This paper analyzes a study by Harding (2003), which used the counterfactual causal framework to explain the rate of teenage pregnancy after exposure to poverty. Harding (2003) used the framework to explain “the difference in outcome between the world in which the subject receives treatment and the counterfactual world in which the same subject does not” (p. 684). Here, treatment was the exposure to poverty. The counterfactual causal framework is a statistical analysis of cause and effect, based on the potential outcomes, as proposed by Rothman and Greenland (2005) in his work titled, “Causation and causal inference in epidemiology.” They used this model in the context of randomized experiments. Harding (2003) used the counterfactual causal framework by applying the concept of potential outcome to explain the probable health outcome after evaluating a set of conditions that would lead to the preferred outcome. Usually, counterfactual conditional statements describe these potential outcomes (Rothman & Greenland, 2005). Some researchers have used the counterfactual causal framework to predict different income levels among people who attend private and public colleges (Sage, 2015). However, to come up with the different potential outcomes for both situations (high-income and low-income), a researcher has to look at the potential outcome of the same person in two probable futures (Sage, 2015). Since it is always difficult to know the outcomes of both potential outcomes, researchers identify this problem as the fundamental problem of causal inference (Sage, 2015). This same problem characterized the work of Harding (2003).
I agree with the author’s selection of the causal model because examining teenage pregnancy in low-income and high-income neighborhoods required an intuitive explanation, as opposed to a formal explanation of counterfactual causal frameworks. Stated differently, the author’s focus of study required a proper measurement of potential outcomes (rate of teenage pregnancy) across two groups of girls who received different treatments (poverty) and those who do not receive the same treatment (high-income). Since it is difficult for the respondents to continuously experience the treatment and then stop experiencing it, it was essential to use more facts to fill the information gap (Harding, 2003). Therefore, the author used the fundamental problem of causal inference because the information gap experienced in the research was the causal inference problem. This assessment often aligns well with epidemiological studies because it borrows the logic and language of scientific experiments (Sage, 2015).
Given the evidence of causality highlighted in the article sampled, there can be no causation without manipulation. This statement significantly impacts public health policy because causality emerges as the net difference in the outcome, through variations in socioeconomic factors (Harding, 2003). Furthermore, it is pertinent to understand the extent that the net difference in causality stems from malleable variables in an intervention. In the context of the sampled article, the net difference would be the change in the teenage pregnancy rate. The malleable variable would be the exposure to poverty. Indeed, exposure to poverty is a manipulative agent. This is why this paper argues that such a finding has significant implications for public health policy because health outcomes could only differ because of manipulation. This statement affirms the views of Sage (2015), which says thoughts of real manipulation affect the definitions of units and treatment (a critical process that determines if a health program would meet its goals). Lastly, based on the outcomes identified in the article sampled, it is similarly essential to point out that the association does not equate to causation. Stated differently, although two variables may have a high correlation, their relationship does not mean that one variable causes the other. Therefore, although Harding (2003) found out that poverty had a direct correlation to high teenage pregnancy, it is incorrect to assume that poverty causes increased teenage pregnancy. Collectively, these issues have a strong bearing on the formulation of public health policies.
References
Harding, D. (2003). Counterfactual Models of Neighborhood Effects: The Effect of Neighborhood Poverty on Dropping out and Teenage Pregnancy, American Journal of Sociology, 109(3), 676-719.
Rothman, K. J., & Greenland, S. (2005). Causation and causal inference in epidemiology. American Journal of Public Health, 95(1), 144–150.
Sage. (2015). Counterfactual Framework and Assumptions. Web.