I chose to determine how much aggregate away from home food expenditures varied with changes in aggregate disposable income in the United States between 1929 and 2014. Given that this was a time-series analysis, positive autocorrelation was a concern prior to running the first regression. The resulting Durbin-Watson statistic of .457 confirmed these suspicions, so PROC AUTOREG was run with a lagged dependent variable. Following this the model failed to reject the null hypothesis that there is autocorrelation with a Godfrey statistic of 3.5 (p=0.0611). As a result, it can be said the model predicts that for every 1$ increase in aggregate disposable income there will be a two cent ($0.02) increase in aggregate away from home food expenditures (p<.0001).