For this problem, your sample means have a distribution of N(mu,sigma^2/n) where n is the sample size, mu is the population mean and sigma^2 is the variance.
If you don't know the population variance, you can use the t-distribution which uses the sample variance and your test-statistic will belong to a t-distribution with n-1 degrees of freedom. Since your distributions are assumed to be normal, using this should be OK.
Also you may want to consider using H0: mu2 - mu1 < 0 vs H1: mu2 - mu1 > 0. Remember you are comparing two means; one from data five years ago and one from current data.