The model you are proposing has great post-hoc explanatory powers within the united states - however it does not appear to have predictive powers over what we will find in other places and times. Therefore I question it's validity, until you can show how the same model applies to other situations, and can show exigent factors that are relevant when it doesn't apply. Again, this is not rage - this is merely not accepting an insufficiently explained model at face value.
FWIW (not that it matters): I am a software engineer/programmer, my parents are in sales and HR, my other family members are not in education either. I just think overly simple explanations should always be questioned.
As for the salary of a government worker - that depends on the job they do, the value they create etc. I don't know the details to make blanket statements. I think there is a tradeoff - good people doing the job requires competitive salary, paying bad salaries is generally more expensive as it requires more workers to do the same job.