Identify a policy that is not usually intended to be a health policy but that you think may have important health implications.
The American Clean Energy and Security Act of 2009 was the first bill passed by the House of Representatives that aimed to reduce greenhouse gas emissions by setting a national cap on total emissions. It had several ambitious plans ranging from 20% reduced emissions from electric companies by 2020 and a strong dedication to the development of renewable energy technologies. It also had a proposed consumer protection from energy price increases, which allowed for estimated average of $160 in additional expenses per year for every American citizen.
Describe why an evaluation of that policy is informative (e.g., determining effects of the policy, or primarily a test of hypothesized mediators).
The bill did not reach the Senate and it was never passed. While there were many counterarguments, one that I saw came from the Charlottesville community group Creciendo Juntos. They wrote to their congressman to explain that although they were in support of climate protections, for many people in their community, the increases in the cost of energy would affect their ability to keep electricity in their homes. The bill did not adequately protect their well-being.
Specify the outcomes and populations you think most affected or least affected by the policy.
There are several health outcomes that would likely be affected if an individual or family loses electricity in the home. At the household level, this may affect the ability to have clean clothes, food in the refrigerator, or hot showers. If money becomes tighter, transportation to and from work may be affected. If entire neighborhoods are affected, there may be increases in crime. All of this could lead to increased stress, and ultimately, biological changes that amount to decreased overall health.
Propose a study design to evaluate
the policy.
The pretty good counterfactual study would be to simply enact
the bill for a random sample of US counties for two years, and measure any changes
to the cost of energy and a specified set of health outcomes and other
variables of interest from all participants in the study. Then ask whether
there are worse overall health outcomes and whether the health outcomes are
differentially affected by measures like income level. While this would give a
good sense of the health disparities and concerns that may arise from a
proposed greenhouse gas limitation bill, it would not be acceptable due to the
ethical issues that stem from the design. Instead, a study that compares a
historical example of differential energy costs in comparable counties may
provide some useful data.
Aim: To measure the effect of
increased energy costs on stress related mental health conditions.
Method: The effect would be estimated by comparing the experience of
differential energy costs between two separate counties within the same region that
had comparable demographic data with respect to average income, net worth,
percent pop below poverty line, and standard of living.
Study population: A suitable percentage
of each relevant income range. Could use census tracts to identify individuals,
or could possibly use regional hospital medical records to obtain demographic
and medical data concurrently.
Describe biggest challenge to implementing and drawing inferences about the impact of the policy on health.
It would likely be hard to find the two counties that are comparable and have suitable populations. When such a situation is identified, it may be difficult to convince hospitals to participate in the study. Assuming the data is obtained, the representation of different income levels within hospitals may not reflect the expected experience for that income level. For example, if someone can’t afford to keep their house powered, they would probably be less likely to use medical services. Therefore, there may be different qualities about the individuals in a given income level who do use medical services that are not accurately portraying the full story about those in the same income level who do not medical services. Each hospital could have additional influences on the outcomes. Frequently, such effects would come down to unmeasured confounders, and the potential generalizability of the study would be diminished. This could possibly sway future estimates of the true effect of increased energy costs out of favor for lower income populations if the bias in the study was towards the null.