Through WW2 and the proliferating age of defense contractors, the government (mostly defense department, but also civilian agencies) stood almost on equal footing with contractors in their ability to design, scope, and evaluate big projects.
You would see scientists and in-house advisers at these departments able to expertly evaluate proposals/designs by contractors with sufficient background knowledge and tools to do so. They even worked closely with contractors to lay out the requirements and designs for systems, or products.
But, through the decades, a couple factors eroded this equal footing of the government / regulatory experts:
-- Shrinking of government budgets for (or unwillingness of the public to stomach) the ranks of Washington "bureaucrats" who represented this expert class of people (what harm is there in cutting "fat" from public servants who don't seem to produce anything tangible?)
-- More attractive pay, career potential, prestige, etc. of working in the private sector
-- Political distaste for being seen as working too closely with contractors
So what happened is that gradually but surely, government lost the tools to do these things themselves, and by sheer need to still have things approved, shifted the work onto industry.
What can you do when industry comes to you with new complicated designs for things, and you have no one who can assess (and no budget to pay for assessments of) whether those designs are safe? You ask the person proposing to critique themselves, and in many cases, they seem to know more than you anyway.
Of course, what in part probably led to our current situation.
As a society, my question is, how do we make it possible to choose to do these things in the way that produces the right outcome? For a start, I think we need to stop asking everyone to make uninformed votes about things we don't understand, yet rely on every day.
Absent a politically impossible change to the way the government handles employment, trimming the ranks is the only way to get rid of the harmful parasites that infest the typical government agency.
As a developer, I can say the same about writing software. Yet, my software is not likely to kill people and comes with no warranties whatsoever.
In this scenario, we should make sure the test engineers are even more competent than design engineers. We can't rely on designers to thoroughly test what they build.
Then you either have to put up and do it or you'll be strategically moved elsewhere or be managed out of the job.