January 20, 2022 In Qualitative Comparative Analysis


For a long time, QCA-researchers have been concerned with time and have endeavoured to account for time in QCA. In this blog, I argue that such efforts are misguided. That QCA is time agnostic (which is not the same as time insensitive) because it should be given its configurational notion of causality.

First of all, the concern for time in QCA is really nothing more than a concern for sequence, for the temporal ordering of conditions in a configuration. However, sequence makes for a very flat ontology of time. Time is much more complex than a simplistic A before B. Social reality is emergent, with things happening simultaneously (synchronic emergence) as well as sequentially (sequential emergence). The concern with the temporal ordering of conditions in QCA follows from a mechanismic understanding of causality. Mechanismic causality infers causal mechanisms from empirically observed causal chains. However, the empirical observation of causal chains rests on two fundamental mistakes. i) Causal chains confess to the above flat ontology of time as a sequence. ii) Empirically observed causal chains are idiosyncratic. They capture what happened in single cases but cannot be generalized into causal mechanisms. (See also my blog on ‘Causal mechanisms, some second thoughts’, 5 April 2021, roelrutten.com.)

Second, and following the above idiosyncrasy argument, QCA is a case-based method and cases are unique time and place specific social entities. Cases, too, are emergent; they have a history. Since no two cases are identical, also within-case dynamics captured as empirically observed causal chains, are case specific (idiosyncratic). A causal chain observed in one case may have no bearing whatsoever on other cases and cannot be generalized into a causal mechanism. That is, the answers to how-questions are case-specific and,for the purpose of generalization, are only relevant as steppingstones for answering the more important why-questions. In fact, recognizing that cases have a history underlines that QCA is indeed time sensitive. Going back to the cases, QCA-researchers can account for time, i.e., for emergence. They can discuss both the synchronic and the sequential emergence that happened in their cases. They can use this knowledge to refine and elaborate configurational explanations of why an outcome occurred. In fact, acknowledging emergence suggests cases as events rather than as, for example, organizations or countries. Case-based researchersinvestigate, e.g., decision-making events in organizations or election events in countries. Thus, case-based knowledge makes QCA time sensitive. Or rather, sensitive to emergence in both its forms and protects it against reducing time to sequence.

Third, in QCA’s configurational causality, sequence is perfectly irrelevant. Configurations describe causal powers rather than capture causal mechanisms. Causal power emerges when all conditions are present. For example, the causal power to pass an exam emerges from two conditions: reading the mandatory literature AND attending the lectures. Individually, these two conditions do not describe a causal power that makes it possible for students to pass an exam. Only when a student does both, does this causal power emerge over the course of the semester, i.e., over the course of the studying-event. Whether the student first attends the lectures and then reads the literature, or first reads the literature and then attends the lectures, or simultaneously attends lectures and reads literature throughout the semester is completely irrelevant. If the configuration [reading literature AND attending lectures] is sufficient for passing the exam, the sequence in which the conditions ‘happen’ is irrelevant. Regardless of the sequence, the causal power has emerged. That is, emergence, or the developing of causal power, is a different phenomenon than causality, i.e., the exercising of an emerged causal power. And these two phenomenon happen on different time scales. In this example, the causal power to pass the exam emerges over the course of the studying-event. However, it is only exercised at the exam, which marks the end of the studying-event. The problem with mechanismic causal explanation is that it collapses the emergence and the exercising of causal power into a single causal
chain, which evidences a flat ontology of time.

QCA separates the emergence and the exercising of causal power. Configurations describe causal powers that have emerged. Doing a QCA-study investigates whether this causal power is sufficient for (explains) the outcome, i.e., whether the causal power makes the outcome possible. How the causal power emerged is a
different question. Going back to the cases may shed light on that; however, QCA-researchers must still keep in mind that what happens in individual cases is idiosyncratic.

Of course, it may be that simultaneously reading literature and attending lectures (synchronic emergence) leads to better exam performance than doing one before the other (sequential emergence). However, this is a different question than: Is the configuration [reading literature AND attending lectures] sufficient for
passing the exam? This question is about exercising an emerged causal power and the answer is (must be) agnostic to how the power has emerged. It only matters that is has. The question whether synchronic or sequential emergence leads to better exam performance is about which kind of emergence ‘produces’ the
stronger causal power. Again, QCA-researchers may shed light on the emergence question by going back to their cases. They may compare the kind of emergence and exam results across cases (studying events). And they may even do that using a truth table analysis. This example thus underlines that exercising causal power
is something different (and thus requires a different explanation) than the emergence of causal power. The two should not be collapsed into a single truth-table analysis (or causal chain).

However, the example also hides the fact that time in social reality is complex. That synchronic and sequential emergence both happen during an event and may be interdependent. Thus, separating synchronic and sequential emergence may not be possible and may not be meaningful. That is, researchers must be careful
to avoid a flat ontology of time. This is the key message of this blog. The emergence and the exercising of causal power are different things that happen on different time scales. QCA’s configurational notion of causality explains whether causal powers, once emerged, are sufficient for an outcome. Worrying about the
temporal ordering of conditions in a configuration conflates configurational and mechanismic causality in an unhelpful way. In fact, in as far as mechanismic causal claims are based on empirically observed causal chains, they at once conflate the emergence and the exercising of causal power and mistake the idiosyncratic (within-case causal chains) for the general (causal mechanism).

In sum, QCA (configurational causal explanation) is time agnostic because it should be. Acknowledging cases as unique time and place specific events suggests that QCA-researchers can address how causal powers emerged by going back to the cases. At the same time, they must be mindful of the idiosyncratic nature of
emergence. QCA focuses on whether emerged causal powers (described by configurations) are sufficient for an outcome; on why they make an outcome possible. Explaining how an outcome occurred is a (largely) idiosyncratic exercise that is at risk of collapsing time into the flat ontology of sequence.

Leave a Reply

Enter Captcha Here : *

Reload Image