Dynamic Scoring

What is Dynamic Scoring?

Dynamic scoring is a method of estimating the budgetary impact of a change in government policy, which accounts for the secondary economic effects of the policy on all sources of government revenue and expenses in addition to the direct effects of a policy on spending and revenue. In dynamic scoring, these secondary effects are estimated using some sort of macroeconomic or econometric model. Because these models can take many forms and include many different kinds of assumptions about the structure of the economy and people’s economic behavior, the results of dynamic scoring can be highly dependent on the specific model and assumptions used. Dynamic scoring can be contrasted to static scoring, which estimates only the direct impact that a policy change will have on government revenues and expenditures without otherwise assuming any change in the economy as a result of the policy. 

Key Takeaways

  • Dynamic scoring is a method of estimating the total fiscal impact of a policy change, including the secondary economic effects.
  • When government policies change, people tend to adjust their behavior as a result of the policy in ways that can impact the tax revenue or government expenditures in other ways. 
  • Dynamic scoring can provide a more complete picture of the impact of a policy change than static scoring, but it is also highly dependent on the type of model and assumptions used to estimate these secondary economic effects. 

Understanding Dynamic Scoring

When any government policy changes, people will tend to change their behavior in response. To a large extent this is usually the point of the policy change in the first place, but we also know that often changes in government policy can come with unintended consequences and that the changes in people’s behavior may involve more than the immediate, direct response to the policy change. 

Because fiscal concerns are such a priority for policy makers, the direct and indirect effects of a policy change on government revenues and expenditures is typically of particular concern. To this end, when a new policy is proposed, estimating and projecting the fiscal impact of the new policy on the government’s budget is normally a major component of the debate around the policy change. This process of estimating the fiscal impact of policy change is known as “scoring”.

Scoring is traditionally done by a method now referred to as static scoring. In static scoring, the direct fiscal impact of a policy is measured or estimated using a simple model. For spending changes this is usually quite straightforward; the fiscal impact is the amount appropriated for the expenditure or an estimate based on simple assumptions around participation or demand on a particular program. For tax policy changes, revenues need to be estimated, but still the assumptions used to estimate the revenue generated are usually simple and noncontroversial.  

For example, if a proposal is put forth to place a $0.05 per gallon retail tax on milk and 50,000 gallons of milk are bought and sold annually in the jurisdiction, then using static scoring the tax could be estimated to raise $0.05 x 50,000 = $2,500 per year. However, because the tax also impacts the total price that consumers would now be paying for milk, and the law of demand tells us that people will tend to buy less at the higher price, the actual revenue will almost certainly be less than $2,500. This is where dynamic scoring comes in. 

With dynamic scoring, economists can use economic models to predict the amount of milk demanded on the market will fall with the new tax, and using econometric models to estimate the shape of the demand curve for milk they can put a number on how much they estimate this effect will be. Note that in theory this technique should arrive at a more accurate estimate of the actual fiscal impact of the policy change. However, because dynamic scoring depends on introducing some economic theory and econometric modeling into the mix, the improved accuracy of the estimate produced will only be as good as the theory, modeling assumptions, and reliability of the model estimates. While more accurate in theory, dynamic scoring also introduces a lot of new potential for error in practice.