Counterfactual Thinking and What-If Analysis

2.1 What Are Counterfactuals

2.1.1 Definition and Intuition

Counterfactuals refer to statements about what would have happened under alternative circumstances that did not actually occur. In causal analysis, they answer questions of the form: “What would the outcome have been if a different action, treatment, or policy had been chosen?” This idea is fundamentally different from prediction, which only concerns what is likely to happen given observed patterns. Instead, counterfactual reasoning compares the observed reality with a hypothetical alternative reality in which some condition is changed.

Formally, counterfactuals describe outcomes under unrealized scenarios, outcomes that are not observed but are conceptually well-defined. To illustrate, in the case of a person who was given a policy intervention, the counterfactual outcome is a situation of what would have occurred to the same person had the intervention not been implemented. It is then given a causal effect that is with a comparison made of these two possible states of the world. This mode of thinking is at the core of the contemporary causal inference in which the causal effects are determined not by the association of data, but rather by comparison of actual and hypothetical outcome.

Recent research stresses that counterfactuals can be used to offer an intuitive connection between the human cognition and the formal causal study. As Wang et al. (2024) [22] emphasize, counterfactuals are inherently stated in terms of what-if options that can assist analysts and decision-makers to reason the cause-effect relationships, as opposed to their correlations. In line with this, Imbens (2024) [23] emphasizes that causal questions are counterfactual in the social sciences, as such always involve the comparison of what actually occurred to what it would have occurred had a different intervention or choice been made. Causal claims cannot be assigned any specific meaning without this hypothetical comparison.

Accordingly, the intuition of counterfactuals is quite straightforward and persuasive, causality concerns contrast between realities, the world as it is and the world as it would be with a different action or policy. This difference is what separates the causal analysis and the mere prediction or descriptive statistical analysis.

2.1.2 Historical Background in Economics

Counterfactual reasoning has a long history in economics, especially program evaluation, policy analysis, and welfare economics. Even when only one policy is actually implemented, thought experiments have been important in the assessment of the implications of alternative policies like tax changes, subsidy changes, or regulation changes, by economists. In this respect the use of counterfactuals in economic reasoning has always been implicit: an evaluation of a policy must involve asking how things would have turned out under an alternative policy choice.

According to Heckman and Pinto (2024) [21], thought experiments are at the core of econometric causality and therefore, they are essentially counterfactual. Their work demonstrates that the fundamental activity of causal econometrics is to build decent comparisons between the actual world and well-defined hypothetical alternatives. In this view, counterfactuals are not of recent origin, instead, it is a formalization of an old tradition in the field of economics of thinking about alternative phenomena to evaluate the effect of a policy and social outcomes.

This tradition has been formalized in the contemporary empirical economics in the form of causal inference structures that render counterfactuals explicit and central. According to Imbens (2024) [23], the contemporary causal analysis in the social sciences defines the causal effects as the possibilities of the outcomes, which are explicitly counterfactual objects: in regard to each unit or individual, the outcomes are conceptualized under various possible interventions, although it is only possible to observe one of them in reality. Such formalization is a significant move to make economic reasoning about causality more specific, transparent and testable.

The general history of counterfactual thinking in economics entails a change in the informal thought experiments of policy to formalized statistical and econometric models that can put counterfactuals at the center of the causal analysis. Counterfactuals are now known as essential instruments of responding to causal questions, assessing policies, and explaining economic processes since without them, the concept of a causal impact would be definitionally unclear.