This document discusses using the Box-Jenkins methodology to forecast unemployment rates in the US from January 2007 to July 2007 using past data from January 1948 to December 2006. It first provides an overview of the Box-Jenkins methodology and its key steps: identification, estimation, diagnostics, and forecasting. It then applies these steps using R: identifying an ARIMA(1,1) model as best fitting the deseasonalized data based on minimizing the AIC, estimating the parameters of this model, and selecting ARIMA(1,1) to forecast future unemployment rates.