Making sense of mainstream and heterodox economics

unification does not rest on a shared perspective so much as a shared opposition to what is often called the neoclassical mainstream, and the promotion of scientific pluralism in economics. Heterodox economists often hold pluralism as a central goal, and argue in terms of opening research in economics to more than the one dominant paradigm.¹¹ But when it comes to shared ideas of more substance, most observers seem to agree that heterodox economics is really only, or at least primarily, unified by its opposition the neoclassical mainstream. But exactly what neoclassical means (or if it is even a useful term), or what unifies heterodoxy beyond that, is the object of heated debates today. Since heterodoxy seems to be a fundamentally relational concept, understood in relation to a neoclassical mainstream, understanding the intellectual divide between mainstream and heterodox economics requires a more refined understanding of that “neoclassical”

mainstream.

2. Making sense of mainstream and heterodox

international and US-oriented, and possible to study as such (Yonay and Breslau 2006:349).

In this section, I will draw on a survey of recent debates among historians of economic thought and heterodox economists on the nature of neoclassical, mainstream and heterodox economics. This preliminary literature review will serve to answer two main questions. First, in what sense can we talk about an enduring mainstream in modern economics? Second, what is the nature of its relation to heterodoxy? Starting with the relational classification concepts

“mainstream”, “orthodox” and “heterodox”, we will then move on to discussions about the substance of neoclassical economics and the contested question of how to characterise continuity and change in modern mainstream economics.

Classification trouble in the history of economic thought

To make sense of any research field, we need some sort of terminology to reduce complexity and cut through the vast mass of actual research being done. We need a classification of different forms of intellectual production, and a rationale for classifying as we do. For economics, the job of creating classifications has largely been the domain of historians of economic thought, though explicit discussion about classification as such is rare even among them (Colander 2000:128).

Nevertheless, every textbook on the history of economic thought has to rely on some sort of classification of at least the major periods or schools of thought. They commonly rely on labels like “classics”, “marginalists”, “neoclassicals” and so on.

Of course, beyond textbooks, historians of thought dispel with such labels in close empirical studies of authors or connections between a smaller number of researchers or local environments. But once big, overarching issues are at stake, like questions about essential attributes of the research of a historical period, or core features of more or less historically persistent schools of thought, such concepts are needed. At first sight, these concepts may seem straightforward. But I will try to show that there is often no clear consensus, and that these concepts are in fact often contested. When it comes to describing the characteristics of contemporary economics and the type of research done in its mainstream, there is no consensus on terminology beyond some of the most basic and broad terms.

Thus, we need to investigate the concepts involved and the way they have been used more closely.

The terms mainstream, orthodox and neoclassical economics are sometimes used as synonyms, which easily leads to conceptual confusion. Mainstream is a common but often rather loosely used term that could be thought of as a

sociologically defined category. According to the renowned historian of economic thought David Colander and his co-authors,

mainstream consists of the ideas that are held by those individuals who are dominant in the leading academic institutions, organizations, and journals at any given time, especially the leading graduate research institutions. Mainstream economics consists of the ideas that the elite in the profession finds acceptable (Colander et al. 2004:490)

To classify something as orthodox should also be rather straightforward. Colander and his co-authors point out that as opposed to mainstream, orthodoxy has a temporal dimension to it as a static representation of a backward-looking approach. To call a person or a school of thought “orthodox” clearly implies a conservative, frozen, unchanging quality, according to Colander.¹² This is also clear when one looks at the purpose of classifying something as orthodox: “in economics at least, the name for the orthodox school usually comes from a dissenter, who is opposed orthodox ideas, not from a supporter of the orthodox ideas” (Colander et al. 2004:491). Clearly then, applying the term orthodox to the currently dominant type of economics implies something more than the relatively neutral term “mainstream”. It is, indeed, a pejorative term used mainly by critics of orthodoxy to point to what is conceived as its unchanging nature.

Sometimes the term “orthodoxy” is used in an even more specific sense, to contrast it with the dissenting “heterodoxy”. Heterodox economist Frederic Lee makes explicit the connection to the etymological origin of the terms in theology.

Just as in the matters of the church, orthodoxy means conforming to the established doxa. Lee (2009:4) claims that the theological distinction between heresy and blasphemy can be transposed to economics, arguing that while the heretic is a true believer who holds some dissenting views, the blasphemer “is a non-believer who explicitly, through reasoned arguments, wit, and ridicule, rejects the state religion and its sacred doctrines and institutions”. Thus, while some degree of dissent is tolerated, “mainstream economists have attempted to suppress the economic ideas and arguments of blasphemous economists, whom they do not generally consider their brethren at all” (Lee 2009:6). This narrative of comparison between modern mainstream economics and the power dynamics of the pre-modern church shows just how strongly some heterodox economists identify as dissenters and oppositional, and it is perhaps also a sign of the extent to which this identification and the relation to the mainstream is bound up with

12 However, against this one could also claim that it is a rational strategy and a sign of a mature science to stick to a well-proven paradigm, and that this is the case in many sciences.

an emotionally charged lived experience for many heterodox economists (Morgan 2015:13).

There seems to be general agreement that the term heterodoxy is clearly defined only in its opposition to the perceived orthodoxy. For example, in a study based on a series of interviews with historians of economic thought, Mary Wrenn (2007) shows that there is no clear consensus about the meaning of heterodoxy and the precise boundaries between the mainstream and heterodoxy. However, most scholars seemed to agree that there is something like a community of heterodox economists and that it is characterised by being “pushed out” from the mainstream, which the heterodox economists in turn more or less reject. In his history of the movement, Lee (2009) similarly describes the heterodox project as united by a central rejection of what he interchangeably calls mainstream or neoclassical economics. Standing closer to the mainstream, Colander et.al.

(2004:492) also observe that among the multiple schools of thought that make up the heterodox movement, beyond the “rejection of the orthodoxy there is no single unifying element that we can discern that characterizes heterodox economics”. Commenting on this and a range of other similar assessments in the search for the essence of heterodox economics, heterodox economist and philosopher Tony Lawson (2006:485) concludes that “we appear to reach an apparently widely shared assessment of heterodox economics only in terms of what it is not, or rather in terms of that to which it stands opposed; the one widely recognized and accepted feature of all the heterodox traditions is a rejection of the modern mainstream project.”

We can now produce some first definitions. Mainstream economics, we might argue, following Colander and co-authors, encompasses the ideas and practices that dominant economists in leading academic institutions, organisations and journals find acceptable at any given time. This definition is agnostic with regard to any historical continuity or change. The term orthodox economics on the other hand implies that the mainstream is also a historically enduring project with some sort of common, stable core that resists change. Defining the self-conscious opposition to orthodoxy, heterodox economics is an umbrella term for the various schools of thought whose minimal common ground is their rejection of orthodoxy, as they understand it. If these three concepts should be somewhat clear by now, it is at least partly due to their relational character.¹³ Let us now turn to the substantial but elusive concept of neoclassical economics.

13 For example, what a mainstream theory is, is an empirical sociological question related to a specific scientific establishment at a particular time and place. Consider for example the role of Marxian economics during the Cold War at, say, universities in the Soviet Union and the United

First attempts to define “neoclassical” economics

The term neoclassical economics has been used for over a century to describe specific schools of thought or ways of doing economics. At least since the 1950s, it has been a common term known to anyone within the economics profession.

Still, there exists no consensus on a stable definition of the term. Indeed, the term can be taken to mean quite different things. For example, some use it to denote a quite well-delineated school of thought, while others use it in a much wider sense to describe a style of thinking that includes many different schools of thought in the more narrow sense. And even in that wider sense, exactly what the term means is highly contested. In general, it is probably uncontroversial to say that the term is most often used by critics, that is, by heterodox economists, even to the extent that Colander (2000:132) irritably exclaims: “I can always tell when I am around heterodox economists by the number of times I hear the term”. Still, the term is hardly foreign to most mainstream economists.

A measure of the extent to which the term is contested is a recent volume with contributions from leading heterodox economists, What is neoclassical economics?

(Morgan 2015). This collection revolves around a recent paper by Cambridge philosopher of economics Tony Lawson (2013) bearing the same title. Lawson traces the origins of the term back to its coinage by Thorstein Veblen in a 1900 paper, in order to reinterpret what Veblen actually had in mind when he introduced the concept. It turns out that Veblen used the concept to shine a light on something very different from the various contemporary uses of the term.

Lawson (2013:34) argues that the term “should be dropped from the literature”.

The ensuing debate has now filled the abovementioned volume (Morgan 2015) of reactions to the paper, which points to the fact that the debates on the definition of “mainstream” or neoclassical” economics is not an issue of orthodox versus heterodox interpretations. There is no consensus on what these terms imply even among heterodox economists, but the level of engagement shown in these debates is an indication of the sense of importance of this classificatory issue.

Let us begin with a simple dictionary definition. Under the entry for

“neoclassical” in the New Palgrave Dictionary of Economics, the historian of economic thought Tony Aspromourgos (2008), a leading scholar on the historical origins of this term, gives us a first start with a historical definition of the term and its adoption.¹⁴ As noted earlier, the term was coined by Thorstein Veblen in

States. See for example the discussion between Barnett (2006) and Backhouse (Backhouse 2006) on the topic.

14 Aspromorgous’s 1986 essay is cited together with Fayazmaneh’s (1998) work by both Colander (2000:134) and Lawson (2013:2 n.3) as the two major in-depth studies on the history of the term.

a 1900 essay to characterise Alfred Marshall and Marshallian economics, that is, the main proponent in Britain of the new marginalist approach to economics that replaced the classical political economy from the 1870s onwards.¹⁵ The point of characterising Marshall’s school of thought as neoclassical was of course to establish a connection between Marshall and the classical economists of the earlier nineteenth century. But for Veblen, this connection was not based on marginalism, which was decidedly not present in classical economics, but in the

“alleged basis of a common utilitarian approach and the common assumptions of a hedonistic psychology” (Aspromourgos 2008).

The term achieved a more general meaning after the Second World War through separate articles by John Hicks and George Stigler in which “neoclassical”

came to denote not only Marshall’s thinking, but marginalist theory in general.

They both saw the unifying core of these theories in two assumptions: first their methodological individualism, and second their marginal productivity theory of distribution, resting upon a subjective theory of value. The term was widely adopted in the 1950s and 1960s, when it became well known in large part through the so called Cambridge capital controversy¹⁶. Aspromourgos (2008) writes that MIT economist Paul Samuelson’s influential textbook Economics (first edition published in 1948) was also influential in popularising the term with its aim to set forth a “grand neoclassical synthesis” from the 1955 edition on.

A second dictionary entry written by the historian of economic thought Roy Weintraub gives a similar view but with a slightly different emphasis. In his entry on “neoclassical economics” in the Concise Encyclopedia of Economics Weintraub (2002b), like Aspromourgos, points to the fundamental difference between classical and neoclassical economics. Whereas everyone agreed in the mid-nineteenth century about an objective “substance” theory of value where the value of a commodity is determined by its costs of production, and the share of value could be divided among the factors of production (land, labour, capital) corresponding to the main social classes, this is exactly what the marginalist revolution of the 1870s changed. The perceived problem with the classical Ricardian value theory was that the inherent value so calculated often differed from the actual market price when people was willing to pay more than the

“worth” of the commodity. The marginalists (Stanley Jevons, Leon Walras, Alfred Marshall and Carl Menger, though the latter was not a proper marginalist) all

15 In the overview of the history of economic thought at the beginning of this chapter, I used the term “marginal revolution” and called the associated authors “marginalists”, though many textbooks call this the “neoclassical” revolution and school. I avoided using the latter term altogether to prevent confusion.

16 For an overview of this episode in the history of economics, see Cohen and Harcourt (2003).

shifted their focus to the relationship between the commodity object and the subjective valuation of the buyer. These economists in different ways all started to think about the relationship between the costs of production on the “supply” side, and the subjective valuation on the “demand” side. Weintraub (2002b) straightforwardly claims that “the overarching theory that developed from these ideas came to be called neoclassical economics.” He argues that “The framework of neoclassical economics is easily summarized” as a set of axioms:

Neoclassical economics is what is called a metatheory. That is, it is a set of implicit rules or understandings for constructing satisfactory economic theories. It is a scientific research programme that generates economic theories. Its fundamental assumptions are not open to discussion in that they define the shared understandings of those who call themselves neoclassical economists, or economists without any adjective. Those fundamental assumptions include the following: 1. People have rational preferences among outcomes. 2. Individuals maximize utility and firms maximize profits. 3. People act independently on the basis of full and relevant information. Theories based on, or guided by, these assumptions are neoclassical theories. (Weintraub 2002b)

Weintraub claims that neoclassical economics became the mainstream of economics after the mid-century, existing in parallel with alternative but marginal schools of thought with their own metatheoretical frameworks for constructing economic theories. Here Weintraub lists Marxian, Austrian, post-Keynesian and Institutional economics as the main alternative schools.¹⁷ These schools, he emphasises,

Are regarded by mainstream neoclassical economists as defenders of lost causes or as kooks, misguided critics, and antiscientific oddballs. The status of non-neoclassical economists in the economics departments in English-speaking universities is similar to that of flat-earthers in geography departments: it is safer to voice such opinions after one has tenure, if at all. (Weintraub 2002b)

The brief explanation for this orthodoxy is “connected to the ‘scientifization’ and

‘mathematization’ of economics in the twentieth century” according to Weintraub. When the neoclassical research programme is increasingly associated

17 As a reference point, Colander et.al. (2014) list the same schools, with the addition of feminist economics. Lawson (2006:484) adds social economics in an article that seeks to elicit the common core of the different heterodox projects.

with “science”, any contestation of that paradigm will also be seen as a contestation of science as such.

Problematising and deepening the concept of neoclassical economics Lawson (2013) recently suggested that there are three main ways in which the term is employed in economics discourse. The two dictionary entries cited above are, in my view, good examples of two of these, and I think his distinction is helpful to better understand the debates about the term. First, Lawson mentions those who use it loosely and without elaboration, mostly as a pejorative term. The mere act of labelling something “neoclassical” is for many a sort of shorthand criticism. Lawson gives the example of Paul Krugman labelling Chicago

“freshwater” economics as neoclassical, while the post-Keynesian Steve Keen in turn criticises Krugman for being neoclassical. Even more common and also worse, Colander argues (2000:130), is its usage “in the discussions by lay people who object to some portion of modern economic thought. To them bad economics and neoclassical economics are synonymous terms.”

After dismissing this first loose use of “neoclassical”, Lawson finds two more approaches that make serious and systematic use of the term, chiefly among economic methodologists and historians of thought. The historical-comparative approach, is found among historians interested in the ways in which the term denotes simultaneous continuity and difference with an idea of “classical”

economics. However, all authors generally seem to conclude that no notion of continuity with something called “classical” economics holds any water. Instead, they claim that late nineteenth century marginalism should rather be labelled counter-classical (Maurice Dobb), non-, counter- or even anti-classical (Milan Zafirovski), and Joseph Schumpeter succinctly held that “there is no more sense in calling the Jevons-Menger-Walras theory neoclassic than there would be calling the Einsteinian theory neo-Newtonian” (Lawson 2013:2 n.3). Lawson also concludes that Tony Aspromorgous (cited above) and Sasan Fayamanesh, who have both studied the spread of the term since its origin in Veblen’s 1900 essay, fail to find any proper grounds for continuity in Veblen himself.

Lawson himself provides an in-depth reading of Veblen’s essays to provide his own, quite surprising, interpretation of what the term should mean, only to argue that even with that interpretation, the term is hardly useful. An important point, according to Lawson, is that when Veblen compared Marshall with the “classics”, his conception differs from the term “classical political economy” originally

coined by Marx (1970b).¹⁸ Whereas Marx used the term in contrast with the superficial “vulgar economists” following Ricardo, in Veblen’s usage the “classics”

are Marx’s “vulgar economists” (Lawson 2013:15). This usage has been followed by others, notably Keynes, and Colander (2000:131) even argues that Keynes lumped his predecessors together in the category “classicals”, that is, pre-Keynesian economics as contrasted with pre-Keynesian economics. It should be evident from this example that even the seemingly basic categories of intellectual history are shifting, and to add to the confusion, these shifts in terminology have not always been properly acknowledged even by historians of thought.

Weintraub’s entry in the Concise Encyclopedia of Economics is an example of Lawson’s third approach, which could be thought of as substantive or ontological.

This approach is taken by a number of authors who try to systematise a coherent account of the core analytical features that characterise this school of thought.

These accounts all seem to share a certain abstract nature, focusing on an abstract metatheory, set of axioms, or even meta-axioms underlying substantial theories.

Lawson argues that most of these accounts seem to agree on some fundamentals:

among the common identified neoclassical axioms is methodological individualism (individuals as units of analysis). Some form of assumption about typical behaviour (often, but not necessarily, a classical conception of self-interested rationality) is normally also included, and often, but not always, equilibrium analysis (Lawson 2013:3). Weintraub’s short definition of the metatheory mentioned above, for example, does not say anything at all about assumptions about equilibrium, or even the use of equilibrium analysis. In contrast, Frank Hahn, a self-identified

“neoclassical” economist, lists the following essential features of neoclassicism:

“(1) an individualistic perspective, a requirement that explanations be couched solely in terms of individuals, (2) an acceptance of some rationality axiom; and (3) a commitment to the study of equilibrium states” (Lawson 2013:3 n.4).¹⁹

18 Lawson (2013:13) analyses Veblen’s interest and purpose for coining the term in-depth. In his view, Veblen is interested in the metaphysical presumptions of economics and especially two competing ultimate “grounds for finality in science”. He thereby wants to contrast the older

“taxonomic” approach which basically compares matters of fact with a normal or ideal state, in late nineteenth century economics expressed as equilibrium analysis. In contrast to this, Veblen sees an emerging Darwinian evolutionary and causal approach to economics. Veblen finds that Marshall (as opposed to the earlier marginalists) realises the need for the latter approach, but cannot let go of the “classical” preoccupation with “taxonomy”, i.e. equilibrium analysis. Veblen identifies an essential tension in Marshall’s work, which justifies the label “neoclassical”, and which he uses interchangeably with terms like “quasi-classical” or “modernised” economics (2013:19).

19 Lawson notes that the “commitment to the study of equilibrium states” does not imply an assumption about equilibrium actually existing in any sense.

Another example of this approach is Christian Arnsperger and Yanis Varoufakis’ (2006) account of neoclassical economics as based on three meta-axioms, as an argument against the thesis of a new mainstream pluralism.²⁰ Davis (2006) as well as Colander et al. (2004) claim that any list of necessary core features of neoclassical economics will soon encounter a modern approach that breaches one or more of the supposedly necessary core criteria, thus rendering any attempt to define the modern mainstream as “neoclassical” futile. In their view, this shows that there is no such thing as a coherent neoclassical research programme, or at least that the contemporary mainstream cannot reasonably be characterised in this way. Against this thesis of a new mainstream pluralism, Arnsperger and Varoufakis argue that the coherence of the research programme should be sought at a higher level of abstraction. Instead of listing necessary conditions, core ideas that must be held true, they claim that behind seemingly different ideas about, for example, rationality (selfishness, altruism, bounded rationality, and so on), we can nevertheless find a minimum common ground of modern mainstream economics. It consists of what they call three meta-axioms.

The first is methodological individualism, which the authors claim underlies the neoclassical school since the marginal revolution, but not the classics, nor Keynes or Hayek. The second meta-axiom is methodological instrumentalism. This means that “all behaviour is preference-driven or, more precisely, it is to be understood as a means for maximizing preference-satisfaction” (Arnsperger and Varoufakis 2006). The standard view holds that all behaviour is fully determined by a given set of preferences. But even in later developments of evolutionary game theory, where preferences are modelled as developing as dependent on past outcomes, or expectations about others’ expectations, the meta-axiom still holds true: “homo economicus is still exclusively motivated by a fierce means-ends instrumentalism”.

In more familiar sociological terms, this amounts to saying that all human action should be understood in terms of Weberian instrumental rationality. The third meta-axiom is methodological equilibration, in which the question is posed of what behaviour should be expected in a state of equilibrium. The point is that the questions about whether equilibrium is probable or even possible, and if so, how it can come about, are left behind in favour of the theoretical study of presumed behaviour, even when no demonstration of the actual emergence of equilibrium is provided.

The common feature of all these approaches is that they attempt to define their object in terms of substantial assumptions. Since these assumptions are rather

20 The approach of Arnsperger and Varoufakis has been adopted by others, for example in the Swedish context by Lars Pålsson Syll (2013).