|
Do these look like pretty good estimates? I pulled MCPS GIS data for school boundaries and linked them to census blocks. Then I pulled median income bins for the block groups from ACS data, assigning them proportionally to the schools based on the household numbers by census block. Then it using the income bins to interpolate an overall median for the school.
But, top income bin in ACS is $200k, so a few schools need augmented data to come up with a median above $200k. This estimates that tail by fitting a truncated lognormal distribution to PUMS data. Not perfect, but I'm not I'm not sure what else to do. Note that this is counting families, not all households. |
|
You can't seem to do tables or monospace fonts here...
[img]https://i.imgur.com/ymMJFY9.png [/img] |
|
| There is no way that’s correct |
|
wow - this seems like a lot of work. good for you i guess.
this does help prove why everyone hates Whitman lol |
| Most clusters are huge! Elementary schools? |
Detailed census data isn't available for small areas. I'm not sure how much I'd trust the binning and interpolation that would have to happen at the elementary school level. Especially in the high income areas. |
|
The above $200K doesn’t seem right or a bunch of people in W district don’t work and life off trust funds. ~40% are below a gs-15?
Median is also a wild measurement. |
What seems off? |
|
WJ seems off.
-DP |
Whitman is not very high. |
| off in what way. |
Lots of retirees that paid much less for their homes zoned for Whitman |
Expand on that. It wasn't quite what I expected, although it isn't terribly far off. It is basically about 10% less than I expect. I need to think about how to handle the interpolation. The current approach might be skewing things towards the middle of a census income bin. Keep in mind "family" doesn't necessarily mean a family with kids. It could be a retired married couple. |
Correct, but even a retiree would need more than that for property taxes, insurance, health insurance and utilities. |