Few years ago, US Airways Flight 1549 had just taken off from LaGuardia Airport en route to Charlotte, North Carolina, when it struck a flock of birds. David Learmount, operations and safety editor of Flight Global, said that the pilot, Chesley Sullenberger “Sully,” made a textbook landing, saving 150 people. When the pilot was questioned about the event, he said that he had landed a plane in similar situations hundreds of times using a flight simulator. Just like Sully used simulation to learn and practice problem solving when facing unexpected situations. Policy makers and managers use computer simulation to explore important policy problems in different domains.

One interesting example of such models involves the work of the Rippel Foundation in trying to better understand Health Policy . Their work in understanding the effect of several policies in the future health of a population (see http://www.rethinkhealth.org/). Producing informed projections of policy options requires data describing the past history as well as an understanding of the problem’s underlying causes, often outside the scope of patient records and health outcome datasets. These predictions are not simple linear extrapolations. They require a deep understanding of the interactions and effects that drive system behavior over time. The Rippel Foundation has produced a simulation model called ReThink Health, which includes tools to project the effects of socioeconomic and government interventions, as well as health models to estimate the long-term costs of action and inaction on our health futures. The goal of this simulation is the development of “an aligned vision, system-wide strategy and a long-term plan” (Fannie E. Rippel Foundation 2015). The model is the result of years of research and development with the involvement of policy makers and other stakeholders in the health system. The ReThink Health Model can be traced back to efforts by the Rippel Foundation in 2007, which adapted simulation and modeling developed at the Center for Disease Prevention and Control (CDC), using a specific approach to systems thinking and policy making called System Dynamics (McFarland et al. 2015, Milstein, Homer, and Hirsch 2009, 2010). Subsequently, the Rippel Foundation decided to use the Rethink Health Model as a key element of a toolset to help local policy makers to better understand reform efforts. As you can read in their website, the model has been used by hudreds of policy makers and local leaders to build that shared vision. You can explore the model yourself at https://rethinkhealth.org/our-work/dynamics-model/.

A second interesting example can be found in the work of Climate Interactive through a couple simulation models developed to understand policy affecting climate change. Once more, the models are publicly available and can be explored in the foundation website (https://www.climateinteractive.org/). Both of their simulations have been used by governments, corporations and NGOs to better understand long-term impacts of policies to reduce greenhouse gas emissions. Both simulations were developed by scientists and policy makers to develop a better understanding on the causes of climate change. Nonetheless, the resulting model can be used in many other ways. These models are used as a substitution of the real world, where making all these experiments is impossible. Although it is not always impossible to run experiments in the real world, sometimes it is too expensive to do it. Simulation then becomes a suitable tool to try alternative strategies or policies. Training (of managers and policy makers), prediction (of potential future scenarios) and discovery (of side effects and unexpected behaviors) are other uses for simulation models.

Simulation Approaches to Policy Making

Although both examples introduced in the previous section involve the use of system dynamics simulation, system dynamics is only one of many modeling and simulation techniques that are used by academics, managers, community leaders and policy makers to explore future scenarios as well as the impact of different policy approaches. Figure 13.3 includes many of these approaches and the ways in which they are related to each other. Three major types of models can be identified in the picture. In the very left of the picture, it is possible to identify a clear line involving system dynamics modeling. Stochastic processes and queuing models are behind a second major category of models used frequently to think about policy problems called discrete-even simulation. Finally, at the right of the picture we find the family of modeling techniques that are commonly refered as agent-based simulation.

Figure 13.3 Simulation modeling approaches (Adapted from Pidd, 2010)

System Dynamics is a systems approach developed in the 1950s by Jay W. Forrester at the Massachusetts Institute of Technology (MIT). Its purpose is to gain a better understanding of certain problems and behaviors in order to be able to design strategies and policies that improve the performance of the system over time. System Dynamics distinguishes from other variants of systems approaches by its intensive use of feedback cycles and emphasis on main accumulations in the system (CO₂ in the atmosphere or number of diabetics in a population). System Dynamics is commonly used to explore and understand long-term dynamics of a policy problem and major trends. Discrete-event simulation is an alternative approach that focus on how important events in a system affect the need for resources within the system. For example, people arriving to the Department of Motor Vehicles to apply for licenses or any ither requests constitute important events that affect the need for public officers to help them solve their needs. A major assumption of this modeling perspective is that these events are stochastic (or random) in nature, that is to say, the rate of customers arriving follows a random pattern as well as the need they have. The major goal of the simulation approach consists of understanding which probabilistic distribution better explains the occurrence of the events, and what is the better combination of resources to maximize efficiency or service quality. The emphasis on optimization of resources makes this approach useful when looking at implementation of policy. Agent-based modeling is a third approach that assumes that the aggregated behavior of a system is the result of the interactions of individuals within the system (or agents). Moreover, the approach also assumes that agents’ interactions can be defined through a short set of rules. Nonetheless, a short set of rules can be used to predict the complex behavior of the overall system. Great examples can be found on swarms of insects or birds, that follow simple rules of interaction among them producing interesting and complex behaviors when they move from one place to another. Agent models are commonly used to understand long-term behaviors of systems.

Simulation Approaches and Stakeholders

Simulation approaches can be applied in most policy and management domains, and applications exist in Public Health, Education, Energy, Project Management, Emergency Management and Response, Finance, etc. According to Bardach (1996), the complexity of analyzing a public policy lies, among other things, in the fact that many actors are involved: interest groups, public officials, popularly-elected public officials, citizens, and civil organizations, among others. In addition, institutional and legal frameworks govern all public policy, which includes laws, standards, regulations, and important cultural aspects. The interaction between multiple social actors and the context means that the analysis of public policy is considered more of an art than a science (Bardach 1996). However, using a consistent methodology can help us understand the complexity since it allows one to identify the elements or parts of a problem and analyze their exchanges. He proposes eight steps for the systematic analysis of public policies: (1) define the problem; (2) obtain data and information; (3) prepare alternative solutions; (4) select criteria; (5) project the effects of results; (6) review costs and benefits; (7) choose a solution that addresses the specific problem; and (8) reveal the solution’s history. Analogically, as we have shown in Figure 1, the System Dynamics modeling process can be represented by stages: (1) state the problem dynamically; (2) create a hypothesis explaining the dynamic behavior; (3) formulate the model; (4) evaluate the model created; and (5) formulate and evaluate policy options.

Simulation approaches have developed tools and techniques for policy making and problem solving with clients (Kim, MacDonald, and Andersen 2013). In the specific case of System Dynamics, there is a clear research interest in working with groups, managers, and policy makers to facilitate strategy and policy development. Since 1987, the field of GMB has made considerable progress in developing and adopting tools for stakeholder analysis, facilitation and collective thinking with managers and pther stakeolders to solve problems.

Attribution

By Luis F. Luna-Reyes and Erika Martin, and licensed under  CC BY-NC-SA 4.0.