How describe IPUMs Data


I’m using data from Decennial 1990, Decennial 2000, and ACS 2012-2017. Each year is a cross section of new survey individuals, but its for the same geography–NYC metro area. How would you characterize IPUMs data–as repeated cross-sectional? I wouldn’t characterize it as panel or longitudinal because we are not following the same individuals, but there is a longitudinal aspect to it.

Thanks. Best,

The US Decennial Census and the ACS samples are fundamentally repeated cross-sections of the US population. This is because, as you point out, the samples do not follow the same individuals over time. If you aggregate the data to geographic areas with consistently defined boundaries, then the data can be described as a panel data at that particular geographic level. For example, if you aggregate the data to the state level, then you can effectively create a state-level panel in the United States.

Thanks Jeff. I appreciate the confirmation.