy works. One key question that divides economic theories is whether growth should be understood as driven mostly by (slowly changing) supply-side factors, such as the skills of the workforce and the speed of technological progress, or by the (more volatile) demand side, that is, spending decisions of firms, households, and governments. The answer to this question has far-reaching implications for economic analysis and, more important, shapes the interpretation of economic crises. Are they due to exogenous, unforeseeable shocks that bring about temporary deviations from an otherwise stable growth path (as implied by most supply-side theories)? Or are they the endogenous outcome of systemic forces that lead to boom-bust cycles, as non-mainstream versions of demand-side analyses suggest?
These questions matter to CPE because they are necessary to understand the economic performances of countries and to evaluate economic policies, but CPE rarely confronts them head on. Herman Mark Schwartz and Bent Tranøy argue that over the past few decades there has been a slow shift in CPE from macroeconomic approaches that emphasize economic instability and issues of political legitimacy to neoinstitutional approaches, in particular the Varieties of Capitalism (VoC) approach, that presuppose (stable) market outcomes which allow for multiple institutional equilibria. 1 This has moved CPE closer to mainstream economics with its supply-side focus and an interpretation of market economies as inherently stable. The global financial crisis and the ensuing weak growth have reignited interest in macroeconomic issues of growth, distribution, and stability and thus the question of the economic underpinning for CPE. Lucio Baccaro and Jonas Pontusson propose basing CPE on the post-Keynesian (PK) theory of demand regimes and use the cases of Germany, Sweden, and the United Kingdom to analyze export-led and debt-led growth models. 2 In a reply, David Hope and David Soskice argue that the more mainstream New Keynesian (NK) theory, which is based on methodological individualism, features a supply-side-determined long-run equilibrium, and regards financial crises as caused by exogenous shocks, is a more appropriate foundation. 3
This article is a reply to this controversy and makes a systematic case for post-Keynesian economics (PKE) as the macroeconomic foundation for the comparative analysis of capitalisms. It argues that CPE lacks adequate macroeconomic foundations; it needs an analytical framework that allows an analysis of the potential instability of growth in a financialized economy and the power relations that underpin inequality as well as financial relations. PKE, in contrast to NKE, offers a (Kaleckian) theory of demand regimes that allows for wage-led as well as profit-led demand regimes, partially used by Baccaro and Pontusson. Importantly, the PK theory of money and finance enables an analysis of financialization that considers the dysfunctional aspects of finance and the emergence of (Minskyan) financial instability. It has a focus on the demand side of growth but considers growth as path-dependent, with (Kaldorian) technological progress induced by demand pressures. Together this forms a basis for an analysis of growth models that is more appropriate than mainstream economics for a world characterized by distributional conflict and financial crises.
The article is at the same time highly sympathetic to and critical of Baccaro and Pontusson. I argue that PKE has more to offer than Baccaro and Pontusson realize, in particular regarding finance and financial instability. The article discusses several specific areas where a PK economics approach can make contributions to CPE debates. First, the PK analysis of endogenous financial instability has implications for our understanding of financialization. 4 Second, I argue that, contrary to what Baccaro and Pontusson assert, neoliberal growth models are premised on wage-led demand regimes and that the stagnation tendencies they encounter in the face of rising inequality are compensated for by debt-driven or export-driven stimulation, both of which give rise to unstable regimes. Third, endogenous financial instability has implications for the political economy of central banks. They act as lender of last resort for private financial institutions as well as governments, which gives them a distinct form of financial power. The overall PK vision of capitalist dynamics is one of an intrinsically unstable growth process, where class relations, financial instability, and government activity shape the growth path of economies.
The article is structured as follows. The first section below situates PKE and CPE within the historical development of the political economy approach. Next, recent debates on the role of macroeconomics in CPE are discussed. The sections following that discussion present NK theory and the three-equation model advocated by Hope and Soskice; set out the core features of the PK analysis of distribution, finance, and path-dependent growth; and highlight contributions of PKE to CPE on financialization and financial cycles, for the interpretation of neoliberal growth models and for the political economy of central banking. The last section concludes.
From Political Economy to Heterodox Economics and CPE
The main objective of this article is to clarify the contributions PKE can make to CPE, but many of the arguments here will also be relevant to other areas of political economy. Political economy started as a holistic approach in the nineteenth century that aimed at a unified analysis of social, economic, and political issues. Toward the late nineteenth and early twentieth centuries, under the influence of neoclassical theory, economics turned into “pure economics” and a disciplinary bifurcation occurred, with economics becoming a discipline distinct from political science and sociology. 5 The political economy agenda, which runs across academic disciplines, thus also split along disciplinary lines. Within economics, the PK and Marxist traditions pursued a political economy approach. Economics experienced a broadening of its theories with the Keynesian revolution of the 1930s, in particular the emergence of the separate field of macroeconomics, and then a narrowing with the monetarist counterrevolution of the 1970s, which delegitimized non-mainstream approaches and macroeconomic theory. To survive in a hostile academic environment, modern non-mainstream economics has increasingly relied on formal modeling, which makes communication with political economists in the social sciences more difficult. This is true for PKE but also applies to other non-mainstream approaches, which (since the 1970s) include evolutionary, institutionalist, and feminist economics. It is symptomatic that these approaches typically use the term “heterodox economics” rather than “political economy” as a unifying label.
Within the social sciences, the fields of international political economy (IPE), comparative political economy, and economic sociology formed in part to overcome the disciplinary divide. Among these, IPE and economic sociology have an eclectic understanding of modern non-mainstream economics; CPE, in the form of VoC, has gone furthest in drawing on mainstream economics. But the gap between heterodox economics and CPE is also reflected in Ben Clift’s excellent book Comparative Political Economy. 6 It has an insightful discussion of nineteenth-century economic theory and debates but ends its coverage with the controversy between Keynes and Hayek in the 1930s and ignores contemporary heterodox economics. This article argues that political economy research in economics as well as in the social sciences more generally was impoverished by the disciplinary split. Modern CPE, in particular the VoC approach, builds on contemporary mainstream economics and incorporates its shortcomings. Contemporary heterodox macroeconomics and specifically PKE offer a valuable alternative that can enrich CPE.
It is noteworthy that debate on the macroeconomic foundations of IPE has also begun. Mark Blyth and Matthias Matthijs argue that IPE has been unable to anticipate and deal analytically with the global financial crisis and its economic and political fallout. 7 The reason is that although (American) IPE has developed its microfoundations by leaning on modern mainstream economics, it is lacking in its treatment of the macroeconomy. Blyth and Matthijs propose the term “macroeconomic regime,” which they define with regard to the “the main target variable for a country’s macroeconomic policy.” 8 Effectively, they distinguish between Keynesian (full employment) and neoliberal (price stability) policy goals but offer limited discussion of different macroeconomic theories.
The Macroeconomics of Current Comparative Political Economy
At the inception of CPE, Andrew Shonfield had a focus on demand formation, economic stability, and political legitimacy; similarly, Peter Hall’s early work was concerned with demand management. 9 However, in subsequent debates CPE moved gradually in the direction of microeconomic questions and a supply-side focus. Schwartz and Tranøy trace this gradual shift and argue that it has resulted in a neglect of fallacy-of-composition problems and financial instability. 10 The culmination of this development is the VoC approach, which regards as the core reference point for the establishment of a viable variety of capitalism its ability to generate competitiveness. 11 VoC takes a firm-centric view. It analyzes firms as a set of relations: relations between firms and their employees, embodied in industrial relations and training systems; relations between a firm and its owners and stakeholders (corporate governance); and relations between a firm and its financiers and competing firms. All these are shaped in part by national regulations and policies that constitute the constraints that firms face. There exist complementarities between different sets of institutions, which led VoC to distinguish liberal, coordinated, and mixed market economies. The subtitle of Hall and Soskice’s influential book encapsulates the approach: the institutional foundations of comparative advantage. VoC offers an institutionally nuanced supply-side analysis of economic performance, centered on the concept of competitiveness that highlights the possibility of multiple institutional equilibria. Traditional concerns of Keynesian macroeconomics such as unemployment caused by demand deficiencies or crises and financial instability are not at the core of its research agenda.
In an attempt to reconstitute CPE, Baccaro and Pontusson suggest an approach that builds on the PK macroeconomic theory of demand regimes, and they develop this into an analysis of growth models that highlights the interaction between distributional changes, demand formation, and export performance. They identify Germany as an export-led growth model and the United Kingdom as consumption-led, with Sweden as an intermediate case and Italy as a case of stagnation. 12 Their growth models are “more numerous and more unstable” than those in traditional VoC analyses. 13 They conceive the growth models as underpinned by social coalitions and a hegemonic social bloc, based on sectoral interests, and they illustrate the notion with reference to coalitions around export interests. This conception is an important step away from static VoC classifications that tries to grapple with instability, but there is clear asymmetry in the depth with which export orientation is covered and an absence of an analysis of financialization and thus of the debt-led growth model.
In a reply, Hope and Soskice welcome the discussion of demand issues but argue that the modern New Keynesian mainstream economics is a more appropriate foundation. They specifically recommend the three-equation model, a textbook version of NKE. 14 This model is anchored in a supply-side-determined long-run equilibrium with monetary policy providing a stabilizing function. (The three-equation model will be discussed in more detail in the next section.) As much of the VoC literature does not discuss its macroeconomic underpinning, Hope and Soskice’s work is welcome in that it makes explicit the implicit macroeconomic assumptions of the VoC approach. There is a basic complementarity between NKE and VoC in that both share a supply-side focus, despite different research agendas. Demand formation plays a secondary (short-run) role; in a longer perspective only institutions and other supply-side factors such as technology matter. Issues like financial stability or the demand effects of rising inequality were sidelined until recently.
While VoC plays a strong role within CPE, there is a substantial and growing literature that has moved beyond VoC. In doing so, many CPE authors take a position and ask questions very close to those of PKE, and at times they build explicitly on PKE. Colin Crouch proposes the concept of “privatized Keynesianism” to describe a regime where private consumption (rather than government spending) is financed by credit. Colin Hay uses the term “Anglo-liberal growth model” to describe more specifically how rising asset prices and equity withdrawal give rise to a growth model based on credit creation. 15 This is very close to the PK concept of debt-driven growth to be discussed later. In their discussions of the Euro crisis, Andreas Nölke and Magnus Ryner prominently feature PK contributions and specifically the juxtaposition of export-driven and debt-driven growth models and how their interaction generates systemic instability. 16 Baccaro and Pontusson’s interest in PKE thus is symptomatic of a growing engagement of CPE and IPE with macroeconomic issues and theories. However, these attempts are as of yet eclectic and limited in scope. They explain specific episodes but do not reflect systematically on the macroeconomic foundations of CPE. Crouch and Hay explicitly analyze the UK and Anglo-Saxon experience during the pre-2008 boom and do not attempt a systematic theory of finance. My argument is that the best-developed stream within CPE, the VoC approach, is closely wedded to parts of contemporary mainstream economics, which hampers its ability to comprehend the changes brought about by financialization and thus fails to understand the instability of neoliberal varieties of capitalism. CPE needs to consider its macroeconomic foundations more systematically and, in particular, how it explains the economic growth process and crises.
New Keynesian Economics and the Three-Equation Model
Hope and Soskice make the case for New Keynesian—that is, mainstream economic—foundations for CPE. To appreciate the significance of Hope and Soskice’s proposal, we need to situate NK theory in the context of competing economic paradigms and their changing dominance over time. In the face of the Great Depression, Keynes had called for an overhaul of both economic policy and economic theory. Starting with the New Deal, Keynesian arguments indeed influenced policymaking in the postwar era. Immediately after the publication of the General Theory, the debate about its interpretation began. 17 John Hicks led the mainstream side by proposing the IS-LM model to marry neoclassical and Keynesian arguments. 18 This model became an essential building block of the neoclassical Keynesian synthesis that dominated economics textbooks in the postwar era. The synthesis regarded Keynesian arguments as a special case of the neoclassical general equilibrium model that arises when wages and prices are not fully flexible, for example because of collective bargaining agreements that prevent immediate wage adjustments. Such wage and price rigidities occur in the short run, but market forces would force wage or price change over longer periods. Analytically it posited that in the short run, involuntary unemployment was possible and thus the economy had Keynesian features, whereas in the long run, markets would clear and the economy would thus be represented by neoclassical general equilibrium where all markets clear: that is, there is full employment. Theoretically the Keynesian revolution was thus domesticated, and for a few decades an uneasy truce between the (mainstream) Keynesians and the neoclassicals emerged and gave rise to the distinct fields of (Keynesian) macroeconomics and (neoclassical) microeconomics. The truce ended with the New Classical revolution of the 1970s and 1980s, an important dividing line in the development of modern economics. The New Classicals postulated that macroeconomics needs to be based on microfoundations, with rational, utility-maximizing individuals and market clearing via the price mechanism. In other words, macroeconomics had to be built on strictly neoclassical assumptions. This transformed the field of macroeconomics, but it also affected the discipline of economics by marginalizing approaches that did not adhere to methodological individualism (such as PKE).
NKE emerged from the New Classical counterrevolution of the 1970s by accepting methodological individualism and the requirement for microfoundations of macroeconomics. But it argued that wage and price rigidities can arise even in a world of perfectly rational individuals if they face transaction costs or information asymmetries. If changing prices is costly (e.g., because it requires physically changing price tags or printing menus) or if management uses wages as an incentive mechanism to elicit higher effort by workers, optimizing behavior will not result in clearing markets. In other words, NKE is an attempt to reformulate some Keynesian arguments, such as that involuntary unemployment can occur and government deficit spending can be socially useful, within a strict neoclassical optimizing framework. NKE has changed over time. In the 1980s and 1990s NKE analyzed specific instances of market failures, such as involuntary unemployment and credit rationing. 19 In the course of the 2000s NKE moved closer to neoclassical general equilibrium models (technically: NK dynamic stochastic general equilibrium models). An important feature of these is that recessions and financial crises are understood as the result of exogenous shocks to an economic system that has self-healing properties and will revert to its supply-side equilibrium.
The three-equation model, which Hope and Soskice suggest as the macrofoundation of CPE, is a simplified textbook version of the NK model. It is widely used for monetary policy discussions and consists of the following three macroeconomic relations. The first is a demand equation, which summarizes the goods market equilibrium and portrays output as, in the short run, determined by demand and depending negatively on the interest rate. The second is a Phillips curve, derived from the labor market, which links inflation and output. Higher levels of economic activity come with higher employment, which leads to higher inflation. And the third is a monetary policy reaction function that depicts how central banks react to changes in inflation and output. Central banks are assumed to be either inflation targeting or following a Taylor rule, that is, responding to higher inflation (or more precisely, the deviation of actual inflation from the central bank inflation target) by increasing the interest rate.
I want to highlight two important features of the model. First, although demand plays a key role in these models in the short run, in the long run they are supply-side-determined. In the words of Hope and Soskice, “The supply side of the economy . . . pins down the equilibrium levels of output and unemployment in the medium run.” 20 The model is a version of the natural rate of unemployment, or NAIRU (nonaccelerating inflation rate of unemployment), models. The natural rate is the unemployment rate at which inflation is stable, and it depends only on labor market institutions such as the degree of collective bargaining, the structure of the welfare state, and the organizational strengths of labor unions, not on demand factors or actual unemployment. The Phillips curve gives a short-run trade-off between inflation and unemployment, but in the long run the economy gets back to the equilibrium unemployment rate. In other words, government spending can manipulate economic activity in the short run, but eventually the economy returns to its equilibrium. The model belongs to the synthesis tradition with a Keynesian short run and a neoclassical long run. For economic policy this means that if governments want to lower unemployment beyond the short term, they need to pursue labor market reform rather than demand management. The NAIRU theory thus has justified the deregulation of labor markets and weakening social protection. 21
Second, the three-equation model does not explicitly model the financial sector. The key financial variable is the interest rate, which is set by the central bank. Financial variables, such as business debt, household debt, or asset prices, do not feature in the model; portfolio decisions or speculative dynamics are absent. This approach implicitly assumes that financial markets are operating smoothly, that is, changes in the central bank’s base rate are duly reflected in lending interest rates. Asset price bubbles or booms in private sector lending are regarded as exogenous, which makes sense only if financial markets are regarded as relatively stable and efficient. Importantly, the three-equation model does not allow CPE analyses to incorporate endogenous financial instability. This is reflected in the precrisis version of the macroeconomics textbook by Carlin and Soskice, 22 which devotes only a single, short section to bank runs (due to asymmetric information and panics) and discusses them largely as historical phenomena. 23 In short, the three-equation model is about consumer price inflation and central bank policy, not about financial instability.
Hope and Soskice advocate basing CPE on the NK three-equation model. Indeed, we notice a strong complementarity between the institutional focus of VoC and the supply-side focus of NKE. Although VoC does not primarily aim to explain economic performance but rather analyzes institutional differences and similarities across countries, it clearly shares some ideas about the working of the economy with NKE. By not considering demand-side developments and focusing on institutions, VoC implicitly assumes that actual demand growth will adjust to the institutional supply-side equilibrium, which is why VoC and NKE make good companions. There is, however, a substantial difference in how NKE and VoC conceive of these supply-side institutions. To illustrate this, consider the labor market. For NKE, labor market institutions such as collective bargaining lead to wage rigidities and consequently unemployment. There is usually a unique (liberal), socially optimal institutional equilibrium. In contrast, in VoC, institutions play a more constructive role, and different sets of institutional arrangements can deliver similar results in terms of competitiveness. There are (at least) two institutional equilibria that allow for competitive outcomes. However, NKE and VoC both neglect financial instability and the role of demand for growth beyond the short run.
Post-Keynesian Economics
PKE is the stream of Keynesianism that emphasizes analytical differences with neoclassical economics and in particular rejects methodological individualism. 24 It has its origins in the circle of Keynes’s collaborators (Robinson, Kaldor, Kahn), who discussed drafts of Keynes’s General Theory. Michał Kalecki had independently developed arguments very similar to those of Keynes, but his work was translated into English only later. 25 Keynes’s own analysis emphasized short-run dynamics and argued that in a world of fundamental uncertainty, decisions—in particular those regarding investment expenditures and financial portfolio allocations—will not be fully rational, as the future is unknown, but will depend on social norms (business sentiment), simple heuristics, and social-psychological phenomena. These result in herding behavior that can temporarily stabilize norms, but they can also give rise to cyclical dynamics if the heuristics include projecting current price movements into the future. As a consequence, market economies will in general not have a self-adjusting mechanism and, specifically, will not guarantee full employment. The employment level will instead depend on investment decisions, which determine aggregate demand and output via the multiplier process. Business cycle fluctuations thus are primarily understood as driven by changes in investment expenditures due to changes in expectations and to financial markets prone to speculation.
Post-Keynesians endorsed the more radical elements of Keynes and rejected the neoclassical Keynesian synthesis. The project was to develop a distinctly Keynesian theory that breaks with rational behavior assumptions and market-clearing models. They sought to generalize Keynes’s (short-run) theory of effective demand into a theory of growth and distribution. This effort involved a criticism of (neoclassical) marginal productivity theory that posits that wages and profits are determined by technological factors and households’ preferences, which culminated in the Cambridge capital controversies. 26 PKE went beyond Keynes in three respects. First, the neo-Ricardian and Kaleckian streams drew on the tradition of classical political economy and its class analysis to interpret income distribution as the outcome of social struggles and power relations. In their macroeconomic models, income distribution plays a core role. Second, monetary Keynesians developed further Keynes’s analysis of fundamental uncertainty, liquidity preference, and endogenous money creation. Hyman Minsky formulated this into the financial instability hypothesis of endogenous financial cycles (to be discussed below). Third, Kaldor’s argument of cumulative causation clarified mechanisms by which supply constraints adjust to demand pressures over longer periods. PKE formed a distinct school of thought in the 1970s when, under the monetarist and New Classical counterrevolution, mainstream economics narrowed theoretically and methodologically as post-Keynesians rejected the requirement of rational behavior microfoundations for macroeconomics. 27
For the purpose of contrasting it to NKE, I will highlight four features of PKE (to be discussed in more depth below). 28 First, income distribution plays a key role in PKE. Distribution is understood to reflect power relations, and PKE offers a flexible theory of demand regimes that allows for wage-led as well as profit-led demand regimes. Second, financial instability is regarded as an intrinsic feature of market economies. That it is intrinsic is due to the credit-driven endogenous money creation in PKE and the assumption that in a world with fundamental uncertainty, actors will adopt simple behavioral rules (often called heuristics) which can lead to herd behavior and generate boom-bust cycles in financial asset prices. Third, PKE asserts that demand matters in the short as well as in the long run. It specifically rejects the notion that economic activity is anchored in the long run in some natural rate of unemployment. This is because induced technological progress and hysteresis in the labor market ensure that the supply side adapts, at least to some extent, to demand. Fourth, what was just stated about technological progress and hysteresis has important implications for the effectiveness of government policies. Post-Keynesians assert that fiscal policy will be highly effective in times of financial crises as it has demand-side as well as supply-side effects. 29
Distribution and Demand Regimes
PKE builds on a long tradition in political economy that highlights the importance of income distribution, in particular between profits and wages. 30 Distribution lost center stage with the shift from political economy to modern economics (in the late nineteenth and early twentieth centuries), and mainstream macroeconomics has until recently not considered income distribution as an important factor (but was concerned with the effect of interest rates or prices on demand). In contemporary PK debates, the Bhaduri-Marglin model has become an important reference point in analyzing the interaction of income distribution and demand formation and has been used by Baccaro and Pontusson. 31 Its contribution is to offer a general framework that allows wage-led as well as profit-led demand regimes and clarifies the conditions under which they arise. An increase in real wages (for a given labor productivity and income) is likely to have expansionary effects on consumption (as workers typically will have higher marginal propensities to consume), but it may have negative effects on investment (as profit margins get squeezed) and on net exports (as higher wages negatively affect competitiveness). The sum of these different effects, which determine whether demand is wage-led or profit-led, will depend on institutional factors (e.g., the degree of inequality, the tax and pension system, and the structure of the financial system) that shape how much firms rely on retained profits for investment finance and how prominently demand and cost factors feature in management’s perceptions and can change over time. The model sheds light on different assumptions in economic paradigms. While mainstream and Marxists theories (often implicitly) assume profit-led demand regimes, post-Keynesians tend to assume wage-led demand regimes.
The Bhaduri-Marglin model has become a benchmark model in PK debates and has inspired numerous empirical studies to identify demand regimes in different countries. 32 Those studies differ in econometric estimation strategy. 33 A majority find that the consumption effect of a change in the wage share dominates the investment effect; in other words, the domestic components of the demand regimes tend to be wage-led. 34 But net exports may make total demand profit-led, in particular for small open economies, because the size of the net export effect will depend on the degree of openness of an economy (and on whether the export industry operates in sectors where competition via prices is prevalent).
The distinction between domestic and external effects and the fact that domestic and total (open-economy) demand regimes may differ have important implications and can give rise to a fallacy of composition. While individual countries can have profit-led demand regimes because of exports, for the world economy overall the export effects will cancel out. In other words, while individual countries may be able to export their way out of a crisis via wage cuts (or internal devaluation, as it is euphemistically called these days), the world economy overall cannot. Stockhammer, Onaran, and Ederer have argued with respect to the Euro area that many of its member states have profit-led demand regimes because of net exports. 35 However, most of their trade is within Europe, and thus the overall (European) demand is wage-led. Thus, a prisoners’ dilemma–type situation will arise: whereas for individual countries it may be advantageous to cut wages (because of export effects), collectively this may reduce demand, as consumption demand falls in all countries and competitiveness gains are canceled out by simultaneous wage restraint. Understanding the prisoners’ dilemma has been useful for understanding demand developments since the Euro crisis, where Troika policy recommendations were biased toward internal devaluation and wage cuts. That policy has resulted in rising European export surpluses but weak domestic demand, with an overall weak growth performance of the Euro area. 36
How large are these effects? Stockhammer, Onaran, and Ederer find, for the Euro area, that a one percentage point increase in the wage share may raise consumption by 0.4 percent of GDP, reduce investment (if at all) by 0.1 percent, and raise net exports by 0.15 percent. The total effect would be +0.15 percent of GDP. In other words, these effects are modest in size. Moreover, Stockhammer and Wildauer report that the effects of financial variables such as household debt and real estate prices have, in the recent past, been orders of magnitude larger. 37 In other words, the Bhaduri-Marglin model does not claim that income distribution is the most important determinant of growth. But even a modestly wage-led demand regime has important theoretical implications for the labor market equilibrium. 38 If demand is wage-led, wage cuts will lead to a reduction in consumption demand, which dominates the investment effect. This had been discussed by Keynes in Chapter 19 of the General Theory. 39 Keynes framed his analysis in terms of money wages rather than in terms of the wage share, but the core argument is the same. If wage cuts lower consumption expenditures (and investment is slow to react or insensitive to the wage cut), then a wage cut will not stimulate aggregate demand. If aggregate demand is falling, firms will have no reason to hire more workers. Falling demand could be offset if the fall in prices (which is likely to come with a fall in wages) has an expansionary effect, but there is no general reason (in a closed economy) to expect deflation to have expansionary effects. In an indebted economy, the opposite is likely. In short, a wage cut in a recession will not lead to an increase in employment. The labor market will not have the self-healing properties of a stable equilibrium.
The prominence of distributional issues in PKE contrasts with the absence of distribution in the three-equation model. Hope and Soskice’s claim that “the three-equation model is perfectly consistent with the income distribution being a determinant of consumer expenditure” is misleading, 40 since a wage-led demand regime fits uneasily with the three-equation model. The three-equation model assumes a stable equilibrium (given central bank intervention); thus it sidesteps the destabilizing effects of a wage cut (or distributional changes) highlighted by Keynes. In short, the three-equation model is not designed to illustrate Keynes’s perverse effects of wage cuts in a recession, whereas the Bhaduri-Marglin model can readily illustrate the contractionary effects of wage cuts in a wage-led economy.
Finance and Financial Instability
A second important difference between PKE and the NK three-equation model is with regard to the role of finance. PKE views the financial system as a source of instability, for two reasons. First, fundamental uncertainty implies that investors’ expectations cannot be fully rational. Expectations about the future are social conventions rather than based on fundamentals, as the relevant fundamentals and facts do not yet exist. People will thus resort to simple behavioral rules (heuristics); they will be driven by what Keynes called animal spirits. 41 Importantly, these behavioral rules will include social norms and comparisons. This can give rise to herding behavior, as people do what they see other people doing. One important convention (or heuristic) is extrapolative expectations, namely, the assumption that the past trends continue. For financial markets, that means that if asset prices increase, this heuristic leads to the expectation of further growth, which can result in a speculative boom. 42 If the bubble gets pricked, prices will fall and uncertainty rise; risky assets will be sold off. As a consequence, the demand for money (or more generally liquid and safe assets) will increase sharply. In a financial crisis there will thus be a sudden increase in a preference for liquid assets and a flight to safety.
Second, money is regarded as created endogenously by commercial banks as a side effect of their lending decisions. Therefore, loans create deposits. 43 This gives modern economies a high degree of flexibility as long as banks’ animal spirits suggest that it is profitable (and safe) to lend. Investment is thus not constrained by saving but rather the availability of credit. However, credit for investment (or more broadly, production-related activities) is only one option. Banks can also channel credit toward financial asset and real estate transactions, which can fuel financial bubbles. 44
PK theory thus regards financial instability and the emergence of financial cycles as a systemic feature of monetary production economies, with Hyman Minsky as the most important pioneer. There are two core mechanisms that can give rise to financial cycles, which will both be in operation in actual economies. 45 First, Minsky’s original writing focuses on a debt cycle driven by business investment. Over the course of a tranquil period, firms will become more optimistic. They will increase investment and start to accept higher leverage. In Minsky’s terminology, an increasing number of firms move from hedge to speculative financial structures. As debt has to be serviced out of the current cash flow, the fragility of the economy increases. Demand shocks that affect firms’ cash flows or increases in the interest rate may push firms toward bankruptcy. At the core, the cycle mechanism consists of a procyclical leverage and a debt-burdened demand regime: during a boom, firms need more external finance, but the resulting higher leverage leads to a negative effect of debt on demand; the interaction of the two effects can give rise to endogenous cycles. Note that this cycle mechanism puts nonfinancial businesses and their debt at the center of the story. A second mechanism, more recently developed, is based on asset prices and speculative behavior on financial markets. As investors follow simple behavioral rules (since global optimization is not feasible in an uncertain world), some investors, often called momentum traders or chartists, will base expectations about future asset price and capital gains on past performance, setting in motion bubble dynamics. During a boom these investors will make higher profits than more conservative (fundamentalist) investors, which encourages emulation by other investors. An asset price boom will thus lead to a recomposition of portfolios toward risky assets (liquidity preference and the demand for money declines), which puts a downward pressure on interest rates. The asset price boom, however, is fragile, as it is built on the expectation of further capital gains and comes with riskier portfolios and higher leverage. Once the bubble bursts, there will be a flight to liquidity, that is, an abrupt increase in the demand for money and safe assets, which drives up interest rates. Such boom-bust dynamics are often discussed with respect to stock markets, but arguably the same mechanisms operate in housing markets, which have a larger macroeconomic impact.
These two financial cycle mechanisms differ in that debt cycles are about business investment and expected future cash flows, whereas speculative cycles are about financial asset prices and expected capital gains. Importantly, both mechanisms, which are complementary, conceive of financial cycles as endogenous cycles that emerge spontaneously without the need for an exogenous shock. Similar mechanisms were also discussed by New Keynesians in the 1980s, in noise trader models. 46 But they are not part of the three-equation model or standard NK models, which treat financial bubbles as the outcome of exogenous shocks. Since the global financial crisis, various NK authors have tried to analyze why the self-healing properties of the market system seems paralyzed. Key to that has been the notion of the zero lower bound, that is, that the central bank cannot set nominal interest rates below zero. 47 However, from our perspective the key question is whether financial crises are due to exogenous and thus unforeseeable shocks or whether they are, as Minsky claims, systemic features. This issue has a direct policy correspondence in the question whether financial crises are predictable (from the point of view of policymakers). In fact, leading crisis indicators, in particular bank capital ratios, liquidity ratios, and house price growth, did predict the 2008 crisis. 48
The Supply Side: Path Dependence and Unemployment Hysteresis
A key feature of the three-equation model and indeed of all neoclassical theories is that it is rooted in a supply-side-determined long-run equilibrium. PKE has a very different view of long-run growth in that it asserts that demand factors, in particular animal spirits and social norms affecting business investment, also play an important role in a growth context. Supply constraints matter, but supply is elastic and will adjust to demand pressures to some extent. 49 Two channels ensure this outcome. First, the available technology and machinery are not simply taken as given but respond to demand and output developments. In particular, productivity growth is not only determined by exogenous factors (such as the extent of research and development activities) but in part driven by demand growth via learning by doing and dynamic returns to scale (the so-called Kaldor-Verdoorn law), and wage growth induces labor productivity growth. 50 Kaldor refers to this as a cumulative growth process, where increased demand feeds into increased productivity growth, that is, supply-side constraints are shifted. Second, in the labor market, post-Keynesians typically accept the existence of a short-run Phillips curve but regard the NAIRU (the unemployment rate at which inflation is stable) as endogenous to economic performance. Cyclical unemployment will turn into structural unemployment. 51 As a consequence, unemployment hysteresis arises: in a severe recession there will be an increase not only of actual unemployment but also of the NAIRU. The macroeconomic implication of this is that the system has a memory and demand shocks have a long-lasting impact on employment. 52
PKE thus conceptualizes the growth process as path-dependent. This notion is also employed by CPE scholars, with, however, a somewhat different meaning. CPE scholars often refer to path dependence as institutional persistence. Once established, institutions, such as a certain type of welfare state or the bank-based financial systems developed during late industrialization, become a permanent feature of an economy because they give a rise to a political constituency that supports them and generate constraints on firm behavior. In PKE, path dependence refers to demand shocks’ having lasting effects, for example, a severe recession like the one in 2008 altering the subsequent growth path and giving way to secular stagnation.
An important consequence of path-dependent growth is that it has strong implications for the ability of the state to shape economic growth, in particular in the face of recessions. In models with a unique supply-side-determined equilibrium, there is not much fiscal policy can do. However, things are very different in a path-dependent economy. Such an economy will not see a swift return to the original equilibrium but may experience periods of prolonged stagnation. A sharp crisis will be followed not by an energetic recovery but by weak growth. 53 One specific issue where this surfaces is the ongoing debate on the size of fiscal multipliers, that is, the effect of a change in government spending on GDP, during a crisis. According to NK theory, fiscal multipliers are modest (typically below or around one) and short-lived. For a year or two the effects of government spending may be substantial, but beyond that the economy will revert to its supply-side-determined natural rate equilibrium. Government expenditures are ineffective over long periods. In the language of macroeconomists, long-run fiscal multipliers are zero. However, if growth is path-dependent, this opens the door for government spending (but also various private sector shocks) to have long-lived effects. More specifically, PKE argues that deep recessions, in particular financial crises, will leave lasting scars on the economy. They cause a drop in the level of income, and they will also affect the subsequent growth path as supply adjusts to lower growth. From this perspective, the so-called secular stagnation that followed the global financial crisis is not surprising but an indication that demand shocks have long-lasting effects. The flip side of this is that fiscal policy will be particularly effective during such crises. Indeed there is mounting evidence that fiscal multipliers are substantially larger in times of recession than during the upswing of the economic cycle, 54 and the International Monetary Fund famously admitted having underestimated the size of fiscal multipliers during the global financial crisis. 55 As demand stimulation will affect the supply side via unemployment hysteresis and induced productivity growth, the effects of fiscal policy are long-lasting. 56 In contrast to NKE, PKE perceives of government policies not only as cyclical stabilization but as having potential for state-led growth strategies.
Much of PKE is concerned with advanced economies. However, there are also versions of (or close relatives to) PKE that address the developing economies. Latin American structuralism and the balance of payment constrained growth models have a close affinity to PKE. 57 They argue that there are capital and technology constraints in developing economies, in particular that capital goods need to be imported. Thus a crucial element of the elastic supply conditions discussed above depends on the ability to import machinery. As a consequence, the balance of payments often forms a constraint for growth as investment goods need to be imported, which requires access to foreign exchange. The structuralist view of international trade is less benign than the Ricardian view. Developing countries mostly export basic goods (agricultural or low-tech manufacturing goods), which tend to have a higher price elasticity and lower income elasticity than the exports of advanced economies. This may trap developing economies in low-income equilibria. The market mechanism will not guarantee the upgrading of these economies.
Modern versions of structuralism as well as of PK theory further emphasize financial asymmetries between developing and advanced economies. Financial markets are typically less liquid and financial institutions weaker. This means that opening up the capital account can easily result in capital flows that are large relative to domestic finance and subject countries to the international financial cycle. 58 Importantly, firms and governments of developing countries often cannot borrow from abroad in their own currency. Thus the Minskyan debt cycle takes a particular twist, as debt is denominated in foreign currency and (because of international capital inflows) is highly procyclical. The currency appreciates during the boom, easing the debt burden, inflating asset prices, and encouraging (imported) luxury consumption. The boom gets amplified but relies on foreign currency debt. Once growth slows down, so do capital inflows, and the currency depreciates, increasing the real debt burden. Such capital flow reversals have been a recurring feature of crises in developing economies, in particular in the East Asian financial crisis. Keynesian liquidity preference suggests a flight to safety during a crisis, which will mean flight to the major international reserve currencies. Thus international capital flows and the associated hierarchy of currencies pose additional constraints on growth and economic policy in developing economies.
Contributions of PKE to CPE
CPE deals with questions of economic performance across countries as well as with institutional diversity across countries and how the two interact. CPE thus needs an understanding of how the economy works as well as how institutions work. At the core of my argument is the assertion that CPE, in particular the VoC approach, has based its economic analyses on mainstream economics. This limits the analysis of the relation between distribution and growth and neglects the role that finance plays in modern economies. It overstates the stability of the capitalist growth process and understates the potential effectiveness of government interventions. PKE offers an approach that highlights the instability of the growth process and lends itself to an analysis of income distribution and power relations. This section highlights specific areas where PKE can offer valuable insights for debates in CPE.
Financialization
Financialization has been one of the major structural changes of capitalist economies over the last few decades and is by now the subject of a rapidly growing literature across several academic disciplines. I will argue that VoC has structural problems in understanding the changes brought about by financialization that are linked directly to its theoretical framework, which has the notion of institutional competitive advantage at its core. Its analysis of finance has thus focused on how businesses finance investment and how financial relations affect training and labor relations. The core distinction VoC has drawn on is between (arm’s-length) market-based financial systems and (relationship-based) bank-based financial systems. These correspond to and complement other institutions in the liberal and coordinated market economies, respectively. In particular, the bank-based financial system allows forms of patient finance to firms, which enables them to provide long-term contracts with and training for their workers. In contrast, the market-based financial systems in liberal market economies created finance with shorter time horizons and financial instruments that are frequently valued at markets. Overall, VoC’s analysis of finance as corporate finance has resulted in a functionalist treatment of finance, in relation to achieving competitiveness, rather than a systematic engagement with current activities of financial institutions. Two issues are worth highlighting. First, as a consequence of the functionalist corporate finance focus, financial instability has long been neglected in VoC analyses. 59 Second, households’ financial relations and balance sheets have been sidelined, as they do not influence competitiveness directly. The focus on competitiveness has hampered CPE’s understanding of the impact of financialization.
So what have been the effects of financialization? It has structurally affected various sectors of the economy and, ultimately, has changed macroeconomic dynamics. For businesses, it has meant shareholder value orientation and has come with a shift in the balance of power between owners, management, and workers (or, more generally, stakeholders), which is reflected in changes in corporate governance, increased dividend payments, and share buybacks. 60 For households, it has come, in many countries, with dynamically increasing household debt, driven by mortgage debt, and a growing importance of capital-based pension systems. For the financial sector, it has led to a shift away from business finance to mortgage finance, a shift toward fee-generating activities (securitizations), and a rise in nonbank financial institutions (shadow banking). 61 The shift from lending to businesses toward mortgage lending and financial engineering (investment banking) is reflected in bank balance sheets. 62 Macroeconomically, it has contributed to rising income inequality and to the return of the financial cycle.
The focus on corporate finance by VoC makes it hard to appreciate a key change in lending practices: banks have moved toward financing financial asset transactions and real estate transactions rather than financing business. This shift of finance from businesses to households and real estate has powerful social and economic effects. Household debt, which has been found to be linked to the severity of recessions, 63 is predominately mortgage debt. Real estate is widely accepted as collateral and thus provides a powerful lever for real estate bubbles, and the price growth also leads to credit growth. But housing also has an important ideological function. 64 Not all of CPE follows VoC closely, but analyses remain partial. Herman M. Schwartz and Leonard Seabrooke propose an analysis of varieties of residential capitalism that highlights links between political preferences around housing and social policy. They offer a rich institutional typology of countries based on homeownership and housing finance, but their analysis describes institutional outcomes and not the macroeconomic implications of the central role of housing and boom-bust cycles. 65 Alison Johnston and Aidan Regan offer an analysis of the determinants of house prices that emphasizes the impact of wage bargaining coordination and the wage pressures that emanate from nontradable sectors on house prices. 66 The macroeconomic effects of financialization, in particular house price cycles, are not part of these analyses.
PKE has made a major contribution to the analysis of financialization. First, PK economists have been at the forefront of analyzing the impact of financialization on income distribution, which fits with the emphasis on the distributional dimension of finance in PKE. Eckhard Hein gives a PK analytical framework for the analysis of financialization and distribution. Petra Dünhaupt provides econometric evidence for the impact of financialization on the wage share. Kohler, Guschanski, and Stockhammer disentangle different dimensions of financialization and find that financial liberalization and rentier payments by nonfinancial businesses have had the strongest impact on the wage share. 67 Second, they have analyzed the impact of financialization and shareholder value orientation on business investment. Stockhammer provides evidence for the negative impact of shareholder value orientation on macroeconomic investment for major economies. Özgür Orhangazi as well as Daniele Tori and Özlem Onaran provide firm-level evidence of that effect for US and European firms, respectively. 68 Third, PKE offers a framework to analyze the macroeconomic effects of financialization, which include a return of financial cycles and, once we consider the impact of financialization in conjunction with rising income inequality, the emergence of two neoliberal growth models.
The Return of Financial Cycles and the Debt-Driven Growth Model
Baccaro and Pontusson have pioneered the use of growth models in CPE debates. They argue that under the influence of VoC, CPE has given excessive weight to questions of institutional configurations and supply-side phenomena. In contrast, Baccaro and Pontusson focus on the demand side of the economy and, drawing on PK macroeconomics, highlight the impact on demand of the income distribution between capital and labor. They apply their analysis to the cases of Germany, Sweden, Italy, and the United Kingdom. All four, they argue, had a wage-led growth model in the postwar era; but each has responded differently to distributional changes since 1980. Germany has pursued wage suppression to further its competitiveness and thus has embarked on an export-led growth model; the United Kingdom has followed a consumption-led growth model, fueled by rising household debt. Sweden, with the importance of its knowledge-based sectors, has been able to pursue a balanced growth path in which both consumption and exports grew. Italy is interpreted as case of an inability to find a new growth driver. That inability, Baccaro and Pontusson hypothesize, is related to Euro membership, which has led to an “equilibrium of fear” in which Italian exports are rendered uncompetitive, but there are substantial sectoral interests against leaving the Euro. From a PK perspective this is a welcome application of demand regimes to CPE, but we note two analytical shortcomings, one related to the treatment of the finance-led growth models, the other to the impact of income distribution (to be discussed in the following section).
Baccaro and Pontusson analyze the United Kingdom as a “consumption-led growth model” fueled by credit growth. This recognizes the importance of credit to households but fails to explain why credit to households has grown. Conceptually they treat household debt as consumer debt, but in fact most of household debt (typically 90 percent in advanced economies) is mortgage debt. Thus most household debt is not taken out to finance consumption but to finance housing transactions. Nor would banks be likely to lend to households to the extent that they have done purely on the basis of households’ desire to consume more. Banks lend based on collateral and expected repayment. As real estate is accepted as collateral, banks will lend in times of rising house prices. Thus to understand household debt dynamics, it is key to understand house prices. 69 Baccaro and Pontusson move away from the VoC focus on corporate finance and acknowledge the role of household debt. But they have little to say about the drivers of debt and therefore fail to offer a general account of financial instability.
PKE provides a systematic theory of endogenous (recurring) financial instability. Historically, financial cycles have played an important role for economic stability, and the Keynesian/Fordist period of tight financial regulation has been rather exceptional in that financial crises (banking crises and exchange rate crises) have been rare. 70 The era of financial deregulation has resulted in a return of such crises. Mainstream economics (including the New Keynesian three-equation model) regards them as the results of (unforeseeable) exogenous shocks. In contrast, Minsky and the post-Keynesians interpret them as a recurring systemic feature. Since the global financial crisis there has been a growing (mainstream as well as PK) literature that empirically substantiates the view that financial crises are indeed part of regular cycles. Claudio Borio notes that private credit growth and real estate prices are the two key financial variables correlated with the financial cycle and notes that financial cycles tend to be longer than regular business cycles. 71 Edward Glaeser surveys house price bubbles in the United States. 72 Mian, Sufi, and Verner provide evidence that increases in household debt lead to positive growth effects in the short run, but after three years these effects turn negative. 73 All of this is suggestive of cyclical dynamics.
From a Minskyan perspective the United States and United Kingdom experienced a house price cycle, which was facilitated, as is often the case during financial booms, by financial innovation, in particular mortgage securitization. 74 Real estate price cycles can arise when households and developers (and their banks) form extrapolative expectations. 75 During the boom an increasing number of households and their banks will adopt the conventional assumption that house prices keep increasing. As real estate is accepted as collateral, rising house prices also come with rising household debt levels, and a small but macroeconomically important part of this rising household debt will feed into consumption expenditures and residential investment, thus driving a boom in real growth. In the run up to the global financial crisis, this pattern was amplified, in particular in the United States, by securitizing mortgages and moving them off balance sheets. During the boom the financial fragility of households will increase and rising consumption will start to depend on rising house prices. Once house price growth flattens, the boom collapses. Effectively, households during the boom had been acting as speculators who took out loans in expectation of future capital gains (i.e., house price increases), even if households typically have a different self-perception and consider their investment in real estate assets safe. Once real estate prices start to fall and credit standards tighten, households have to engage in a lengthy and painful deleveraging process that requires them to reduce consumption expenditures (and in many cases default on their mortgage payments).
The Interpretation of Neoliberal Growth Models
Baccaro and Pontusson build on PK macroeconomics and the Bhaduri-Marglin model, but they take some analytical shortcuts. PKE distinguishes between demand regimes and growth models, whereas Baccaro and Pontusson conflate the two. Demand regimes describe the effects of a change in income distribution on demand (and its components: consumption, investment, and net exports). 76 If an increase in the wage share leads to an increase in demand, the regime is called wage-led. The definition of this regime is independent of the actual change in the wage share. A growth model is defined with respect to the main driver of growth; for instance, a debt-driven growth model is one where changes in debt actually contribute substantively to actual growth outcomes. Baccaro and Pontusson collapse the two concepts and refer to the Fordist period as a wage-led growth model that went into crisis in the 1970s. By implication, they conceive of neoliberalism as a shift to a profit-led growth model. Specifically they state, “Some heterodox economists use ‘financialization’ as an umbrella term for institutional or regulatory changes that have moved advanced capitalist economies onto a profit-led growth path.” 77 However, this is misleading. In fact most PK economists argue that demand regimes in most major economies remain wage-led in the neoliberal era. 78 The emergence of debt-driven growth models is understood as a demand stimulation, due to rising real estate prices within a wage-led demand regime. While neoliberalism and financialization have changed the growth model, they have not changed the demand regime.
The distinction may seem pedantic, but it has material implications for the interpretation of neoliberalism and for economic policy. If neoliberal economies were indeed profit-led, then a wage cut in a recession would have positive effects on demand and, presumably, on employment. If demand is wage-led, then the wage cut in a recession will contribute to a prolonged stagnation. Thus the Troika’s strategy of internal devaluation via wage restraint will be assessed differently. But the distinction between profit- and wage-led demand regimes also leads to different interpretations of the performance of countries. If Germany has had a profit-led demand regime in the last few decades (as Baccaro and Pontusson assert), then the Hartz reforms, which contributed to weak wage growth, would have stimulated the economy. If it has been wage-led (as Stockhammer, Hein, and Grafl claim), then the Hartz reforms help to explain why Germany had one of the weakest growth performances of the Euro area (prior to 2008). 79 Most post-Keynesians interpret neoliberal growth models as debt-driven or export-driven within a wage-led demand regime. This amplifies potential instabilities of the regimes in the face of crises and downward wage pressure and raises fallacy-of-composition problems for national wage policies.
PKE offers an analysis of the impact of financialization and inequality on growth models that differs from Baccaro and Pontusson’s. Lavoie and Stockhammer emphasize that both the export-driven and debt-driven growth models are based on financialization, and they are both unstable. This is straightforward for the debt-driven growth model, which relies on domestic financialization in the form of real estate booms and the causal mechanisms described above. An important implication of this is that the downturn of the cycle and the stagnation with deleveraging should be interpreted as part of the debt-driven growth model. Debt-driven growth and debt-driven stagnation are two sides of the same coin. But the seemingly more industrial export-driven growth models also rely on financialization. The flip side of the growing current account surpluses of export-driven growth is the rising external debt of their trade partners. Although the periodic exchange rate realignments during the Bretton Woods system put a limit on the extent of international trade imbalances, in the era of neoliberalism, the limit becomes the ability of the debt-driven economies to mobilize credit. In other words, external financialization (the liberalization of international financial flows) is a precondition for export-driven growth. There is an important asymmetry in the instability that the two growth models generate. While the debt-driven growth model relies on growing household debt within its own national economy, the export-driven growth model generates trade imbalances that rely on growing foreign liabilities of their trade partners. In other words, the instability is to some extent externalized.
Central Bank Policy and the Political Economy of Money
The PK theory of financial instability also has important implications for the role of central banks. Central bank policy and its interaction with wage bargaining structures featured prominently in CPE debates in the 1990s. This was in response to shifts in mainstream economics and policymaking that regarded price stability as the primary objective of the central bank and argued that independent central banks were best suited to fight inflation. They could use monetary policy, that is, interest rates, to counter inflationary pressures without regard to the short-term social costs. As monetary policy was neutral with respect to output over the longer period, this would help create a low inflation environment without damaging long-term growth. CPE analyses highlighted the central role of wage-bargaining institutions and qualified the claim that monetary policy was always neutral with respect to long-term growth. Peter Hall and Robert Franzese argued that the effectiveness of central bank signaling depended on wage-bargaining institutions. 80 That is because in coordinated bargaining systems unions would internalize wage restraint, as they anticipate central bank reactions to high wage demands. European unification would effectively lead to decentralization of wage bargaining and thus would undermine central bank effectiveness. Torben Iversen develops a nonlinear model of the interaction of central bank policy and wage-bargaining systems and argues that nonaccommodating (conservative) monetary policy would result in higher unemployment under medium degrees of bargaining centralization. 81 He concludes that central bank policy does have real effects. CPE contributions have added institutional detail to mainstream macroeconomic analysis and thereby modified some key conclusions.
While the nuanced analysis of the impact of wage bargaining is welcome, these CPE analyses accept the premises of mainstream economics and take a narrow view of the role of central banks. Price stability rather than financial stability is put center stage. It is telling, with the benefit of hindsight, that concerns about cross-border financial flows and instability on sovereign debt markets (as they erupted in the Euro crisis) did not feature in Hall and Franzese’s and Iversen’s analyses of European monetary integration. They take as a starting point that the purpose of the central bank is to ensure price stability, and thus they downplay its role in ensuring financial stability. Assuming that the central bank’s main goal is price stability either presupposes that financial markets are intrinsically stable or that any potential instability can be taken care of by microprudential regulation. Consequently the political economy of the central bank’s lender-of-last-resort function is not discussed. This downplays the range of instruments that central banks have but also understates the power they possess.
PKE offers an analysis of money as created by lending decisions of banks. This statement presupposes the existence of a monetary system. Post-Keynesians endorse a credit view of money, but several post-Keynesians have also contributed to (chartalist) state theories of money that emphasize that money is not a private institution; instead, issuing money is part of a state’s authority. 82 While historically this theory was closely tied to imposing tax liabilities and with them a currency in which they have to be paid, in today’s economy the dialectic between state authority and banking interests shapes the financial system. Which institutions are allowed to offer deposits, to what extent they are part of deposit insurance schemes, whether they are allowed to borrow from the central bank—in short, to what extent a financial institution’s liabilities are guaranteed as substitutes to central bank money—depend on state policies. These questions surfaced in 2008 when the Federal Reserve decided to extend the range of institutions that can access emergency liquidity. Financial institutions will issue different forms of assets and liabilities. The central bank sits at the apex of the hierarchy of moneys and decides where on the hierarchy different private institutions (or their assets) are. 83
Periodically recurring financial crises have important implications not only for the range of a central bank’s policy instruments but also for its position of power. Katharina Pistor highlights in her “Legal Theory of Finance” that although during normal (noncrisis) times debt commitments have to be honored, in times of crisis the enforcement of all legal obligations may result in the financial system’s self-destruction. 84 Thus central banks often act as lender of last resort and thereby suspend normal market rules. In 2008 Lehman Brothers was allowed to go bankrupt, with devastating effects. Thereafter Western governments and central banks were committed to the survival of systemically important financial institutions. If in times of acute crisis part of the rules get suspended, the question is, For whom exactly? At this point the hierarchical structure of finance and the power relations underlying monetary authority become apparent. In times of crises the law is enforced asymmetrically (some financial institutions are bailed out, others forced into bankruptcy). It is national sovereigns (and their central banks) who, in times of crisis, have the power to issue money and thus save (or not) different institutions and actors. In short, the role of lender of last resort is not only an issue for financial stability but also a power relation.
These power relations are about the stratifications of different players within private financial actors, but they apply also to states. Central banks are now routinely acting as lenders of last resort for private financial institutions. However, they were historically founded as funders of governments. 85 Much of modern macroeconomics is built on the assumption that monetary and fiscal policy can and should be strictly separated. A part of this separation is ruling out direct government financing by the central bank (as this would create moral hazard problems and would anyway be unable to generate growth in the medium term). The importance of this can hardly be overstated in context of the Euro crisis. While the Fed and the Bank of England used quantitative easing to indirectly finance government expenditures, the European Central Bank (ECB) initially refused to play that role. It relied on (private) ratings of government bonds and threatened not to accept some member states’ debt. This arguably explains the escalation of the Euro crisis. Among the advanced economies, the Euro area was unique in that the financial crisis turned into a sovereign debt crisis. 86 This created a situation in which southern European governments had to submit to the Troika rescue packages and impose austerity on economies in recession. The sovereign debt crisis ended after the ECB committed to “doing whatever it takes,” in particular to buying government debt from states under pressure. There is little in the three-equation model to help us understand the significance of the central banks’ readiness to commit to buying government debt.
From a policy perspective, the CPE of central banking understates the range of policy instruments that central banks have. Since around 1980 the set of policy instruments has been purposefully restricted, and central banks have tried to influence credit volumes through open market transactions. However, credit guidance and banking regulation have been used historically to steer the economy, and since the crisis central banks have rediscovered instruments to counteract financial bubbles and lending booms. Specifically, CPE downplays the significance of the central bank balance sheet. The central bank is a bank and as such it can lend and it can buy financial assets. Since 2008 quantitative easing has effectively been used to indirectly finance governments, as most of the assets acquired are government bonds (these are bought on the secondary market; thus their purchase does not constitute direct government financing). Central banks could also be used, say, to finance direct transfers to households (“QE for the people”), to directly finance government expenditures (say, government investment), or to finance a national investment bank.
In short, CPE has so far failed to fully appreciate the impact of financialization and in particular the significance of financial instability. The VoC approach, analytically centered on the concept of competitiveness, has conceptual blind spots as regards household debt and the dysfunctional aspects of finance. The shift to lending to households and the return of financial crises thus have been underappreciated. Baccaro and Pontusson go part of the way of reconceptualizing comparative capitalism in terms of growth models and incorporating debt-led growth but without a systematic analysis of housing and financial cycles. I have argued that PKE offers a useful starting point for the macroeconomic analysis for that reconceptualization because it offers both financial instability and a framework to analyze neoliberal growth models. The PK theory of money and finance highlights systemic instability as well as power relations arising from the fact that the central bank sits at the apex of a hierarchy of money. Central banks act as lenders of last resort, both for private banks and for states, but whom they lend to, as revealed in the Euro crisis, is a power relation as well as a key macroeconomic act.
Conclusion
CPE is the study of institutions and economic performance across countries. It requires a theory of the economy as well as a theory of politics and institutions. Much of current CPE, in particular the VoC approach, relies explicitly or implicitly on mainstream economics. This article has argued that this reliance leads to an overstatement of the stability of market systems and fails to appreciate the changes brought about by financialization, namely, the return of financial cycles. PKE is proposed as an alternative economic grounding of CPE. It offers, first, a theory of demand regimes allowing for both wage-led and profit-led growth, which has been extended to analyze debt-driven and export-driven growth models. This aspect has already been recognized by CPE research, in particular through the work of Baccaro and Pontusson. However, their approach lacks an analysis of financialization and financial instability. The PK theory of finance is based on credit-created money and a theory of endogenous financial cycles. It thus offers an enhanced understanding of the process of financialization such as the shift to financial asset transactions and the return of financial cycles. Finally, PKE is based on the concept of fundamental uncertainty and pursues a class-analytic approach that regards income distribution as the outcome of power relations. In addition, its theory of finance and central banking incorporates power relations.
The overall vision of capitalism that emerges from the PK approach is of a dynamic system in an uncertain world. The growth path is not anchored in an institutional equilibrium, but rather one where growth dynamics, financial structures, power relations, institutions, and state interventions coevolve. Demand regimes may generate periods of growth as well as systemic instability. Political coalitions will form around growth models and states that stabilize an unstable economy. The growth path is temporarily stabilized by institutions and state interventions, but because these serve many purposes, in particular crystalizing power relations and enabling class compromise, they will not always be conducive to growth over longer periods. A key source of instability is the financial sector. Asset prices and credit volumes, in an uncertain world, are guided by expectations and social conventions, which will often lead to overreactions and speculative bubbles. Financial instability thus is a pervasive feature of capitalism, but it has more than merely cyclical effects. First, financial crises leave long-lasting scars on the economy because of hysteresis effects. Second, in times of acute crisis, states often intervene and thereby critically shape the distribution of the costs of recessions and the path to recovery or stagnation. States also mediate distributional conflicts (or reinforce social domination) and can shape the sectoral composition of the economy.
This article has emphasized the analytical contributions of PKE relative to NKE and tried to illustrate how it can help illuminate areas where CPE has deficiencies in explanation. However, what is at stake here is not merely a matter of academic elegance and explanatory power. Ultimately, the choice of macroeconomic theory allows us to interpret economic and social problems and thereby frame policy interventions. In a time of secular stagnation with slow-growing economies, a large debt hangover, and persistent income inequality, the question is what CPE has to offer in terms of policy analysis and advice. Orthodox economic policies have arguably exacerbated these social crises, such as in the Euro crisis. While NKE offers a vision of limited but targeted intervention, 87 it remains wedded to a vision of market efficiency that discourages radical policies. The PK focus on financial instability, persistent involuntary unemployment, and the possibility of wage-led growth allows a broader set of policy proposals that may include QE for the people, growth via public development banks, job guarantee programs, substantive redistribution, and state-led innovation and decarbonization policies. In short, PK macroeconomic analysis not only offers a richer understanding of macrodynamics than NKE; it also enables CPE to develop a richer set of policy interventions.
Footnotes
Acknowledgements
Earlier versions of the article were presented at the workshop “Growth Models and the Politics of Macroeconomic Policy,” Max Planck Institute for the Study of Societies, Cologne, at the University of the West of England, and at the IPEG conference. The article has benefited from discussions there and from comments by Joseph Baines, Louis Daumas, Karsten Kohler, Paul Lewis, Inga Rademacher, James Wood, and the journal’s editorial board. The usual disclaimers apply.
Declaration of Conflicting Interests
The author declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author received no financial support for the research, authorship, and/or publication of this article.