Did insurance premiums really increase because of Obamacare? I am continuing to write my essay on Universal healthcare, showing both the pros and cons, but I haven't been able to find any evidence Obamacare actually increased premiums, all I can find are conservative media outlets saying so. Trump often claimed Obamacare caused insurance prices to rise, and even some to lose their insurance, but I'm have trouble finding any evidence of that. Does anyone know anymore about this, or have any links to credible sites showing evidence of it?
You're viewing a single comment thread. View all comments, or full comment thread.
Comments (46)
sorted by:
Your firsthand experience would be howled at by reddit, who would call you a liar and vehemently persecute you because aside from perversion they don't believe in outliers. Because muh science, muh data. Nevermind the reproducibility crisis, of course. That is a deep rabbit hole indeed.
I for one don't care, nor am I sufficiently arrogant as to tell someone not to believe their own eyes.
Except for when the firsthand experience fits with their own narrative and worldview. Because everything they approve of is perfect science and anything otherwise is "word we goggled to sound smart when dismissing it." The same way stats are fake when we don't like it here, except when its 13/52 or 41% then stats/data are infallible truth.
Their retardation doesn't mean we need to gimp ourselves just to be their opposite. Firsthand is greatly beneficial for many things. Writing essays and compiling "credible evidence" like OP is not one, because it fails to hold up when poked. The same way "my gay uncle never molested anyone" doesn't pass versus the sheer number of predators out there.
You shouldn't dismiss firsthand. You need only consider how small of a person you are and how that gives you a very narrow field of view on the world, and factor around that.
I'm not suggesting anyone gimp themselves. I'm suggesting that statistical data is not a be all, end all, especially when science is so blatantly for sale AND in the middle of a reproducibility crisis like never before seen in history.
If you aren't going to examine the data and the methodology yourself, then it's not only worth not assuming any given statistic is correct, but I would argue it's actually worthwhile to assume it false. Especially in cases like OP, wherein he is dealing with something that has been actively suppressed from study.
That's the industry's problem. A person doing their own data collecting has all the tools necessary to create a perfectly valid collection of data without that ever coming into play, other than if they use other sources without considering how those might have been changed by it.
That's what I have been saying from the start, except you must also apply it to your own collecting. And asking random people you know has easy holes you can poke in it with barely any knowledge of how stats work to begin with.