Good morning everyone, hope you're all well - and hope that you can help with some confusion.

I have a dataset where I am looking at a categorical variable describing monthly household income, and the dataset has two sources of clustering - over time, and from the randomisation procedure (cluster randomised trial, at health clinic level). The same people have responded at baseline and at follow-up. The dataset is in long format. I want to look at whether the stated household income has changed between baseline and follow-up. I'm working my way through the 'multilevel and longitudinal modeling using stata manual', and am having difficulty finding the relevant section and code that would take into account the clustering over time and account for clustering at the clinic level as well. Can anyone help with suggesting code for examining whether there has been a change in how respondents have answered the household income category between baseline and endline?

My variables are as follows:

hhincomecat:
0 "0-2,000 rand"
1 "2,000-5,000 rand"
2 "5,000-50,000 rand"

time:
0 Baseline
1 Post-lockdown

Health clinic:
categories 1-12 with name of clinic


Happy to give further information. Thanks in advance.