3 ms·
==I work in government relations and this is neither news, nor anything new. == Lots of people in this article disagree: > A high-level former official who he
by take_a_breath 7y ago
==I work in government relations and this is neither news, nor anything new. ==
Lots of people in this article disagree:
> A high-level former official who helped address public health outbreaks in the George W. Bush administration said “it’s not normal to classify discussions about a response to a public health crisis.”
> Attendees at the meetings included HHS Secretary Alex Azar and his chief of staff Brian Harrison, the officials said. Azar and Harrison resisted the classification of the meetings, the sources said.
> One of the administration officials suggested the security clearances for meetings at HHS were imposed not to protect national security but to keep the information within a tight circle, to prevent leaks. “It seemed to be a tool for the White House - for the NSC - to keep participation in these meetings low,” the official said
- creaghpatr 7y agoThose quotes appear to prove OPs point about giving people in the discussions the ability to speak freely without second guessing from people like "a high-level former official".
- JamesBarney 7y agoThe point of state secrets isn't to allow elected officials to say embarrassing things.
- Simulacra 7y agoI didn’t say embarrassing, but if you’re talking about a mass casualty event and how many body bags are going to need, and limitations that are in the system, and where resources are going to be diverted, how many might die, etc etc the last thing you would want is the public questioning every single thing. Because they will. If they don’t, the opposition party certainly will.
- take_a_breath 7y agoThe same logic used to keep lots of important information out of the prying eyes of voters and oversight committees. ==the last thing you would want is the public questioning every single thing. Because they will. If they don’t, the opposition party certainly will.== And if you classify it wouldn't that also cause the public to question "every single thing" just the same? Seems like the case we have proves that to be true as they hid it and people are still questioning.
- JamesBarney 7y agoUsing the classification system to shield yourself from criticism is abusing it. The criteria is damaging to national security, not damaging to an officials reputation.
- ApolloFortyNine 7y agoSomething such as "we should prioritize efforts in cities besides Seattle/New York as the virus has clearly spread beyond reasonable containment, yet can still be contained in Atlanta/Miami" could be reported as "U.S Government elects to leave the people of Seattle on their own." Replace the cities or issue with anything you want. The point is, something has to be prioritized, and the ones who aren't prioritized could be seen as victimized.
- JamesBarney 7y agoClassifying doesn't keep you from having to prioritize things. Or people finding out that you prioritized certain regions over others.
- edanm 7y agoSometimes in life you actually do have to prioritize. The alternative is maybe doing a half measure and being far less effective. So the rational course is often to prioritize. However, no politician/official necessarily wants to be on record as making this kind of priority decision. So if you insist on radical transparency in these situations, you're insisting on people being less effective.
- JamesBarney 7y agoNot classifying meetings that have never been classified before is not insisting on radical transparency. But even if you assume that politicians have misaligned incentives because they care more about optics than saving lives, why reduce transparency? You're making the assumption there are more good decisions with bad optics than bad decisions with bad optics. Which seems like an unlikely assumption to be true.
- edanm 7y ago> Not classifying meetings that have never been classified before is not insisting on radical transparency. Agreed. I was referring to the principle of the thing, not this specific instance. > You're making the assumption there are more good decisions with bad optics than bad decisions with bad optics. Which seems like an unlikely assumption to be true. I think it might be true in certain situations, e.g. emergency situations like we're in now.
- take_a_breath 7y agoExcept that second guessing still happened after the classification, which disproves the point entirely.