Ian Moore, a post-doctoral fellow at IWH, will discuss a study that investigated three forecasting models to assess which one works best with time-series data on workers’ compensation lost-time, no-lost-time and total claim rates.
The researchers compared business cycle fluctuations between monthly workers' compensation claim rates and the unemployment rate, over the same time period. They determined that most of the significant periodic behaviour in monthly workers' compensation claim rates is seasonal. A three-month positively correlated lag was found between the seasonal components of the unemployment rate and workers' compensation claim rates.
Moore will present overall findings that show workers' compensation claim rates from 1991 to 2007 were driven by (in order of relevance) a strong negative long-term growth trend, periodic seasonal trends and business cycle fluctuations.