Explaining changes in family population by geographic type.

I’m examining the population among people in families by geographic area. Adding together the totals in principal cities and suburban and rural areas, it looks like the total family population is about 99 million in 2011 and 2012. Then in 2013, the total drops significantly to about 71 million. Why is that happening?

At the same time, it looks like the missing data rate for geography type jumped from 37% to 55% between 2012 and 2013.

Could you please help with an explanation for those changes?

The 2012 ACS was the first sample to use 2010 Census-based PUMAs instead of the 2000 Census-based definitions. These PUMAs were then mapped onto the 2013 Office of Management and Budget metropolitan definitions to create METRO for 2012-onward ACS samples. The ACS samples prior to 2012 use the 1999 OMB metropolitan delineations. Due to these changes in both PUMA and metro area definitions, the population totals you have calculated are not comparable before and after 2012.

Hope this helps.