you only need to complete the excel part:
Homework 2: Conducting a Sensitivity Analysis in a model of the cost-effectiveness of alternative drug control policies
The goal of this problem set is for you to review the concept of net present value and conduct a parametric sensitivity analysis using a simplified version of the RAND two-stage model of cocaine control, documented the RAND mandatory minimums policy analysis. You should start your analysis from this template: PADT2024_HW2_RAND_Cocaine_template.xlsmDownload PADT2024_HW2_RAND_Cocaine_template.xlsm
In this analysis we will primarily stick with the parameterizations of the cocaine control model that were presented in the original report. However, the recent draft revision to Circular A-4 (the US federal government’s guidance document for regulatory analysis) suggests 1.7% as an appropriate long-term social discount rate. This is considerably lower than the discount rate used for the government in the original report. Please change the government discount rate in the Parameters tab to 1.7% and use this value for Questions 1 and 2.
1. Compare outcomes associated with programmatic spending on conventional enforcement and mandatory minimum sentencing policies:
- Finish populating the calculations in the Enforcement tab of the template.
- Plot the annual expected costs of an arrest for years 1-10 after an arrest for both the government and the supplier, under both conventional enforcement and mandatory minimums. (So your plot should contain four lines. Make sure to label those lines and your axes.) Include fixed costs in costs for Year 1, as these costs are incurred with arrest regardless of conviction or sentencing.
- Use this plot to explain in words what the RAND analysts meant when they stated that mandatory minimums “dilute the efficiency” of enforcement spending.
2. Understand the sensitivity of policy outcomes to underlying parameter assumptions:
- Populate the Policy Outcomes table.
- Conduct a parametric sensitivity analysis of cocaine consumption reduction outcomes by varying: (a) price elasticity of demand, (b) supplier cost of incapacitation ($/year), and (c) dealer discount rate. Do this first one by one (one parameter at a time) and then for extreme value combinations.
- Explain which parameter combination comes closest to favoring spending on long sentences relative to treatment, and why. Given the outcomes under this parameter combination, how would you characterize the robustness of a recommendation to spend on treatment rather than enforcement?
3. Consider what your analysis tells you about the varied objectives of mandatory minimum sentencing policies (these can be found in Chapter 2 of the RAND report, starting ~ p. 11). In 400 words or less, comment on the following:
- For which of these objectives can the model we studied in class be used to assess the efficacy of mandatory minimum sentences and why?
- Which of these objectives can you imagine a (different) quantitative policy analytic model could be built to speak credibly to the policy’s efficacy in meeting that objective? Give an example of what primary outcome such a model might project and what causal processes or data sources that model might draw upon.
- Which of these objectives is unlikely to ever be appropriately investigated using a quantitative policy analytic model and why?