I would like to estimate to what degree (e.g. 20% fixed) the costs of e.g. a central function like HR develop with the growth of e.g. Revenue within a company.

My first though would be a regression for estimating HR costs Y if sales (X) go up by $1. However I'm supposed to classify HR costs to what degree they are fixed. Meaning, if they are 0% fixed, they are totally variable and they grow with the same growth rate as the Revenue. If they are 100% fixed they only grow with inflation growth.

How do you suggest should approach this problem? At the end I need to calculate that HR costs are e.g. on average 46% fixed, growing with inflation and 44% are variable and grow with Revenue growth.

Should I try a multiple regression where Y are the HR costs and the independent variables are Revenue (R) and inflation (I)? In this case, do you have a suggestion how to interpret the coefficients, so that at the end I can get the fixed / variable % share?

Merely trying the correlation between Revenue and HR costs seems wrong to me due to causality. Am I right?

Thank you and best regards