Austrian Macroeconomics in Search of Its Uniqueness.

Author:Luther, William J.
  1. Introduction

    Austrian macroeconomics is experiencing a revival. Some scholars have used the Austrian business cycle theory to explain the Great Recession (O'Driscoll 2009; White 2009; Boettke and Luther 2010; Horwitz and Luther 2011; Koppl 2014). Others have employed Austrian insights to consider the effectiveness of monetary and macroprudential policies (Garrison 2009; White 2010; Selgin 2010; Salter 2016; Salter and Smith 2017; Salter and Tarko 2017). Moreover, as Koppl and Luther (2012) and Cachanosky and Salter (2017) make clear, these efforts have not been limited to self-ascribed Austrians.

    Given the renewed interest, we consider the essential features of an Austrian macroeconomic model and then ask whether these features are unique. We argue that the temporal aspect of the structure of production is not an essential feature. Malinvestments in any dimension (e.g., time, geography, type, etc.) can generate the predicted boom-bust cycle so long as there are costs to reallocation. However, the view that nominal shocks have long-term consequences because costs are incurred to remedy past mistakes is not uniquely Austrian. In particular, we note similarities with the New Keynesian notion of hysteresis.

  2. Essential Features

    What is Austrian macroeconomics? In any progressive research program, one will find disputes over what it means to be working in a particular tradition and what distinguishes the in-group from the outgroup. Austrian macroeconomics is no different in that respect. One might use the term to refer exclusively to the historical Austrian business cycle theory advanced by Mises (1912) and Hayek (1935). Others might use it to denote a more modern variation, which Garrison (2005) has called "capital-based macroeconomics." Still others refer to a "new Austrian" or "neo-Mengerian" macroeconomics (Salter 2017; Wagner 2007, 2010, 2012).

    We use the term Austrian macroeconomics to denote a class of models that account for macroeconomic fluctuation in a way that is broadly consistent with (but not defined by) the views put forward by Ludwig von Mises and F. A. Hayek. (1) The class of models we have in mind includes the traditional Austrian business cycle theory (e.g., Hayek 1935) as well as modern restatements (e.g., Garrison 1984, 2002; Horwitz 2002). More generally, it includes any model with the following essential features:

    * microfoundations

    * monetary non-neutrality

    * capital specificity and the potential for malinvestment

    * costly errors

    We discuss each in turn.

    A. Microfoundations

    Methodological individualism has been a central tenet of Austrian economics since its founding--and this approach is no less necessary in the domain of macroeconomics. Individuals acquire information, devise plans, and make decisions. They might coordinate their actions in households or firms. Nonetheless, the individual is the ultimate unit of analysis, and understanding why individuals do this or that --what is sometimes referred to as the deep parameters of the model --is essential for assessing the welfare consequences of a given shock or policy response (Lucas 1976). As such, macroeconomic models in the Austrian tradition must be based on microeconomic foundations.

    In advocating the need for microfoundations, some Austrians have taken a hard stand against aggregate or representative agent models. Bagus and Howden (2012, p. 274) claim that a majority of "errors stem from a too aggregative approach to economic theory" and dependence on "the Keynesian terminology of 'aggregate demand'." They echo Hayek's (1931) concern that "Mr. Keynes' aggregates conceal the most fundamental mechanisms of change." Others have stressed the need for heterogeneous agent models and a corresponding refocus on the interactions between agents (e.g., Koppl 2011; Wagner 2012). (2) Such views would seem to imply that microfoundations are necessary but not sufficient; that a specific type of microfoundations is required.

    We do not go so far. Rather, we maintain that a good model identifies a specific and potentially significant mechanism in a clear and concise way. The relevant question is not whether aggregate or representative agent models are appropriate. Instead, one should be concerned with the appropriate level of aggregation or the appropriate degree of heterogeneity given the task at hand. For example, White and Selgin (2017) depict the Austrian business cycle theory using the standard textbook aggregate demand-aggregate supply framework, where the microfoundations are implicit but fairly well understood. Given their objective, the aggregate model seems appropriate. (3) This is not to deny that excessive aggregation might lead one to inappropriate conclusions or that important details are glossed over when agents are assumed to be homogeneous. It is merely to acknowledge that a simple model is, in some cases, sufficient to make one's point.

    A related issue concerns the behavioral and epistemic assumptions of agents populating Austrian macroeconomic models--specifically, whether they should be assumed to have rational expectations. Bilo (2017c) summarizes the debate. On the one hand, we reject the assertion that an Austrian theory of macroeconomic fluctuation necessarily "conflicts with the rational expectations" assumption (Bilo 2017c, p. 14). Salter and Luther (2016) offer an Austrian business cycle model where agents have rational expectations. (4) On the other hand, we do not believe Austrian macroeconomics is "bound to remain on the fringe of discussions among professional economists" unless rational expectations are adopted (Bilo 2017c, p. 21). As Reis (2017, p. 21) makes clear, mainstream macroeconomists "have made much progress in the last three decades to provide alternatives to the assumptions of [...] rational expectations." As such, we view the rational expectations assumption as an optional feature of Austrian macroeconomic models.

    B. Monetary Non-Neutrality

    Modern macroeconomists can be crudely divided into two camps: those who maintain that macroeconomic fluctuation is driven entirely by real shocks and those who leave some scope for nominal shocks. The Austrians are firmly in the latter camp. (5) Austrians maintain that money is non-neutral: nominal shocks have real consequences.

    Belief in the non-neutrality of money is not particularly alienating in the economics profession today. Most modern macroeconomists accept that money is non-neutral in the short run. In standard textbook models, agents face a signal-extraction problem (Lucas 1972; Sargent 1991) or are subject to information cascades (Mankiw and Reis 2002, 2007) or find it prohibitively costly to collect information and update prices instantaneously (Alchian 1969) or encounter some other friction that enables real disturbances from nominal shocks. These mechanisms for generating non-neutrality seem entirely consistent with Austrian macroeconomics. (6)

    To the extent that Austrians part ways with the bulk of the profession on the non-neutrality of money, it is with regard to the time horizon over which non-neutrality prevails. Most macroeconomists accept that money is neutral--or, approximately neutral--in the long run. Austrians, in contrast, argue that money is non-neutral in the long run. Bilo (2017d, 2017f) and Bilo and Wagner (2015) have revived the ideas of the sixteenth-century economist Richard Cantillon in noting that monetary disturbances alter individual decisions and, as such, affect the resulting distribution of wealth. (7) Hence, money has lasting effects. In the model put forward below, we go further in arguing that monetary shocks also affect the equilibrium (or, natural) path of real output in the long run. In any event, we maintain that an Austrian macroeconomic model must acknowledge some potential for monetary non-neutrality.

    C. Capital Specificity and the Potential for Malinvestment

    Austrian macroeconomics developed, in part, as an extension of Austrian capital theory, which took it for granted that capital was specific to particular production processes. If capital were homogeneous, one need only consider whether individuals accumulate the optimal amount of capital, or, alternatively, whether they might be induced to accumulate too much or too little capital. In asserting the specificity of capital, Austrians accept that there is an optimal composition of capital: that it is possible to have too much of one kind of capital and too little of another. In other words, Austrian macroeconomists are not merely concerned with the prospect of over- and underinvestment, but also with malinvestment.

    Historically, Austrians have emphasized the temporal dimension of production and, correspondingly, the prospect for malinvestment to occur over the time structure of production (e.g., Hayek 1935; Garrison 2002; Bilo 2017a, 2017e). By pushing real interest rates below their equilibrium values, the familiar Austrian story goes, an unexpected monetary shock induces entrepreneurs to invest more heavily in the earliest and latest stages of production at the expense of middle-stage investments. (8) "Attention to the time element in the process of production and, more specifically, the incorporation of an intertemporal capital structure," Garrison (1991, p. 303) writes, "are what characterize Austrian macroeconomics and set it apart from the more widely accepted formulations of macroeconomic relationships." Indeed, recent efforts to revise Austrian capital theory continue to assert that the temporal dimension is...

To continue reading