creating area-based socioeconomic and demographic variables in registry data

[alex seger]

Hey,

I am working with registry data and I am looking to create area-based variables for ethnicity and SES per postal code.

Specifically, I need to create 2 new variables:

1. Area (postal code) with high concentration of foreign-background or foreign-born population
2. Area (postal code) with high concentration of low SES households

The existing variables (name) I can work with are the following:

• postal code (pn)
• mother tongue (mt)
• country of birth (cob)
• citizenship (ctz)
• SES (ses)
• average family income per postal code (inc)

My thought process is that I should use categorize postal codes in reference to the % of foreign-background/foreign-born and low SES households living in the specific area (e.g. 1-5 categories, indicating overall % SES distributions and native/non-native pop.).

The tricky thing is that there is no clear-cut data on whether person is of a non-native ethnic background as many individuals may write the dominant language of the country where they reside as their "mother tongue" despite having a different language spoken at home. In this case, the variable indicating a high % of non-native households per postal code should take into account 3 variables; mother tongue, citizenship, and country of birth.

Any tips for how I could code these 2 new variables I need for my analysis?

thanks for the help in advance