Reply #26: And here's the problem with that claim...

[gcomeau]

Let's take his own example... programmers.

Looking at the H1-B LCA disclosures, sure you average them out and the average for "programmer" is lower than the BLA survey average wage for "programmer".

OMG! They're paying people less to do the same job when they're not supposed to!" you promptly declare, and people on DU turn red in the face and start stomping their feet and throwing tantrums.

Only, last time I checked "programmer" covered a HELL of a lot of ground, with a HELL of a wide salary range. And the pool of H1B workers is extremely small relative to the total programmer workforce. And we have this thing called "margins of error" and "measurement uncertainty".

Sure. The average salary for a "programmer" in the H1-B disclosures was mid 50k. Did you happen to know the standard deviation? I do. I just downloaded the figures and crunched the numbers. It's over $15k. There are people on there getting 30k a year, and there are people on there getting 150 grand. And there's the little detail that the BLA and the H1B LCA don't have 1 to 1 job title equivalency between them.

Starting to see some issues with this "study" yet?