According to this paper from 2002, the OECD measures structural unemployment using the NAWRU
In recent years, the OECD has measured the structural rate of unemployment
using a specific indicator, the non-accelerating wage rate of unemployment
(NAWRU) indicator, first suggested by Elmeskov and MacFarland (1993)
and Elmeskov (1994). Crudely, this indicator measures the structural rate of
unemployment as the rate of unemployment at which wage growth is stable.
However, the paper also concludes that the NAWRU is not stable and does not accurately reflect real-world unemployment.
This is evident from its poor
predictive power; as shown in Section II, the NAWRU indicator has virtually
no explanatory power even within sample.
This paper from the Canadian International Labor Network talks about differing definitions of 'unemployed'
For example, Australia and the United States require “active job search” for
classification as unemployed, while Canada and most other OECD countries include both
“active” and “passive” searchers among the unemployed. The difficulties are also
illustrated by the fact that within the same country there have been changes over time in
key procedures. For example, “discouraged workers” were classified as unemployed
prior to 1975 in Canada and prior to 1967 in the United States, but are now treated as
being out-of-the-labour force in both countries.
The paper isn't long (conclusions are on page 9) but they do conclude the US and Australia should include passive seekers in unemployment numbers.
To try to give you a more satisfying answer than 'structural unemployment is measured badly,' I'm going to mention something I heard on this podcast by NPR's Planet Money team. Basically, they said that the unemployment rate among skilled, college-educated professionals is so low at around 2.2% it's not theoretically supposed to be possible. This could account for the lack of "qualified applicants" in many jobs that require a good deal of investment in education and training. Hopefully it's OK to cite a recording with no transcript.