The census is a large survey with relatively consistent methodology, and likely to be responded to in roughly equal measure across large numbers of people. This consistency leads me to think that their geographic representation is likely accurate, despite an inaccurate absolute number (because the census is an unreliable way to count illegals). This group also publishes how they create estimates from the census data, documenting their own methodology and therefore generating a useful number.
That's the part I'm happy to acknowledge. An analogy would be if you take a large bag of Skittles and dump it on a map of the country, they have a great skittle-counting machine, but they're only counting one color. And I happen to know that more Skittles are out there because my machine counts Skittles by weight. That's why I know they undercount.
I've seen proprietary evidence of remittance flows that demonstrate significantly more individuals than this group is counting. You don't have to believe me if you don't want to, but that's how I know the undercount is severe. I also know that my methods don't produce perfect numbers either (not all illegals send remittances; many remittances are sent by people who aren't illegal aliens; each remitter may send multiple times), but I'm confident my methodology relies on a better proxy indicator than theirs.
Independent validation that you can check yourself comes from public reports published by central banks that document remittance inflows, like Banxico in Mexico or the Central Bank of Kenya. From time to time the Pew Center and other institutions have published remittance flow studies with fairly accurate average amounts per remittance and/or per individual: use those numbers and divide.
Believe what you want, but other methods consistently show higher numbers as well. (I kind of like tracking the growth of specific ethnic food products compared against reported legal immigration growth from that area, which is obviously not speculative at all.)
Ok. I get it. You're just trolling.
The census is a large survey with relatively consistent methodology, and likely to be responded to in roughly equal measure across large numbers of people. This consistency leads me to think that their geographic representation is likely accurate, despite an inaccurate absolute number (because the census is an unreliable way to count illegals). This group also publishes how they create estimates from the census data, documenting their own methodology and therefore generating a useful number.
That's the part I'm happy to acknowledge. An analogy would be if you take a large bag of Skittles and dump it on a map of the country, they have a great skittle-counting machine, but they're only counting one color. And I happen to know that more Skittles are out there because my machine counts Skittles by weight. That's why I know they undercount.
I've seen proprietary evidence of remittance flows that demonstrate significantly more individuals than this group is counting. You don't have to believe me if you don't want to, but that's how I know the undercount is severe. I also know that my methods don't produce perfect numbers either (not all illegals send remittances; many remittances are sent by people who aren't illegal aliens; each remitter may send multiple times), but I'm confident my methodology relies on a better proxy indicator than theirs.
Independent validation that you can check yourself comes from public reports published by central banks that document remittance inflows, like Banxico in Mexico or the Central Bank of Kenya. From time to time the Pew Center and other institutions have published remittance flow studies with fairly accurate average amounts per remittance and/or per individual: use those numbers and divide.
Believe what you want, but other methods consistently show higher numbers as well. (I kind of like tracking the growth of specific ethnic food products compared against reported legal immigration growth from that area, which is obviously not speculative at all.)