The world's top destination country for migrant workers, the United States, is enacting a set of unprecedented taxes on high-skill immigration that include charging employers roughly $100,000 for each foreign worker they hire using the H-1B visa. The US government has relied on one academic study to design the tax and evaluate its impact on high-skill immigration under the H-1B quota, an impact the study estimates at zero by measuring the wage differences between H-1B and native workers. This paper precisely replicates the study's method and finds that its conclusion rests on three major flaws, each of which renders its results difficult to interpret. The study mismeasures the wage gap between H-1B and native workers, misinterprets any negative wage gap as reflecting markdowns rather than task specialization, and carries the wage results into a policy simulation whose central estimates are invariant to the wage gap: they predict an identical (null) effect on immigration for any H-1B wage gap including zero. Improved methods are required to offer regulators useful information on the effects of this large shift in immigration policy.