Create more dynamic households about june HOT 4 CLOSED

It is a good point, but we haven't included them because they appear to be very rare in the UK. You can find more info here https://www.ons.gov.uk/peoplepopulationandcommunity/birthsdeathsandmarriages/families/bulletins/familiesandhouseholds/2019 . If I understand Nomis classification correctly, multi-generational families would enter in the category of multi-family households, which are about 1% of all households. We would need to take care of inter-generational mixing differently, perhaps simulating weekly visits to grandparents or something like that.

from june.

Multigenerational households is one proposed reason for the apparent higher mortality rate in the BAME population so might be an important factor even if they only represent a small proportion of overall housing.

from june.

I'm not sure that ONS number quite captures it @florpi (although I agree it's not totally clear). I think that number has a specific definition for families (e.g. requiring a child and the child's partner to live with the child's parents) but I don't think it captures e.g. a grandmother living with her child, her child's spouse and their children. In fact, do we allow for two adults in their e.g. 50s to live with someone retired e.g. a parents in their 80s? I'm also not sure how much we allow for a child to live with their parents if they are of working age either (e.g. I think more that 20% or people 25-35 live with their parents).

This ONS data doesn't seem to match the number given in the other link and I think is clearer and shows that there are a non-trivial number of households, particularly if you scale it to 2020 numbers: https://www.ons.gov.uk/peoplepopulationandcommunity/birthsdeathsandmarriages/families/adhocs/008846threegenerationhouseholdsuk2001to2013

Also this study, and other cited, might help: https://www.tandfonline.com/doi/pdf/10.1080/14036096.2019.1653360

My thinking was exactly that @rhayes777. We had discussed at one point about whether trends correlated socio-economic and racial factors would come out naturally form the model, but this would seem like a key element (although for now I think we don't include any race information dn just distribute).

Hopefully this shouldn't be too difficult a task but could add bit more detail. Then maybe we address other multi-generational household issues?

from june.

Arnau and I are working on this at the moment, also as part of a more general issue (see #49)

from june.