When I was conducting my first true policy analysis in graduate school, I set out my policy analysis around three major criteria: effectiveness, efficiency, and political feasibility. I had seen from a recent report that Ohio’s worst indicator compared to other states was food insecurity, so I wanted to see what policy options could potentially turn Ohio’s bad problem with food insecurity around.
As I conducted this analysis, I found the most useful contribution I could make was around efficiency. Even more importantly, I defined efficiency in a specific way: how many dollars would it take to make someone no longer food insecure given a specific intervention? In the end, I found job creation programs pulled someone out of food insecurity for about $700,000 per person, cash transfers pulled people out at a cost of about $70,000 per person, and nutrition education pulled people out of food insecurity at a cost of about $700 per person.
This was my first time doing cost-effectiveness analysis.
Doing cost-effectiveness analysis is not rocket science, but before you dive into it, it is a good idea to get a good understanding of what you are doing. This starts with understanding what cost-effectiveness analysis is.
What is cost-effectiveness analysis?
At its heart, cost-effectiveness analysis is the question of how much bang for your buck you get for a given project. I have written before about how benefit-cost analysis contrasts with cost-effectiveness analysis but if you understand what benefit-cost analysis is, it is basically a type of cost-effectiveness analysis, but the variable of interest you are trying to maximize is how many dollars of social benefit you get compared to social costs.
The simplest equation you can use to calculate cost effectiveness is dollars divided by a given outcome. The cost of a program divided by the number of people it pulls out of poverty is a cost-effectiveness metric. In the example above, you can see how my study focused on dollars divided by people pulled out of food insecurity. You could imagine a range of different denominators that you could choose for a cost-effectiveness analysis.
How is cost-effectiveness analysis different from cost-benefit analysis?
If you want a full treatment of this question, you can refer to the blog post I mentioned above. But at its heart, cost-effectiveness analysis is a broader type of analysis that is answering a narrower question. Cost-effectiveness analysis tells you how many dollars a given approach costs. In a way, it is putting a valuation on an intervention, albeit a narrow one. It tells you how effective an intervention is at bringing about a specific outcome. Cost-benefit analysis, on the other hand, focuses on dollars in, dollars out.
A nice benefit of cost-effectiveness analysis is it gives a simpler scale for both analysts and policymakers. By focusing on one outcome, the impact identification phase of analysis is simplified greatly.
When should policymakers use cost-effectiveness analysis?
Cost-effectiveness analysis is a great tool when policymakers have a specific social goal they are trying to bring about when a policy is implemented. For instance, if policymakers are looking at different interventions to reduce chronic homelessness, having an idea of how much it costs under each intervention to pull a single person out of chronic homelessness helps policymakers compare alternative approaches against each other.
Cost-effectiveness analysis is also a powerful tool for helping policymakers deal with a neverending problem in public policy: budget constraints. Often, a policymaker who is working for an agency will be given a budget and a mandate and will have to figure out how to carry out that mandate with that budget. In this case, cost-effectiveness analysis is a powerful tool for figuring out how to help as many people as possible with a fixed budget.
Examples of cost-effectiveness analysis
We have already talked about a couple examples of cost-effectiveness analysis, but cost-effectiveness analysis has been used in real policy contexts to bring real improvements to lives.
The United Kingdom’s National Institute for Health and Care Excellence uses Quality Adjusted Life Years to recommend treatment options. Since the United Kingdom’s health care system is entirely run by the government, this helps the state get the most bang for its buck for interventions, maximizing the number of healthy years provided given the cost of health care in the country. The State of Oregon takes a similar approach with its Medicaid program, prioritizing cost-effective treatments that demonstrably improve health outcomes. The Centers for Disease Control and Prevention uses cost-effectiveness analysis around vaccine guidance, estimating the relative benefit of vaccines compared to alternate uses of resources.
Limitations of cost-effectiveness analysis
Cost-effectiveness analysis is not a panacea, of course. A huge problem with cost-effectiveness analysis is the problem of tunnel vision. For instance, while health impacts of Medicaid coverage have been mixed, the evidence around how Medicaid helps people with financial security is much stronger. After all, Medicaid is an insurance program. Focusing on the most cost-effective treatments may not be the best deployment of Medicaid resources. It may make more sense to go after the most costly programs, using the market power of the state to hold prices down and make these treatments more affordable for people.
Cost-effectiveness analysis can also lead policymakers into a McNamara fallacy of thinking that because they have outcomes quantified that they know how a policy will turn out. Proper sensitivity analysis has to be applied to cost-effectiveness analysis to make it a valuable exercise, and analysts must also disclose assumptions as much as possible so policymakers have an idea of what considerations are left off the table during an analysis like this.
Bottom line
Policymakers will not get an answer to every question about a given policy using cost-effectiveness analysis. But deployed correctly, cost-effectiveness analysis can tell a policymaker a crucial insight: how well does this policy fulfill a core goal of the intervention compared to other comparable policies? If a policymaker sets out to reduce homelessness and an alternate policy can do it at half the price per person, that alternate policy deserves consideration. And a policymaker deserves to have that analysis at her fingertips.
