Good morning!I am working on trying to compare cities to their suburban counterparts. From my national dataset,

Good morning,

I want to compare cities to their suburban counterparts. From my national dataset in Stata, how to I cut out MSA’s? And then how do I cut out cities from the MSA’s?

For example, from my national dataset, I would right now just like to look at the city of New Orleans and its data. How do I do that in Stata?

Thank you!


In order to identify an MSA, you need to use the appropriate metro area variable (METAREA or MET2013, depending on the year). For New Orleans, the METAREA value is 556. From there, you want to identify whether the person living in the metro area was located within the central city (METRO=2) or outside the central city (METRO=3).

However, there are two issues that make an analysis like this difficult. First, you should read this User’s Note on the difficulties of identifying metro areas. As you’ll notice, the New Orleans MSA could not be completely identified in most post-1980 samples. Therefore, while it is possible to create metro area estimates from the microdata, these estimates will be subject to error. Second, you cannot always identify each individual central city. For New Orleans, there are three central cities within the MSA: New Orleans, Metairie, and Kenner. Thus, the METRO variable would let you know if a person lived in any one of those three cities within the MSA, not just New Orleans alone.

I hope this helps.