A Brief Overview of the BSIM
What Does the BSIM Forecast?
The BSIM generates forecasts of economic variables in future years. It first calculates forecasts for a baseline case intended to represent the path the economy would take assuming no major changes (“shocks”) occur. It then calculates forecasts for an alternative scenario in which some major change/shock occurs. For our purposes, these shocks take the form of changes in public policy. The difference between the forecasts for the alternative and baseline scenarios is the end result of this process. Economic variables forecast by the BSIM include (among other things) levels of private sector employment and real output.
What Is the BSIM?
The Business Size Impact Module (BSIM) is a dynamic, multi-region model based on the Regional Economic Models, Inc. (REMI) structural economic forecasting and policy analysis model which integrates input-output, computable general equilibrium, econometric, and economic geography methodologies. It has the unique ability among forecasting models to forecast the economic impact of public policy and proposed legislation on U.S. businesses differentiated by size of firm, allowing analysts to estimate how policy changes might impact small businesses and their employees. The REMI model is the leading forecasting and policy analysis model in use and is employed by hundreds of governmental agencies, universities, consulting firms, nonprofits, and others.
When You Say “Model”, What Exactly Do You Mean?
An economic model like the BSIM attempts to describe the economy using mathematical techniques intended to capture past relationships between economic variables (consumption, investment, government spending, interest rates, etc.) and extend them into the future. Such models usually include formal descriptions of major components of the economy, like consumers and households, firms and producers, and government.
The REMI policy model, on which the BSIM is based, is a 70-sector NAICS (an industry classification system) model and provides a high level of detail on the producer side. The BSIM provides even finer detail by describing private sector employer firms down to categories defined by the number of employees per firm.
How Does the BSIM Differ From Other Forecasting Models?
The principal distinguishing feature of the BSIM is its ability to forecast the impact of policy changes on U.S. businesses differentiated by size of firm. No other model (of which we are aware) has this capability.
What Do You Mean by “Dynamic” Model?
Economic models come in two varieties: static and dynamic. Both varieties attempt to describe the behavior of certain economic phenomena vis-à-vis mathematical modeling techniques used to capture both (a) the historical relationships between economic agents and variables and (b) forward-looking assumptions regarding such relationships. Such models usually include descriptions of consumers and households, firms and producers, and government. The primary value of these models from a policy perspective is their ability to forecast how the economy might respond to a so-called economic “shock” (some event that occurs “outside the model” like 9/11, a tsunami or other natural disaster, or a change in policy).
Static and dynamic models differ in their forecasting approaches. A static model will evaluate the shock consistent with a predetermined, fixed response by economic agents (consumers, producers, government) to the shock. The shock, along with the response of economic agents, help determine the forecasts produced by the model (output, employment, price levels, etc.). The scoring methodology employed by the Congressional Budget Office is one example of a static approach used to evaluate the impact an economic “shock” (a change in tax policy, in this case) may have on government revenue.
Like a static model, a dynamic model will also evaluate the shock and predict the initial responses by agents. But unlike static models, a dynamic model will allow for adaptive behavior by economic agents as a consequence of the shock and a changing economic environment. Because economic agents in the real world adapt their behavior when confronted with new information, dynamic models may be thought of as more “realistic” representations of the economy than static models in at least one sense.