Hey. I'm Professor Paula Lantz. Today we're going to be talking about policy simulation modeling. Our roadmap for this discussion is first, I'll talk a little bit about what are policy microsimulation models. I will define a couple important terms in these models, cost-benefit and cost-effectiveness analysis, also talk about sensitivity analysis, and then later on after I'm finished, you'll have the pleasure of learning from Professor David Mendez about a microsimulation model that he's developed. Just as a reminder, we've talked a little bit about this already. Policy simulation models are a type of prospective policy analysis. This is a kind of policy analysis where we attempt to forecast or predict specific outcomes or conditions in the future, usually under status quo conditions. What would the future look like under the current policy regime and then what might that future look like under one or more policy changes? Now, policy simulation models are pretty technical endeavors and they involve a lot of kinds of programming, analysis and many different types of data. In Decision Science, there are four general types of simulation models; microsimulation, sometimes called Monte Carlo or risk analysis models, also, there's discrete event simulation models, agent-based modeling and simulation, and also system dynamics simulation models. What we're going to really focus on is microsimulation models because these are the ones that are used primarily in policy analysis. What is a microsimulation model? It's a computer-generated forecast that attempts to imitate or mimic the operation of government programs or policies, an individual or micro units in a population under different sets of assumptions in different kinds of policy interventions. By micro units, we mean people mostly, but also we can do this simulation model to look at what the effects of policy changes might be on households, businesses, or other small-level organizations. Now, these policy microsimulation models are mostly designed to look at what the effect of policy change might be, it says on the slide, but almost always take demographic processes like the ones we've talked about, birth, death, and migration into account as well. The role of the data analyst in this policy simulation modeling is really important. First of all, you're going to be the member of a team because again, this is a pretty technical endeavor and so there will be people with a lot of different training involved with developing and then using a microsimulation model. But a data analyst as a member of this team could have a number of different roles. First of all, identifying data sources, where might the data come from that's going to be used as inputs into the model. Also, cleaning that data and then doing some re-coding of the data and managing the data is a big part of the project and data analysts really get involved with that. Also, when there are some results, creating data visualizations of the results is going to be very important. The major responsibility of a data analyst, however, is communicating any known problems or limitations with the data to the team. Often, you're going to be the one who knows the ins and outs and the strengths and also the limitations and problems with the data that the microsimulation team wants to use. That's an important role. Also, to do that, you need to know the purpose of the model and the questions that the model is attempting to answer, so you can help ensure the optimal use of data and that the model is producing high-quality results. The things you're going to need to know are, what are the goals of the simulation? That can be under the heading of what projections or forecasts into the future look like under status quo or no further policy changes or different interventions. I mean, really what we're trying to design here is a crystal ball that's going to help decision-makers predict what the future might look like. You also need to know as a member of the team what the assumptions are that are going to go into the model, so we can try to predict what the future impact of new policies or programs are going to be. These assumptions include not only what might be the effectiveness of a policy change, but also we want to look at the differential effectiveness in terms of creating new or exacerbating existing inequities. There's going to be cost issues that people want to look at and other societal impacts. Also, these microsimulation models often are used to look at the potential impact not only of policy changes but other shocks we might say, economic shocks like a recession or rising inflation. There could be environmental shocks like a natural disaster of some sort, other social shocks rather than intentional policy changes. Models can be used to predict these things as well. I've just gone over at a very high level some of the basics of policy microsimulation models and the important role of the data analyst. Let's jump into a few examples.
