What do dummies variables exactly capture?

[Jelle de Vries]

Hello!

I have a question regarding the use of dummy variables. I want to use an i.industry dummy in my re model to adjust for industry specific effects.
I'm interested in the effect of marketing intensity on firm growth. Every industry will have different average marketing intensities. Will my i.industry dummies capture this?
I.e. is the dummy sufficient, or should I create a new variable with a firm's relative marketing intensity depending on its industry?

Once again, all help is much appreciated!

[Jared Greathouse]

Well, dummies are usually intended to capture unobserved confounding or (sometimes) unit specific variation.

Assuming that you've no metric for marketing intensity, then yes, in theory, this captures that. But, you write you're

interested in the effect of marketing intensity on firm growth

so presumably you can just calculate marketing intensity, right? Why use the dummy for that purpose?