Hi,

I am a graduate student and I am currently writing a thesis studying the effects of one of the high speed rails in China on local economic activities. I am trying to study at a micro-level since my dependent variable (night time light intensity) has an advantage of doing so, and use DID as an identification strategy. However, now it seems that other control variables such as income is available only at county level or above.

My question is whether it is okay to cluster(?) and fix controls at different levels of cluster.

Thank you.