This student project shows the effects of joining disparate time series. The candidate notes that the apparent population growth in 2000 might reflect data collection techniques that underreported population during the 1990's and then finally reported an accurate population in 2000 or might reflect an exogenous intervention, such as people who wanted to have "Y2K" babies. The eleven year lag from 1989 to 2000 causes the pattern of the correlogram, with the highest negative autocorrelation at 11 years. For the student project, you need not validate the data, but you should note the effects of possible data errors, as this candidate does.