"BUT I would have to say yours is totally incorrect. SORRY. ;-)"
Hi Simon,
You must have been talking to my wife - I hear that all the time!
Let me try to back off a little as I was convinced that I had been able to derive 'new information'; but it doesn't seem to be working.
First of all let me get the emotional bit out of the way. I dearly want OC to have 3%. If not 3%, ok 2% would be nice. But, deep in my heart, I suspect that 2% sounds high. If my message indicates that it doesn't look like being that high, it is certainly a blow I would rather not be delivering. It is tense enough round here as it is.
Back to basics on how the allocations between Regions are arrived at. You say that visa quotas have NOTHING to do with Population. I regret that doesn't ring true. The method of allocation between regions is specified by law and can be easily read in FAM 42.33 instructions. Transliterating, what is specified is that the Attorney-General (no less) must using set criteria identify the 'high admission countries' and kick them out of the analysis. (We all know the high admission countries, the manual requires them to be promulgated.) Then [quoting the manual] 'the Attorney General must then determine the population of each of the six regions (excluding the population of any high admission foreign state) and use those totals to determine the apportionment of the 55,000 worldwide DV limitations. Quotas for the six regions will be established'.
Using publically available figures and the above rules, I calculate the resultant 'world population' after removing those high admission countries (and the US) is about 3300 million. SA (after removing high admission countries) is about 120 [call it 3%] OC at 38m is 1.2%.
Quite candidly, if SA and OC were ever to receive exactly the same quota, the screams from Latino Senators/Representatives would reach Kentucky. (After all, this whole DV thing only exists because of the screams of Irish-descended Senators.)
I guess that is my starting point.
Next, we both know that OC and SA have been receiving (almost exactly) the same number of selectees per year. Hence there must be another 'picking mechanism' at work to make sure that the Attorney General is not breaking the law. One such mechanism would be to run OC at 1.5% [we're dealing with bureaucrats here so rounding is allowed] and SA at 3% - and that would come out close enough to be defendable at congressional hearings. I suspect that you would object to that and stick with 2%/2% for OC/SA and I am not trying to argue you out of that position; but I, myself, cannot find any objective basis to justify it.
The next thing I came across was the recent CEAC data; loved it and, yes, I have statistics qualifications. And using such analysis, the data are telling me that OC is running at 1.5% and SA and 3%. That's the message in the data, I'm just the messenger.
.... and, you know, those 1.5% and 3% figures just kinda keep popping up, don't they?
And another thing I learned from the CEAC data is just how brilliantly someone is managing the monthly CN cut-offs to keep the statistics aligned so closely; it sure isn't happening by magic.
Finally let me note and accept your criticism of the 'data set'; I have even deeper misgivings with some of the data. But we are trained to work with that (after all it is probability). And we now have (grossly) 50% of the data as our sample. I can assure you that the Actuaries who priced your house and car insurance had nothing like this quality of data. And, I guess - and hope - you paid the premium - despite their crappy data sets.
So let me leave it there and, perhaps, agree to disagree.
But in doing that, wise men navigating in uncharted waters - like Chris Columbus, who started this whole adventure - prudently keep a weather eye out for the unexpected.
We both just might be counselled to do that.
Cheers.