Alright, I often kid about how disappointed I am in MSNBC, but to be honest, the really problematic “liberal” media darling is NPR. You heard me right – they’re a mess. I’ve written elsewhere about how they’ve equated the testimony on marriage equality by members of queer families with heterosexist screeds by members of straight-headed families, acting as though those fundamentally dissimilar acts from people with widely divergent social perspectives can be easily reduced to “people talking about their families”. Since then, they’ve started riding the explicitly abilist part of the austerity gravy train. In that vein, they revealed yesterday that they don’t know anything about statistics or really anything like basic logic. I don’t think NPR’s editors need to go back to journalism school, since at this point I’d be surprised if they can accurately identify colors and numbers. Maybe they should try kindergarten?
In response to continued analysis of the “natural experiment” of Oregon residents previously just a smidge too wealthy to qualify for medicaid being given it by lottery, NPR made a lot of statements. For instance, they introduce the data with a quote from one of the statisticians that “We don’t see any improvements in this window in hypertension, high cholesterol or diabetes”. At least in the following paragraph they admit that said window was from 2008 to 2010 – a two year period. Considering that the experimental group began receiving treatment at the beginning of that year, why would we necessarily expect statistically significant improvements in those conditions? All of them are also easier to prevent than treat, and easier to treat then ignore. Doesn’t it make more sense to analyze how many people were able to receive medical care or suggestions to prevent them from developing those diseases? The time period is too narrow to actually analyze how people under different conditions manage those diseases.
NPR’s piece goes on to note that while people with those conditions didn’t necessarily see a reversal or improvement in them, it does admit that there’s a stark increase in economic security and self assessed happiness among medicaid-recipient families. Naturally enough, amid the claims that only those less “deep” indicators improved, NPR admits that depression diagnoses dropped by about 30 percent among medicaid-recipients. In spite of that, NPR makes it clear that there “was no statistically significant difference between the Medicaid group and the control group” in terms of blood pressure or cholesterol level.
(The non-rejection region is where the data value cannot allow for either the conclusion that the hypothesis [medicaid improves biomedical quality of life] is right or that it is wrong – it’s non-evidence in essence. From here.)
Yeah, hold the phone. There’s some weasel words there. It sounds like there indeed was a difference between the two, just not a very large one given the size of the population being studied (which happens a lot in studies, actually). So it’s important to present this in terms of what it means. The data don’t counter-balance the point that having medicaid drastically improved people’s mental health, so much as fail to suggest that there were also physical benefits. The null hypothesis (that medicaid has no medical impacts) couldn’t be rejected, but that doesn’t mean that the null hypothesis is necessarily true.
Speaking of counter-points, the ever-“liberal” and ever-“balanced” NPR had only one policy wonk on to discuss this analysis, and what do you know it’s someone from the American Enterprise Institute (why is that so familiar?)! Scott Gottlieb, yes that Scott Gottlieb, raises his concerns that “we’re not paying attention to a lot of the problems with Medicaid even as we try to expand it”. That’s vague, but fair enough, let’s tax the rich to make the coverage more effective and capable of covering specialty needs that currently aren’t adequately addressed. The NPR coverage apparently hasn’t even considered that, however, as they confusingly describe the eventual inclusion of the control group to Oregon’s medicaid as having happened after the state “found some extra money”. Governments apparently just randomly find funds, like they were pennies in between couch-cushions.
That last note, that thousands of people without medicaid were finally added to the rolls in 2010, is actually the somber note that NPR closes on. Yes, seriously, it talks about how that’s “not so good for the research”. Even if it’s a “good” that people aren’t drowning in medical bankruptcies quite as often or badly or quite so concerned about that risk, it’s about equivalent to the “bad” of research not being able to conclusively prove in at least three different ways that this is a good thing for society. Just like straightspawn talking about marriage equality as the end of motherhood and fatherhood is the same as queerspawn talking about it as ending the stigma against their families.
Quick, NPR, tell me what color you think this text is written in?