Abstract
In the early 1980s Washington State developed a simulation forecast for inmate populations that was used to estimate the impact of the state's pioneering sentencing guideline initiative. This process presented forecasting challenges that occurred in three phases over a 10-year period. Different types of forecasting challenges were specific to each of the three phases. These challenges are described along two dimensions: forecast horizon and level of detail. Various combinations of horizon and detail sometimes required mutually exclusive assumptions that yielded much different results.
Get full access to this article
View all access options for this article.
