Chris Dillow, Marxist economics writer for the Investors Chronicle, blogs regularly on all sorts of topics at Stumbling and Mumbling. A recent post of his discusses structure versus agency and the neglect of systemic analysis in economics.
Structures and structural forces tend to be left to macroeconomic analysis. In modern microeconomics, individual agency, or the ability to act in order to influence some outcome given a particular environmental context, dominates theory. But structure shapes the environment, enabling or constraining agency. Individuals may therefore act in good faith but be unable to achieve the best economic outcome for themselves in the absence of a favourable context.
Given the limits of human agency in a particular structural context, an understanding of that structure, whether it is an institution such as the state, or macroeconomic in nature, such as the force of competition throughout the economy, can enable a response from policy-makers to try to improve social welfare, or not as the case may be!
There is plenty of debate over structure versus agency in social science, but it is either confined to heterodox or non-mainstream economics, or outside economics entirely, in sociology for example, as Dillow describes.
Where it gets political is with the notion of responsibility. If structure influences individual agency, then individuals are not entirely responsible for what happens to them. ‘Positive’ intentions may lead to undesirable outcomes for the individual or society concerned.
This conclusion makes neoliberals, conservatives and libertarians of an individualist bent uncomfortable or angry, but seems to make sense for leftist interventionists who may be more drawn to collective solutions for social and economic problems.
Of course, human agency need not be individual. It can come from collective action, such as a mass protest or strike. These might be led by individuals or they could be decentralised and even anarchic. Here the lines get a little blurry with defining what scale of agency dominates a particular action.
The most satisfactory arguments over structure and agency are found in social theory and have influenced heterodox economists such as post-Keynesians and Marxists, who usually employ a more systemic or macroeconomic analysis.
There is a danger that a neglect of the individual in the analysis can lead to a bias towards seeing all problems in economics or sociology as a result of ‘the system’ rather than as the responsibility of the individual. As already mentioned, this kind of critique will find more favour with the political right.
But the danger also goes the other way, with the neglect of the social or systemic in the analysis leading to an absence of responsibility by the state or other bodies representing collective interests to improve matters.
So the theory we favour matters greatly to what sort of economic and social policies and actions, if any, are proposed and enacted by our political leaders, as well as the rest of society.
Both structure and agency matter. They influence each other in dynamic fashion, evolving through time. Individual or collective agency is enabled and constrained by structure. But structures such as institutions are themselves products of history, of human agency in the past, creating and modifying them up to the present. These ideas are the stuff of social theory but, as already mentioned, they are strangely absent in much economic thinking.
With respect to economics, the dominance of microfoundations in mainstream analysis has led to the neglect of macroeconomics. The latter is only deemed to be sufficiently rigorous if it is founded on particular microeconomic assumptions regarding individual behaviour.
Keynes’ paradox of thrift, or Marx’s theory of the capitalist imperative to accumulate driven by competition, are examples of macroeconomic outcomes which ’emerge’ from the interaction of individuals, but which cannot be predicted by simply observing and extrapolating from individual action. In such cases, the whole is greater than the sum of the parts.
Social theorists have argued that it is important to avoid the reductionist extremes of atomism (everything is founded on and is caused by the micro level, or the individual) and holism (everything is founded on and is caused by the macro level, or the larger whole).
If one is to assign importance to both parts and wholes, one can argue that wholes are made up of constituent parts, whose interaction gives rise to the behaviour of the whole. In the terminology, wholes are emergent from their parts, but are not reducible to them. Wholes can therefore influence their constituent parts, as much as vice versa.
Having said all that, how should one define wholes and parts? Is a whole the national economy, with firms as its parts? Or are firms the wholes, with employees as their parts?
Taking this further, and beyond economics, we could see the individual employees as the wholes, and particular biological or medical categories as their parts. At each level, the parts interact to produce an emergent whole. But it is human thinking and analysis which has named the wholes and parts, and decided, depending on the field of study, at what level they are so-defined.
This kind of thinking can take us beyond any simple division between micro and macro analysis, towards a more conscious way of seeing things, and make us aware of the categories and definitions that we have been drawn to, and that we have made a decision to use or even invent.
To finish, here is a quote from post-Keynesian John King’s fine book The Microfoundations Delusion (p.235-6), which is understandably critical of the way micro has come to dominate modern mainstream economics. He addresses all economists in particular, himself included, but there are also lessons for social scientists more broadly:
“First, mind your language. Metaphors are not mere rhetorical adornments. They can have substantive consequences, not always for the good. Be especially careful with constructional analogies, which have proved to be misleading and a source of serious confusion. Do not be tempted to replace ‘microfoundations’ with ‘macrofoundations’, which is not much better. Use horizontal rather than vertical metaphors: ‘bridges’ or ‘links’, rather than ‘underpinnings’ or ‘cornerstones’.
Second, be a good neighbour. Do not treat your colleagues in the other social sciences as colonial subjects, or feudal vassals, or disobedient children. Economics imperialism was always a bad idea. It was never going to succeed, and it should be abandoned now. Your neighbours in anthropology, political science, social psychology and sociology have a lot to teach you, and probably also much to learn. Cooperation will be beneficial to all concerned, but conquest is an idle dream. Remember Edgar Kiser’s metaphor, and be prepared to take part in two-way trade with the other social sciences.
Third, be kind to methodologists and philosophers of science. They do not bite, unless severely provoked, and they have a lot to teach you. They are for the most part modest folk, who are also prepared to learn from you. Perhaps in the future, when the dust has settled on the ruins of the microfoundations project, they will write it up as a salutary lesson: a case study in how not to do social science.
Fourth, and most important, be happy. The end of the microfoundations dogma is good news. Read up on the recent history of the other social sciences, where intellectual pluralism is alive and well, as confirmed by several contributors to the recent volume edited by Roger Backhouse and Philippe Fontaine (2010). Send your DSGE texts to be re-filed on the science fiction or science fantasy shelves of your library, and get ready to read up on some real macroeconomics.”