INSTITUTIONAL ECONOMICS: ECONOMIC METHODOLOGY – CRITICAL REALISM

We shall examine the nature of the social and economic world, highlighting the positions of critical realism and positivism. Then we shall compare the different types of economic methodology, explaining why critical realism is more preferable than to the others. We shall also investigate the nature of econometrics and its limitations. Finally, we will examine other details of critical realism.

The social world consists of three different layers of social objects (see Table 1). Structures, irreducible to other objects, usually cannot be observed, but whose effects can be felt. Actual objects are typically observable and measurable, and are explainable in terms of their underlying structures. Empirical objects refers to our senses and impressions of the actual objects; i.e. our interpretation and imagination of the actual e.g. half-full or half-empty glass of water

Table 1: Different layers of objects in the social world

Critical realism (an approach to economic methodology favoured by old institutional economists, Marxist economists, feminist economists and some Austrian economists) recognises the nature of the social world as structured into different layers of social objects. For critical realists, the aim is to uncover and explain the actual and empirical objects (or manifest phenomena) in terms of their ‘deeper’ (or real), underlying structures, and to reveal more clearly the operation of mechanisms (i.e. the connections between the actual and structures). This analytical process is called retroductive inference (from manifest phenomena to structures).

Positivism (an approach favoured by neoclassical economists and many new institutional economists) only focuses on actual entities and the empirical objects – it refuses to acknowledge that there are unobservable objects that have real effects. For positivists, the aim is to objectively measure and quantify observable ‘facts’ and put into a numerical form, and search for and predict empirical regularities of actual events, or constant conjunctions. Its main analytical reasoning is called deductive inference (from general laws to particular predictions).

I will suggest that critical realism is superior to positivism in terms of its conception of science and social world. Positivism is both useless and false, having an inadequate conception of the science and the natural and social world.

Methodology denotes the study of method, an activity concerned with the procedures and aims of a particular discipline. The term methodology has a narrow and a broad sense of meaning. The narrow sense refers to techniques applied to specific analyses (especially in econometrics and modelling). The broader sense refers to the philosophy of science as applied to economic issues – it focuses and elaborates on ontological issues – i.e. questions of the nature of being.

Methodology is important to economists, because the task of methodology (or philosophy of science) is to act as an under-labourer of science, facilitating a set of perspectives on the nature of the economy and society, and how to understand them. Methodology affects how economists approach their science – i.e., what questions to ask (of deep structures or of events and impressions), and how to conduct economic research (e.g. a large statistical sample size or a small qualitative set interviews).

Traditionally, there are two perspectives of economic methodology.

• The naturalistic / positivistic tradition views science as an activity grounded in positivist principles. It is concerned to identify, or to test or refute claims about empirical invariance between discrete events or states of affairs. Rooted in Humean notion of laws and causality, it dismisses the possibility of any ontology of deeper structures and the possibility of an independent existence of things or the operation of natural necessity. Indeed, all we have are impressions. Experiences of atomistic events and their conjunctions exhaust our knowledge of nature. The natural (or positivist) tradition is identified with the search for constant conjunctions of events. In econometric realm, any methodological focus is almost exclusively upon the comparison of competing procedures for estimating the parameters of empirical regularities, or methods of comparing two or more such empirical relations. This positivist perspective has been the most dominant influence in economics.
• The hermeneutic / subjectivist tradition contends that social science is, or ought to be, concerned with matters of elucidation of meaning, the tracing of conceptual connections, constructing social wholes (collective entities) out of individual opinions, beliefs and actions - i.e., methodological individualism. This perspective is associated with mainstream Austrian economics (e.g. Hayek and von Mises).

However, there is an alternative: critical realism. Realism rejects the view of science as concerned with seeking out, or with testing claims formulated as, empirical regularities. According to critical realism the world is structured, open and intransitive.

• The world is structured in the sense that underlying manifest phenomena at any one level are irreducible structures, powers, mechanisms and necessary relations that govern them. For example, firms’ actions (manifest phenomena – i.e., empirically observable events and actual objects) are governed by structures such as property rights, habits, routines, competition, creativity, labour relations, state regulations, business networks and so on These structures cannot be reduced to individual manifest events, and are, in fact, collective entities.
• The world is open in the sense that manifest phenomena are typically governed by various countervailing mechanisms simultaneously, so that the deeper structures can rarely be directly ‘read off’. For example, a consumer goods market can be structured by the need to create novelty, fashion and innovation, but simultaneously by countervailing set of mechanisms such as tradition, loyalty, habits and routines. The purchase of the commodity can not be attributed (or ‘read off’) from one or either set of mechanisms. The dominant mechanism is ascertained through concrete analysis of people’s purchases.
• The world is intransitive (i.e., ontological world - the existence of objects and beings - is independent of its identification by social agents). The social world cannot be reduced to epistemological issues (i.e., the knowledge of objects and being – the social construction of objects and beings, and concept-dependent identification of objects). The social objects and events are socially constructed, but are not determined by concepts. In other words, the social reality exists and acts independently of agents’ conceptions. Furthermore, because the social world exists independently of agents’ conceptions, the social world can be inadequately or falsely conceptualised by social agents, and thus prone to rational conceptual critique. For example, causes of unemployment, such as exchange rate movements and industrial uncompetitiveness, exist irrespective of their conception by unemployed people. Furthermore, an unemployed person may falsely and wrongly attribute his lack of work due to immigration (‘too many blacks’) and/ or women working (part-time jobs) rather than to underlying structures of unemployment.

According to critical realism, it is essential to move from manifest phenomena to the deeper (real) structures and relations that govern them. Note that the essential mode of inference required here is not induction (particular to general) nor deduction (general to particular) but retroduction (manifest phenomena to deeper conditioning structures). For example, the observation that a few small firms do not undertake much research and development (R&D), can lead to the inductive inference that all small firms will undertake little innovation. Or starting with the claim that all small firm undertake little R&D can lead to the deductive inference that the next small firm must carry little innovation. But starting with the observation of one or more small firms undertaking little R&D, the task is to identify the mechanisms intrinsic to small firms which dispose them to carry out little innovation.

To re-cap, critical realism aims to reveal the unacknowledged conditions, tacit skills and unconscious motivations that are necessary (and never sufficient) for some activity to take place, as well as to identify the unintended consequences of such activity. And being potentially revealing and enlightening, critical realism also lays the basis for emancipatory development – rather than mere small changes in the state of affairs (for more details see the section ‘Critical Realism).

To repeat, methodology is essentially an underlabourer for economic science, being concerned to reveal the nature of the social world, and on how to understand it. In this sense, methodology (particularly critical realism) can certainly make a difference to the conduct of economic science, always remembering that it constitutes at most a necessary (and never a sufficient) condition for enlightened change and development.

What are the possible limits within which econometrics can be legitimately applied?

Broadly speaking, econometricians conceive of themselves as attempting to provide quantitative content to hypothesised relationship connecting economic variables (or events). This quantitative aspect relates primarily to parameter values of such relationships. Central to econometric analysis is the attempt to identify and quantify empirical regularities. Mainstream economists argue that the world is constituted by empirical regularities. It is this premise that is questioned.

The econometric analysis is only operative in closed and controlled social world, where the conditions are fixed and given (i.e. ceteris paribus). For example, econometricians will analyse the effects of number of firms in a product market on prices and profit, predicting that a large number of firms would reduce price and profit levels to a minimum, and proving the competitive nature of a large number of firms. There are a number of important conditions required for this prediction and analysis to work – one of which is to assume no innovation or creative process. In effect, this prediction works only in a closed and controlled social world.

Yet, this is an illegitimate premise of the nature of the social nature. In reality, the social world is open, reflecting the concepts of human choice, intentionality, purpose and creativity. Accepting that agents can act otherwise implies that the world is open to many possibilities. The concepts of choice and human agency are real so that the world must be open and changeable. In this respect, any possible economic law is only a tendency and not an empirical regularity – in short, the task of making predictions will always fail, and indeed, is useless and false. In other words, for an event to act the same way on different occasions (i.e. empirical regularities) the social world must be constant, given and fixed. Yet, clearly this condition fails to hold in the social world where human agency is real, so that choice, intentionality and creativity mean that the social world is transformed.

The legitimate scope of econometrics is limited to situations that are closed. In open systems that constitute the economy and society, econometric relationships breakdown. Consequently, the realist conception of the economic world as structured and open means that economic science must be explanatory, rather than predictive.

A legitimate scope for econometrics would be summary statistics of a body of data, such as means and growth. While this type of econometrics is not contentious, the majority of econometric work is characterised by the search for empirical regularities.

Critical realism is theory of being (i.e. a theory of ontology) not primarily a theory about the knowledge of that being (although it entails epistemological implications). Critical realism explicitly acknowledges the two realms -–that of object of analysis and that of knowledge about the object – in other words, intransitive and transitive dimensions respectively. It is a condition of the possibility of science that objects exist and act independently of their identification. (Indeed, to be a fallibilist about knowledge is to be a realist about its objects.)

Critical realism can lead to certain insights. As already mentioned, the social world is structured, open and intransitive. For example, the output of the firm is necessarily governed by labour relations, market competition, intuition and hunches, demand expectations, norms and technological innovation and so on. These structures can work against each other. Some of these structures may operate without the businessowner’s acknowledgement of them, or are falsely conceived. What emerges at the manifestation level (e.g. output) arise from complex and plurality of underlying structures.

Critical realism does not only identify the nature of major constituents (structures, mechanisms, tendencies and powers) of the social realm, but also elaborates on the nature of the connection between human agency and social structure. In opposition to individualism and collectivism, critical realism emphasises an essentially relational conception of the social world. And in opposition to structuralism and subjectivism, critical realists elaborate a transformational conception of the agency/structure relation.

• The relational conception of the social refers to how all forms, structures and systems (e.g. economy, the state, organisations, trade unions, and households) depend upon, or presuppose, social relations. For instance, the existence of a firm presupposes capitalist product and labour markets and property rights– indeed, without these the firm could not exist. Of special concern is how individuals fit into their positions, tasks, obligations and responsibilities prior to their occupation and conception. For example, a lecturer fits into its job and its rights and responsibilities (i.e. position) irrespective the identity of the lecturer (i.e. man or woman, black or white, working class or middle class, and so on), and its knowledge of the conditioning structures.
• The transformational conception of social activity refers to how the existence of social structure is the often unacknowledged but necessary condition for an individual act, as well as an often unintended, but inevitable outcome of, individual action taken in total. For example, in order for an individual to purchase a car in ‘the market’, it requires social structures such property rights, state regulations, international trading system, traffic police, finance, buyers, roads, and so on. Moreover, the numerous individual consumption acts will have unknown and unintended consequences for the car market; e.g., higher/lower prices, increase/fall in output, or an introduction/delay of a new model car. In short, social structures are the unmotivated condition of our motivated production, and the reproduced/ transformed condition of our daily economic/ social activities.

For critical realists, the objective of economics is to identify the structures governing a particular economic phenomenon; this means describing agents’ unacknowledged conditions of their economic practices, and evaluating their existing conceptions of their economic actions. Indeed, while society and the economy are the skilled accomplishment of active agents, they remain to a degree obscure to the individuals concerned.

There are three further considerations that bear on method, policy and scope of analysis.

• First, because critical realism acknowledges the openness of the economic world, and accepts the primary aim of science as the identification and elaboration of deeper structures that govern manifested phenomena, the central criterion of the theory assessment (what makes a good theory?) must be explanatory power, not predictive accuracy.
• Second, because critical realism recognises that event regularities are rare in the economic sphere, and that the aim of economic science is instead to uncover and identify underlying structures, it holds that economic policy should be concerned not with predictive control but with emancipation. Emancipation occurs through being informed about the knowledgeable transformation of structures that govern and facilitate human actions. For example, government policy should not change taxes and subsidies to affect firms’ pollution, but transform the conditions under which firms produce pollution, so dealing with underlying structures such as property rights, technological investment, alternative goods, nature of the business organisation and so on.
• Third, a broadly philosophical project like critical realism does not, and cannot, licence any specific political substantive theory. Economics’ primary aim is to elucidate the nature of the economy and society, and never lead to a particular political and moral philosophy. Critical realism is compatible with a wide range of politically implicated schools of economics, such as radical political economy, feminist economics, Marxian economics, Austrian economics and old institutional economics.

Finally, a note on why this perspective of economic methodology is referred to as critical realism. Because social structure is dependent upon human agency, it is open to transformation through changing human practices, which in turn can be affected by criticising the conceptions and understandings on which people act. The social sciences such as economics are unavoidably susceptible to change through critique. For example, neoclassical economics suggest a particular relationship between the state and the market. Yet, by highlighting the inadequacy of this conception, a more fruitful conception of state, market and society emerges that changes the nature of society and individual practices. Economics is dependent on (and requires) social criticism.