Thanks for the reply. Sadly I found two errors in my database. I have to check all entries again but expect a considerable change (predictions will be -5% off).
Hi Roger,
I have been away from the forum for quite a while mainly working in the three C's: Connecticut, Colorado and California!
Your data looks very cl0se to mine - which I posted back in Feb using the formulas and census data.
(I used 80.9% as the regional split factor; which back-fits the previous years precisely!)
Back then, I was able to predict that for 50k it would turn out:
[The number in brackets is an interesting prediction (mainly extrapolation) that used the CEAC data at the half-way mark.]
AF 20863 (22164)
EU 18725 (20248)
AS 8225 (7142)
SA 1325 (1468)
OC 862 (816)
I'm not too worried about the OC mismatch; anecdotal travels round Nolita in NYC indicate that most of the ANZ population there have interviews.
AOP is likely to be of significant impact.
So in the end, I guess that all in all the data reflects pretty accurately - what the law says it should.