The previous output showed simple OA-OA pairs, with OAC coding.
We can do more with OAC, however. This example looks at aggregating origins by OAC supergroup. Thus, for given destinations, we can look at the supergroups of migrants’ origin OAs. NOTE: It is easy to assume that a migrant who has come from a given supergroup has particular characteristics matching the profile of that supergroup – e.g. that someone from a ‘multicultural’ OA is non-white. This is not necessarily true. e.g. those who can afford to leave a ‘constrained by circumstance’ OA may themselves *not* be constrained…
Picking destinations: we could just use some sample OAs, but flows would not be big enough to generate meaningful comparisons of mix of people’s origins. Instead, we aggregate destinations to district level (thus, looking at migrants’ origin OAs to all OAs in target district). Flows within the district are retained in this example, but could easily be excluded.
Which districts to pick? Use exemplars at top, bottom and middle of IMD (2004 version) ranking in England. IMD is based on SuperOAs, so: take the score (not rank) for each SOA in each district, and calculate a mean score for the district. This was not population weighted, but still gives sensible answers…