The mainstream view of economics as a science

It is well established now that mainstream economics evolved by emulating physics (Reference MirowskiMirowski, 1989). The aim was to develop theories which were both internally logically consistent and testable against empirical evidence. According to logical positivism, such an aim marks out science from non-science. Indeed prior to the emergence of logical positivism in the 1930s, the marginalist methodological approach had started the shift of economics away from the class-based framework of classical political economy towards a focus on the supposedly universal optimising behaviour of individuals within competitive markets. Marginalism required and thus promoted the increasing use of formal mathematics to such an extent that the deductivist formal mathematical method for analysing choice has come to define the discipline as far as mainstream economists are concerned. The scientific nature of economics is seen as being completed by empirical testing.

Explicit references to methodology are in fact rare in mainstream economics, such that methodological stance has often to be inferred from practice. But mainstream introductory textbooks do provide illustrative statements about the nature of economics and its methodology. Reference Mankiw and TaylorMankiw and Taylor (2006) and Reference Parkin, Powell and MatthewsParkin et al. (2012) make typical statements, introducing students to economics by specifying the subject matter in terms of rational choice under scarcity. They then briefly set out economic methodology (what Reference Mankiw and TaylorMankiw and Taylor, 2006, call ‘the economist’s way of thinking’ (p. 19)). Theory is equated with mathematical models which they assert to be simplifications of reality, tested against evidence: that is, they make a clear statement of logical positivist methodology. Here, we have a particular understanding of economics, and its methodology presented as the definitive understanding for all of economics. But it involves huge methodological presumptions. If there are (unacknowledged) alternatives, the implication is that they must be less scientific.

But this view of scientific method has long been discredited in the philosophy of science literature. All three recent accounts of the methodology literature attest to this and therefore devote more attention to the ways in which the field has moved on (see Reference Boumans and DavisBoumans and Davis, 2010; Reference DowDow, 2002; Reference HandsHands, 2001). All three accounts explain how logical positivism suffers from the Duhem–Quine problem, which refers to the difficulty of identifying what exactly has been falsified by contrary evidence. Finding evidence consistent with a theory does not prove it to be correct. Reference PopperPopper (1963) had argued that scientists should instead test very narrow propositions and reject them if contradicted by the evidence. But Reference BlaugBlaug (1980) demonstrated that, where economists test theories, they persist in relying on confirmatory evidence. Thus, even among those who share a positivist approach to economics, differences of opinion on austerity policies have not been resolved.

Because of the difficulty in arriving at definitive conclusions on the basis of empirical evidence, the higher ground of mainstream economics has been held by pure theorists whose theories produce definitive conclusions, regarded only as testable in principle. Thus, while there has been an increasing challenge from evidence arising from applied fields like behavioural economics, the aim of the challengers is stated as being to build a better formal framework (Reference DowDow, 2013). Reference ShillerShiller (2013) reiterates the compatibility between formalist theorising and experimental evidence, within a logical positivist view of science as demarcated from non-science (‘charlatanism’). But the view, that for economics to be a science theory must be tested empirically, has continued to pose a challenge for mainstream economics. Such a challenge is endemic to logical positivism.

According to this ‘scientific’ (indeed scientistic ^{Footnote 3} ) approach to economics, theory appraisal and policy prescription are matters of logic and evidence; there is no place for belief as far as economists are concerned. Were economists’ beliefs to be analysed, they would be constrained to converge on the correct model, identified on logical positivist grounds. Beliefs do feature in agents’ expectations in mainstream models, reflecting the subjectivity of expectations and the learning required to adjust expectations in the light of evidence. Agents’ beliefs are particularly important with respect to the behaviour of state agencies; central banks are enjoined to behave in a consistent way which enhances credibility and thus the correctness of beliefs. But beliefs are constrained eventually to settle on the final equilibrium values which are identified by technical analysis, based on deductivist logic tested against empirical evidence.

Not only has philosophy of science moved on, so have the physical sciences. Reference BuchananBuchanan (2013) argues that the physical sciences have shifted their focus from pure theory to attempting to understand real disequilibrium systems with positive feedback mechanisms; the nature of the subject matter is determining the type of theory that is being developed and the type of mathematics used. He contrasts this development with the continuing dominance of mainstream economics, whose deductivist equilibrium theoretical system bears little relation to the real subject matter. Reference ShillerShiller (2013) too draws attention to developments in the physical sciences in the direction of becoming more applied, using this tendency as evidence of economics’ continuing scientific status. Yet he does not challenge the identification of mainstream theory with mathematical models: ‘The challenge has been to combine its mathematical insights with the kinds of adjustments that are needed to make its models fit the economy’s irreducibly human element’. This is not the fundamental shift from axiomatic deductivist equilibrium theory to the study of disequilibrium positive feedback processes we find occurring in the physical sciences, but rather a shifting balance in the continually uneasy relations between theory and evidence which characterise mainstream economic methodology.

It is the economist’s task, according to Reference LucasLucas (1980), to build ever-better analogue models, where models are again equated with theories. Since formal models are closed systems, this means that mainstream theories too are closed systems. Although the crisis provided overwhelming evidence that the economic system was not closed to unexpected structural shifts, this evidence has not been taken by the mainstream as a challenge to the closed-system approach. Much of the discourse responding to the challenge posed by the crisis was conducted, not in terms of rethinking the reliance on formal models, but in terms of finding a better technical model (Reference LawsonLawson, 2009).

But, in spite of all these problems with the unworkability of logical positivism and the fact that physics has moved on to systems very different from economics, arguments in favour of austerity policies are still presented as technical propositions (see Reference Ball and MankiwBall and Mankiw, 1996, for a typical statement). The underlying argument for austerity is indeed a technical result of the assumptions and structure of mainstream models, but these in turn are built on a range of metaphysical and methodological assumptions. Arguments either for or against austerity are not purely technical.

Yet, the assumption of rational expectations (or some variant of them) portrays agents as arriving at the same technical conclusions as economists. Indeed, the argument for austerity has increasingly been expressed in terms of the assumption that financial markets have accepted the technical result that deficits crowd out private sector expenditure and reduce growth, such that they will factor it into the pricing of assets, including government debt, constraining the government to accept an austerity stance. This thinking was embedded in the Maastricht rules, limiting the size of deficits and public sector debt relative to gross domestic product (GDP) in order for fiscal policy not to interfere with the priority of controlling inflation through independent monetary policy. The force of the pro-austerity reasoning appeared to be strengthened by the research findings of Reference Reinhart and RogoffReinhart and Rogoff (2010), whose historical evidence of budgetary positions and growth rates appeared to confirm the technical results. Reason and evidence are assumed to be sufficient for generating and accepting the austerity conclusion. Beliefs are seen only to enter into the policy arena once politicians input their political preferences (Reference ColanderColander, 2002). The presumption is that the theoretical structure is purely technical, yielding definitive ‘scientific’ results.

Yet, even physical scientists talk in terms of belief, as in belief in global warming. This is a clear indication that there is too much uncertainty surrounding climate change to be able to demonstrate definitively that global warming is occurring. The physical world, like the social world, is an evolving, complex, open system. ^{Footnote 4} Reason and evidence support the belief in global warming, but others find support for alternative beliefs. While promoters of global warming regard their critics as wrong-headed, nevertheless the matter is treated as one of belief, requiring persuasion to change from one belief to another. Pure deductive reasoning, built on premises which all agree to be true, is not enough; any deductive chain is vulnerable to premises not being reliably true. Pure inductive reasoning, extrapolating from past observations, is not enough; there is no guarantee that past structures and relationships will continue as before. Even in the physical sciences, then, arguments require belief – that premises hold true, or that structures and/or relationships will continue to hold – even when this cannot be demonstrated definitively. Knowledge is in general uncertain.

Knowledge and belief

The inadequacies of logical positivism and the centrality of belief for knowledge were established at the beginning of modern economics in the 18th century. Reference Hume, Selby-Bigge and NedditchHume ([1739–1740] 1978) grappled with Cartesian deductivism and also specified the problem of induction, concluding that knowledge rests on belief. As far as deductivism was concerned, no system based on pure reason could legitimately be applied to real circumstances. As far as empirical testing on the basis of evidence from the past was concerned, there could be no basis for assuming that the same causal mechanisms would operate, or operate in the same way, in the future. Reason and evidence were necessary for knowledge, but not sufficient: they relied in turn on belief.

This belief applies at a variety of levels, starting with a conventional belief in existence and in the nature of the real world but also incorporating belief that some theoretical propositions are reliable, on which to build other propositions. Hume, like Smith, adopted a Newtonian approach, whereby knowledge is built up through a process of abduction. Any resulting theory must be regarded as provisional since its premises may not apply to other times and places than those from which evidence is drawn. Even as explanation rather than prediction, a theory is provisional given the reliance on belief at various stages of reasoning. Knowledge is more reliable the more it draws on experience and on reason – the philosopher (scientist) is someone drawn to applying more reason than others, and thus to challenging conventional beliefs. But knowledge is always subject to some degree of uncertainty.

Reference KeynesKeynes (1921) pursued Hume’s agenda by exploring how we establish grounds for belief under the normal condition of uncertainty, where uncertainty eludes the quantification which allows mainstream theorists to generate what purport to be demonstrable propositions. Keynes understood establishing grounds for belief as being an objective process, drawing on evidence and reason; anyone in the same circumstances, environment and psychological state would arrive at the same belief. But circumstances, environment and psychological state differ, such that, not only does the evidence brought to bear differ, but so do the conventional beliefs and intuition employed and also what Keynes called ‘animal spirits’. Beliefs therefore differ. Beliefs can be justified up to a point by reason and evidence. Thus, the more different types of evidence support any belief, the greater the evidential weight. But what is admissible as evidence depends on judgement as to its relevance which depends on the underlying theoretical perspective and ultimately on belief (Reference Dow, Dow and HillardDow, [1995] 2012).

Critics of austerity policies have marshalled a compelling range of arguments challenging both the deductive and empirical elements of the case for austerity. For example, Reference Chick and PettiforChick and Pettifor (2010) provide empirical evidence from a range of historical episodes in the UK which contradicts the mainstream arguments for austerity. They show in general that fiscal expansion has been associated with falling public sector debt, higher growth and lower interest rates. A key argument is that austerity cannot achieve its object of reducing public sector debt because of automatic stabilisers: austerity causes revenues to fall and expenditure to increase. Increasing expenditures on the other hand, in underemployment conditions, have a multiplier effect on incomes and thus on revenues, while reducing reliance on unemployment benefits and spending on social services. Similarly, Reference Herndon, Ash and PollinHerndon et al. (2014) provide a detailed critique of the evidence on the relationship between deficits and growth behind the Reference Reinhart and RogoffReinhart and Rogoff (2010) thesis. These are just two examples of what would appear to be powerful direct challenges to the arguments for austerity on the same ground of reason and evidence. So why does the argument for austerity still persist?

While apparently misguided (even delusional) beliefs can persist, it is also possible for them to change in the light of evidence, particularly when it has a direct effect on experience. Thus, for example, it was the initial experience of defaults which punctured the belief that financial markets had reliably priced risk and protected themselves against default, setting off the crisis. Mainstream theorists too have responded to the experience of the crisis by adapting theory to incorporate a range of constraints on the free operation of markets and on rational optimising behaviour. But, although it would seem that pointing out false logic and contrary evidence should change attitudes to austerity policies, this has been true only to a limited extent. Belief in austerity policies is necessarily based on more than reason and evidence; these are filtered through the underlying structure of beliefs.

At the level of belief in the nature of the economy, the arguments for austerity are all predicated on economic structure being unchanged. But the crisis and its real consequences provide ample evidence of the scope for the structure of social systems and behaviour within them to change in a non-deterministic way, that is, for the subject matter to be an open system. Thus, an important aspect of the Reference Chick and PettiforChick and Pettifor (2010) argument is the reference to changes in the policy-making architecture (particularly with respect to monetary policy) in the periods under study. Rather than separable deductive reasoning tested against evidence, their argument involves abduction, whereby observed structural changes alter the interpretation of the data and the reasoned argument.

And yet the mainstream persists in adopting a closed-system approach which implies a closed-system subject matter (Reference LawsonLawson, 1997). It is this approach which allows conclusions to be treated as reliable when they rest on deductive argument applied to an unchanging structure or on extrapolation from empirical evidence. Mainstream arguments in favour of austerity are only ‘technical’, independent of belief, within such a system. Yet, the system itself rests on the belief that the social system can be treated as if it were closed. The disparity between an open-system reality and closed-system theory lies behind the difficulties with empirical testing, notably the problem of induction. The statistical version is that tests can only refer to the past and cannot address uncertainty with respect to the future. Hume’s more fundamental problem is that causal mechanisms and their interrelations are too complex for us to fully understand them, far less predict their outcomes.

But closed-system theories yield definitive conclusions. While these conclusions cannot be related directly to evidence, the resulting certainty can be more persuasive to non-economists than the provisional conclusions of open-system theory. This is particularly the case with equilibrium theory where any dynamism en route to the final equilibrium occurs in mechanical time, that is, it does not correspond to historical time. As Reference Chick, Dow and HillardChick (1995: 33–34) points out, the closed system puts the focus on a fictional long run, while knowledge about a real historical long period is fundamentally uncertain. She quotes Reference RokeachRokeach (1960) as follows: ‘Knowledge about the remote future is impossible to refute and, hence, one can be safely preoccupied with it … For this reason, a narrow, future-oriented time perspective is … seen to be a defining characteristic of closed systems’ (p. 64).

Because reason therefore has limited scope, efforts to change others’ beliefs are a matter of persuasion rather than pure reason. As Reference KeynesKeynes ([1934] 1973, emphasis in original) put it,

Reason and evidence are sufficient for argument according to mainstream economics, at least as far as the official discourse goes. (Reference McCloskeyMcCloskey, 1986, demonstrates that the unofficial rhetoric of the coffee room is – necessarily according to our argument – more pluralistic.) But reason and evidence are not even in principle sufficient from a Keynesian perspective. Thus, for example, while Joan Robinson demonstrated a logical inconsistency in mainstream theory in the Capital Controversies (an exercise in pure reason), the principle of factor substitution was not rejected; pure reason was insufficiently persuasive. On the other hand, the Monetarist/Keynesian debates about the relative slopes of the IS and LM curves remained unresolved in spite of extensive empirical work; empirical testing was insufficiently persuasive.

Argument rather needs to address the complex structure of knowledge, encompassing also the nature and role of beliefs within that structure. In Kuhnian terms, beliefs underpin the worldview of any paradigm; in critical realist terms, beliefs about the nature of reality (ontology) determine epistemology. But since theory develops by abduction, involving a going-back-and-forward between experience and theory, beliefs continue to penetrate all subsequent reasoning, including the interpretation of evidence. We have focused so far on belief in the epistemological sense. Indeed, much of the critiques of austerity draw on the large literature which spells out the foundation of orthodoxy, like heterodoxy, in beliefs as to what constitutes reliable knowledge. There is no basis in modern philosophy of science for mainstream economists to claim that their research is more scientific than that of heterodox economists because of their reliance on formal deductive mathematical reasoning and their appeal to what are presented as independent facts. The grounding in undemonstrable beliefs is there, unacknowledged or not.

It is challenging enough to get this argument across to fellow economists. But for persuading public opinion more generally to challenge the mainstream results, these epistemological arguments may have limited traction. What may have more purchase is the argument that, among these beliefs, all economic reasoning involves moral judgements, as Hume had contended (Reference KayatekinKayatekin, 2014). Thus, for example, arguments for austerity prioritise efficiency over distributional concerns. But the moral judgements are even more deep-rooted. Where it is assumed that workers are paid the value of their marginal product (VMP), this is taken as a technical result without moral content. Yet, the suggestion that government should support the market in paying CEOs of banks and low-paid workers, their market value entails a moral judgement in favour of forces for a particular distribution of income. The moral content is more evident in alternative approaches which understand pay levels to be politically and socially determined. But this is taken by mainstream economics to be an indication of poor science – reasoning invaded by ideology.

What is widely ignored is that this is in the nature of a social science and that mainstream economics incorporates ideology. Reference StrettonStretton (1969) sets out the argument that all theorising inevitably incorporates values, in the selection of the focus for theory and the selection of causal forces identified. The belief that the market produces a morally acceptable distribution of income has of course been widely challenged as a result of austerity policies. Yet, the presentation of mainstream arguments for austerity as a technical result has been persuasive that moral arguments about income distribution are separable from the content of mainstream economic theory. But Stretton points out that the mainstream methodological stricture to select for generality itself involves a value judgement. Rhetorically, the misguided conventional view that mainstream economics is the only scientific approach further diverts attention from any discussion among economists of belief on the grounds that any such discussion would not be ‘scientific’. It also diverts the moral argument in public discourse away from mainstream economics. Stretton points out the immorality of concealing the values embedded in theorising which has the consequence of stifling debate over these values.

The fact is that some beliefs hold more sway over the policy process than others. In the next section, we consider the process of persuasion which ensures that some beliefs become conventional in spite of the fact that critics of austerity policies regard the basis of these beliefs in reason and evidence as very weak.

Rhetoric, power and belief

If conventional beliefs spread through persuasion, then we need to consider further why the rhetoric of austerity has been so successful. We consider three elements in this process of persuasion: the content of the rhetoric, the means by which it is communicated and the relative power of those attempting to persuade.

It was noted at the start that the primary forces behind austerity policies were ideological (the aim of reducing the role of the state) and self-serving (the aim to protect capitalist economic interests). These forces are associated with socio-economic power and thus the capacity to influence what is regarded as reliable knowledge. There is a range of mechanisms for exercising that power. For academic economists, it is the editorial policies of leading mainstream journals, research funding and the hiring process in universities. These mechanisms operate by ‘peer review’, where the peers predominantly promote mainstream economics as the most ‘scientific’ approach to economics. That governments actively promote this kind of peer review reflects their tacit acceptance of the mainstream closed-system view of knowledge. Since governments have the power to challenge such a view, the fact that they don’t implies that it is not in their interests to do so. Even where political interests would not appear to be best served by mainstream economics, as is the case particularly for left-wing politics, the power of financial markets over governments can be seen as an effective constraint. Globalisation and financialisation have increased the power over governments of financial markets and their trading in sovereign debt.

For the general public, the mechanisms are different, relying heavily on power exercised over various communications media. A growing body of work in discourse analysis is building up a picture of communication strategies in persuasion with respect to government policies. Reference PigeonPigeon (2008) provides a rich account of how the Canadian government used a variety of communication strategies to great effect to produce conventional belief in the wisdom of inflation targeting and fiscal austerity. Reference MercilleMercille (2014) concentrates on the role of the mass media in Ireland in promoting a conventional neo-liberal view in support of fiscal consolidation policies. Reference Herndon, Ash and PollinHerndon et al. (2014) discuss how the Reference Reinhart and RogoffReinhart and Rogoff (2010) evidence had been used in the policy discourse. The language used itself can be very effective in persuading the general public to accept ‘expert’ opinion. Reference GaborGabor (2010) emphasises the importance in Romania of the presentation of macroeconomic policy in technical, apparently non-political, terms, allowing the government to pursue a neo-liberal agenda. Her work is path-breaking in demonstrating the power of presenting political economy arguments as neutral technical results, an argument on which this article has attempted to build.

Reference Smith and WightmanSmith ([1795] 1980) had emphasised the role of rhetoric in persuasion, given the impossibility in general of fully demonstrating a proposition by reason or evidence. Metaphor performs a powerful role in the communication of theories (Reference McCloskeyMcCloskey, 1986), but also in their origination and development (Reference LawsonLawson, 2003). Indeed, Reference Smith and WightmanSmith ([1795] 1980) had likened theories themselves to imaginary machines. In economics, metaphors such as ‘the market’ have been very powerful in guiding thought. With respect to fiscal policy, the most powerful metaphor employed by politicians (though not generally by economists) has involved likening a national budget to a household budget; fiscal austerity is presented as normal prudence. Such a device is persuasive, as Reference Smith and BryceSmith ([1762–1763] 1983) suggested, because it builds on personal experience (unlike proper macroeconomic arguments). Households are then susceptible to the argument that, once debt reaches a particular level at a particular interest rate, debt servicing will make it increasingly difficult to control the budget. A budgetary surplus is therefore seen to be required to reduce both the debt and the debt service burden; it is perceived that the way to achieve this is to implement austerity policies.

While this may seem like a purely technical argument, it rests on the belief that a government is like a household. But, even setting aside the confusion in the public discourse between current and capital deficits, and the different capacities of governments and households to finance a deficit, a current deficit at the national level is not like a current deficit at the household level. When a household is earning from employment less than it is spending on current consumption, there is no mechanism to avoid ever-increasing debt other than to cut back on spending. A government’s earnings are primarily from taxation, which depends on the employment and spending of the population. Cutting back on government spending therefore reduces tax revenue and requires increased spending on social services, making the budgetary situation worse. It is not just government capital spending which can improve long-run fiscal prospects, as can investment in human capital by households. The point is that government current spending can have multiplier effects on incomes and thus on the fiscal position; there is no such equivalent for households. The fallacy is to use a microeconomic metaphor, tapping into the public’s personal experience, for a macroeconomic problem.

A further rhetorical device is to make the recipients of austerity feel responsible for their plight. It is a clear implication of mainstream economics that (barring market imperfections) workers are paid low wages because of the low value of what they produce, or are capable of producing. Similarly, unemployment is portrayed as an unwillingness to search adequately, move location, acquire appropriate skills, or accept the lower wage appropriate to that worker’s VMP. Vulnerable household borrowers, further, are portrayed as having been greedy and profligate, when in fact banks had had a duty of care not to make loans without a reasonable expectation that they could be serviced. Of course, this can be explained to some extent by banks’ buying into the expectation of continued asset price rises purveyed by ‘technical’ economics and finance experts, such that they had not appreciated the risks for them or their clients.

Psychology clearly plays an important role in rhetoric, as in cognition more generally. Reference Smith and BryceSmith ([1762–1763] 1983) pointed to the importance of bearing in mind the beliefs and prior knowledge of the audience. This explains the effectiveness of the household metaphor. Smith also stressed the importance for successful rhetoric of appealing to what the audience would find psychologically satisfying (given that knowledge is sought to calm any sense of discomfort at unexplained events; see further Reference Smith and WightmanSmith [1795] 1980). He was fully aware of the aesthetic attractions of abstract argument, however, misleading it may be:

Furthermore, Smith distinguished between the natural and social sciences in terms of how far, on the strength of their own specialist expertise, they could sustain their theories in the face of contrary arguments. He maintained that the natural sciences were protected from much public scrutiny because their theories are far removed from normal experience:

But as far as moral philosophy, and thus the social sciences, was concerned, he predicted that they would be held in check by the fact that the public understood the subject matter:

The device of mimicking the natural sciences and mathematics has allowed mainstream economics a degree of immunity from critique which Smith had not thought possible for economics. A key element of that device has been to present economic results as independent from moral sentiments.