That is a different david shor

On some level you might have more confidence in these numbers than historical Census-a much much greater share of the people hardest to survey now interact with the USG on their way in and while here. Barring like massive CBP data problems, reasonable disagreement w CBO is probably low or mid 100ks

CBO also looks at CPS, which inherits some ACS problems and probably underweights FB recently. They do a not-confidence-inspiring adjustment for this. But even the unadjusted data shows huge recent immigration, all consistent with the admin data, just from who’s getting picked up in the sample

No double counting here, no laggy ACS or survey coverage issues, no questionable legal status imputation, definitely no Fazel-Zarandi et al-level sloppiness. Other than maybe EWIs, none of the assumptions matter much to the bottom line (CBO uses DHS “gotaway” estimates). It’s all about CBP releases

For departures they use standard (not necessarily good) emigration rate assumptions. Huge uncertainty here but not huge implications. Any assumption in a reasonably wide range yields the same headline. And lots of corroborating evidence (immigration court records, EADs) that people have stayed

In case you haven't seen it, CBO has a fairly detailed description of where their numbers are coming from: www.cbo.gov/publication/...

Can quibble with many of their details (I do!) but the big picture all comes down to the sheer number of NTAs/paroles/other releases in CBP administrative data