I'm using a number of indicators in my research which are not available at a sub-national level. Is there a reliable method for disaggregating the country data into city ones? (preferably with a reference)
please note that population and GDP per capita are available at the city level.
https://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=&cad=rja&uact=8&ved=2ahUKEwjW6YHdlvPuAhW2WxUIHXoLCLMQFjANegQIFxAD&url=https%3A%2F%2Fwww.iisd.org%2Farticles%2Fdisaggregated-data-essential-leave-no-one-behind&usg=AOvVaw2uxPXt9FkX1JQam_Btwtxt
No, there is no "reliable method for disaggregating the country data into city ones".
Any approach to 'disaggregate' data must be based on assumptions and additional information that allows to distinguish the regions. The validity of an approach is based on whether the assumptions are intuitive and true or not.
One example: you have GDP data for a nation but you aim to have GDP data at a sub-national/regional level. Now you can 'disaggregate these data by assuming all people have on average the same GDP per capita across the regions. Knowing the number of inhabitants, you can then easily 'disaggregate' GDP. If the new disaggregated measure is of any validity depends on your assumption and the quality of your data on the number of inhabitants per region.
I guess you aim to do that 'disaggregation' in order to analyze regional differences, right? As the regional differences may be driven by your approach to 'disaggregate' your indicators, the empirical results that stem from such an analysis can be considered questionable.
Stephan Huber many thanks for your explanation. Indeed, I want to compare different cities according to different indicators. Unofourtionalty, sub-national level data isn't available and as you mentioned the assumption of equal weights won't be appropriate.
- Badges
- Science topic