Combining 1-year supplementals

Wondering if it is statistically valid to average data from two or more 1-year supplementals. If so, can I also average the MOE or would I use another approach to find the combined MOE?

Purpose: I'm doing data visualizations of socioeconomic changes over time in a 5-county area, and each of the counties are 20-30K population. 

I'd like to use data from five periods: 

2000 Census

2005-2007 3YR

2008-2010 3YR

2011-2013 3YR

And ... since the 3YR was discontinued... in October when the 2016 supplemental is released, I'd like to combine the 1 YR supplementals for 2014-2016 to make a 3YR for that period.

Thanks for any feedback on this.