3 ms·
Yeah, that's the problem, that's some crappy data - it's a whole bunch of correlation, and not a lot of causation. I want to see some actual data that suggests
by cheald 9y ago
Yeah, that's the problem, that's some crappy data - it's a whole bunch of correlation, and not a lot of causation. I want to see some actual data that suggests that the cause was the ACA, not a recovering economy post-massive recession.
That article uses two crazy misleading graphs: One starting in 2010 (right after the recession, when personal bankruptcy had just peaked) and one starting in 2006 (right after the passage of BAPCPA). There was a massive dip in personal bankruptcies in 2006 (2.1m filed in 2005 because of people rushing to file in 2005 before BAPCPA, 600k filed in 2006) and a massive spike in bankruptcies during the recession (peaking at 1.6m in 2010).
Using those data as bounds, then suggesting that the ACA and not bankruptcy legislation or the prevailing economic winds are the cause is statistical malpractice.
- sjg007 9y agoRead the primary research papers cited in the article. https://assets.documentcloud.org/documents/2153989/daniel-austin-medical-bankruptcies.pdf https://assets.documentcloud.org/documents/2153989/daniel-au...
- cheald 9y agoDid you read it? It does quite successfully make the case that Massachusetts' healthcare system has probably improved medical bankruptcy rates. It doesn't make any such claim for the ACA at large. In fact, it possibly suggests the contrary - look at Tables 3 and 4 - their estimation is that overall, the percentage of bankruptcies that are medical in nature has risen, and the distribution of medical bankruptcies in the "basket" jurisdictions - the percentage of bankruptcies reporting medical debt increased sharply post-2010! This paper even specifically calls out the Himmelstein paper (which was broadly used as a justification for the ACA) as having far to broad a definition of medical bankruptcy. I am very willing to accept that the ACA reduced medical bankruptcy rates (it makes intuitive sense!), but I maintain that there's practically no data that validates the claim, and the way that the statistics are manipulated to support that point via correlative inference makes me especially suspicious that it's not actually happening.