Skip to main content Skip to navigation


The term ‘causality’ refers to a very precise connection between a cause (X) and an effect (Y). For example to say poverty (X) causes low educational achievement (Y) generally means that poverty and educational achievement are directly related; that poverty ‘precedes’ educational achievement and is not somehow a consequence of low attainment; that there is a plausible explanation as to why poverty may cause low educational achievement (for example poverty creates a lower sense of self efficacy) and this explanation is more plausible than others (for example the claim that ‘bad teaching’ ‘causes’ low educational achievement).

There is an instinctive desire to identifying causality and repeated attempts have been made to present the natural and social world as one in which cause and effect can be observed and discovered with some degree of certainty and generalisability. Drawing out cause and effect is invariably problematic (e.g. Hammersley, 2014) for several reasons :

  • most ‘real world situations are inescapably complex. Staying with the idea of educational under-achievement, it can easily be seen that teachers, ethnicity, language, funding, parenting as well as the homogeneity, or otherwise, of schools will all play a part in achievement.
  • in many instances too the direction of cause and effect is often unknown. For example in relation to education and well-being Desjardins (2008) sees educational outcomes as a set of ‘dynamic interactions’ rather than one way cause and effect. They further describe education as a problematic area to research as the aims of education are contested or conflicting.
  • what seems causality may often be what the Scottish philosopher Hume described as 'constant conjunction’: X and Y seems to be regularly associated but X is not the cause of Y. In natural science, for example, thunder follows lightning but is not caused by it, a third factor is in play. While in education symbolic factors such as school uniforms, even homework, are sometimes seen as causes of learning outcomes when in practice the relationship between one and the other is uncertain.
  • connections often occur unpredictably as in the well documented cases of schools, or particular teachers, ‘bucking a trend’.
  • causality is philosophically a difficult notion in respect to human activity as human beings have a level of agency, consciousness and choice over their actions (e.g. MacIntyre, & Korbut, 2013). We might say we ‘had to do X there was no choice but in reality there are always choices. Of course how far we should or want to account for agency is a never-ending debate.

We could try to do without causality as we cannot escape it. For example even if we want to write a biographic or autobiographic account of a life celebrating the agency of the subject, we could not do so without recognising the events that shaped meaning in that life and ways in which opportunities were presented / constrained by material conditions.

A more sophisticated view of causality, pretty much conventional wisdom amongst social researchers, sees the world as much more complicated than it first appears and takes claims to causality as ‘tentative’ or ‘a balance of probability’ and subject to countervailing examples. In other words social research can provide illumination of, and insight into, situations, events, issues, policies and practices and can show important connections and correlations but it cannot show direct causal relationships or identify causal agents. We can, to take our earlier example, say that based on the available evidence there is an association between poverty and educational attainment but we cannot say that poverty causes low educational achievement.

If we are particularly wedded to the idea of causality what should we do?:

  • we could recognise the objections and simply say instead that while we recognise the picture is more complex a search for causality is valuable for the insight it affords us. It is almost as if we are providing almost a ‘what if’ account based on the hypothetical proposition that all actions are caused by external factors.
  • if trying to stand back and look at relationships between ‘variables’ we can pay attention to the language we use and avoid blanket statements of the kind X causes Y and replace it with phrases such as there is an association between X and Y which seems to suggests X as influencing Y. There are many ways of showing subtlety in our understanding of causality for example we can speak of ‘necessary’ (it has to be there to result in X) and ‘sufficient’ (if it is there it will result in X) conditions. But we can also talk of causal factors as necessary ones but intervening contextual factors which influence outcome but do not cause them. We may end up pointing to a series of interlocking factors which influence / have an impact on / affect attainment in certain contexts.
  • we can avoid the ecological fallacy that sees associations at the group level as necessarily impacting at the individual level. For example even if we know that even if there is very strong correlation between X (say belonging to group A) and Y (say voting for party B) this does not mean that each individual will vote in the same way. We have to allow for for human conscious some degree of self control.

Causality is a difficult concept not least as politicians, newspapers and other opinion leaders play on our ‘instinct’ to find simple causality when the picture is more complicated. Politicians often try to blame this group or that group for our misfortunes and offer simple solutions based on this or that observed association. Social research offers a more rational and measured arena in which causality can be pursued.


Desjardins, R. (2008) ‘Researching the links between education and well-being’, European Journal of Education, 43, 1: 23–35.

Hammersley, M. (2014). The Limits of Social Science: Causal Explanation and Value Relevance. London: Sage.

MacIntyre, A. & Andrei Korbut (2013) Russian Sociological Review, 12, 1:139-157.