Why structural explanations are needed
15 Dec, 2025 at 18:57 | Posted in Theory of Science & Methodology | 1 Comment
All explanations of a phenomenon have underlying premises which limit the number of alternative explanations. These premises significantly affect the capacity of different potential explanations to truly explain anything. If we have a system in which underlying structural factors govern the functional relations between the parts of a system, a satisfactory explanation can never disregard this premise. Certainly, explanations that take their starting point in the parts (‘micro-explanations’) can often describe how and through which mechanisms something occurs, but without the structure we can never explain why. Women’s lower level of education can indicate a mechanism behind the gender pay gap, but is in itself insufficient and often not necessary. If women acquire higher education, discrimination can still occur through some other factor.
Behind every individualistic explanation lie structural premises. The methodological individualist’s ‘hyper-concreteness’ therefore most often means the explanations become unstable. When we ask for the cause of why women, on average, have lower wages than men, it is not a sufficient answer to state that Ada has a lower wage because …, Beda has a lower wage because …, and so on. If one changed the premises in these concrete cases, the outcomes would certainly be different. But in this context, we are not really interested in either Ada or Beda. These objects of explanation are simply too specific for the question we are asking. We are looking for more stable and general causal factors.
But can one not simply say that structural and individualistic explanations are different, without needing to grade them as worse or better? No, I do not think so. That would be too relativist. They are, to be sure, two different forms of explanation answering entirely different questions, but the structural ones answer most often the more relevant questions. In social science, we often seek explanations for phenomena so that we can thereby avoid or change certain outcomes. We want to know why unemployment is a common phenomenon in a market economy, in order to then be able to alter institutions or economic policy to prevent or reduce it. We want to know what causes gender-based pay discrimination in order to be able to prevent it, etc. If we give individualistic explanations — Bo is lazy, Eva stayed at home to look after the children for two years — this becomes impossible. The individualistic explanations do indeed state sufficient conditions for creating the outcomes in the individual cases. But they do not identify the necessary factors behind unemployment and discrimination. And it is these we must know in order to be able to prevent or avoid undesirable social phenomena.
If these phenomena are reduced to simply being about individuals and their characteristics, there is a great risk of actually ‘explaining away’ the very phenomenon one wishes to explain. If one begins to reduce gender discrimination down to ‘just’ being about women’s absence from the labour market for childcare, etc., the discrimination soon dissolves into nothing.
If one sets out to explain the discrimination of immigrants in the labour market, one can similarly make it completely disappear with the help of explanations such as, ‘for this particular type of job, the foreign qualification is not adequate,’ or ‘for this particular job, we rather envisioned someone with a different skills profile.’ Indeed, this type of explanation can often even be used to mask sexist and xenophobic attitudes.
What this highlights is simply that certain types of questions require a certain type of answer. If we want answers to macroeconomic questions about the relationship between inflation and unemployment, it is not possible to give non-trivial explanations of this using individualistic explanations which take as their starting point precisely what, in a more relevant explanatory approach, constitutes that which must be explained — the price system itself and its relation to the labour market.
That unemployment rises when prices fall is not a law of nature, but a result of a specific economic structure and its mode of operation. Explaining why Bo is unemployed and Lidl is raising the price of milk is not the same as explaining why an economic system, for its functioning, presupposes a certain degree of unemployment and why this is affected by the rate of inflation.
That women often have lower wages than men does not mean that gender per se causes the lower wage. It does mean, however, that a certain gender — being a woman — causes one to often be discriminated against, and that it is this social discrimination which is the real cause of the lower wage.
Structural explanations have the advantage of being more stable than individualistic ones. Models are often products of some kind of idealisation, and problems arise when the idealising assumptions lead to significant errors in the model. If the value of any of the model variables changes — even ever so slightly — the model can change qualitatively. This often happens in individualistic explanatory models and indicates a lack of stability which structural explanations do not exhibit.
Even if stability — like adequacy and relevance — is a pragmatic criterion for a good explanation, it is no less important for that. All types of explanation in the empirical sciences are pragmatic. We cannot say that one type is false and another true. Explanations have a function to fulfil, and some are worse and others better at this.
Even if individualistic explanations can reveal a pattern, the pattern in itself does not constitute an explanation. We want to be able to explain the pattern, and for this purpose, a structural explanation is usually required.
By studying labour market statistics, we can certainly establish that not all those willing to work have jobs. We might even note a pattern — those living in rural areas, women, and the elderly are more frequently unemployed. But with these data, we cannot explain why this is the case, and why a certain level of unemployment might even be a precondition for a functioning market economy. The individualistic ‘explanatory framework’ gives a false picture of the kind of causal connections involved and, consequently, also gives a false picture of what must be done to bring about a change. For this, a structural explanation of the type that critical realism speaks of is required.
