tag:blogger.com,1999:blog-6837159629100463303.post5887692340654132250..comments2023-06-18T01:25:08.748-07:00Comments on Information Transfer Economics: Leaning over backwards: health care editionJason Smithhttp://www.blogger.com/profile/12680061127040420047noreply@blogger.comBlogger12125tag:blogger.com,1999:blog-6837159629100463303.post-51951628882775580582022-12-12T17:01:43.220-08:002022-12-12T17:01:43.220-08:00Hey, I read some other RCA posts (mainly about ach...Hey, I read some other RCA posts (mainly about achievement gap stuff) and his arguments seem maybe solid but I want a good rebuttal. You don't need to make a post, just give me examples for where he gets wrong (other then the post you responded to).Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-6837159629100463303.post-41716476932428424662020-02-23T10:51:29.790-08:002020-02-23T10:51:29.790-08:00Since income is larger in level, a 1% rise in inco...Since income is larger in level, a 1% rise in income is larger than a 1.8% rise in health care spending until health care spending is about 50% of income.<br /><br />At €10,000 income and €1,000 health care spending, if income rises €100 then health care spending rises €18 — leaving €82.<br /><br />At €100,000 income and €50,000 health care spending, if income rises €1,000, spending rises €900 leaving only €100.<br /><br />And yes, at some point the curve would have to bend away from the y = x line instead of intersecting it — preventing health care spending from being more than %100 of income.Jason Smithhttps://www.blogger.com/profile/12680061127040420047noreply@blogger.comtag:blogger.com,1999:blog-6837159629100463303.post-52421926572876003392020-02-23T07:23:52.311-08:002020-02-23T07:23:52.311-08:00Thanks for the additional chart. I have a followu...Thanks for the additional chart. I have a followup question.<br /><br />RCA says 'the slope in the health expenditure-income plot below implies that a 1% increase in underlying value on the x-axis (income) predicts a 1.8% increase in value on the y-axis (health spending).'<br /><br />This seems to be saying that as income rises any additional income will be increasingly spent on healthcare. While this does mean that health care spending rises faster than income I think it prevents healthcare from ever reaching 100% of spending and indeed it would mean that as income increases people get richer both in terms of healthcare and other goods (at least until the point where all additional income is send on healthcare). This would prevent healthcare spending from ever becoming 'unaffordable' (at least for society as a whole).<br /><br />Am I interpreting RCA's statement correctly ? Are his charts perhaps inconsistent with this statement ? robhttps://www.blogger.com/profile/04682517711551179057noreply@blogger.comtag:blogger.com,1999:blog-6837159629100463303.post-14913711113656995132020-02-22T14:42:02.640-08:002020-02-22T14:42:02.640-08:00Yes, eventually. In fact, he says that on his blog...Yes, eventually. In fact, he says that on his blog:<br /><br />"... it’s entirely possible to spend an ever-larger share of income on health without reducing our real consumption elsewhere."<br /><br />i.e. Eventually we will make $500,000/year median income and spend $450,000 of it on health care.<br /><br />The disposable income measure he uses (i.e. AIC as a proxy for "disposable income") on the x-axis includes health care, so (aside from the funny bit about regressing x on x that I didn't get into) the values on the y-axis cannot exceed the values of the x-axis — putting essentially a wall at the y = x line.<br /><br />I'll add a graph above in update III.Jason Smithhttps://www.blogger.com/profile/12680061127040420047noreply@blogger.comtag:blogger.com,1999:blog-6837159629100463303.post-64320512673165437532020-02-22T07:41:32.121-08:002020-02-22T07:41:32.121-08:00I have a question on: 'Let's say RCA's...I have a question on: 'Let's say RCA's claim is true that the US is not an outlier — and that globally health care spending rises twice as fast as income''<br /><br />Am I correct in thinking this would (if continued indefinitely) lead to 100% of income being spent on healthcare - at which point the trend would have to end ? robhttps://www.blogger.com/profile/04682517711551179057noreply@blogger.comtag:blogger.com,1999:blog-6837159629100463303.post-51144231580063651202020-02-22T04:49:24.950-08:002020-02-22T04:49:24.950-08:00That was amazing, Jason. Thank you!That was amazing, Jason. Thank you!Vasu Malhotrahttps://www.blogger.com/profile/03996642953789719400noreply@blogger.comtag:blogger.com,1999:blog-6837159629100463303.post-77041229514680835822020-02-21T18:51:02.352-08:002020-02-21T18:51:02.352-08:00Are you equating health care with entertainment?
...Are you equating health care with entertainment?<br /><br />What is it with libertarian weirdos getting surgery for fun?Jason Smithhttps://www.blogger.com/profile/12680061127040420047noreply@blogger.comtag:blogger.com,1999:blog-6837159629100463303.post-41415121077267457602020-02-21T18:49:01.998-08:002020-02-21T18:49:01.998-08:00Thanks ... and *lol* :)Thanks ... and *lol* :)Jason Smithhttps://www.blogger.com/profile/12680061127040420047noreply@blogger.comtag:blogger.com,1999:blog-6837159629100463303.post-40402057158201318832020-02-21T18:48:05.459-08:002020-02-21T18:48:05.459-08:00So the defense of RCA's analysis based on regr...So the defense of RCA's analysis based on regressions is to say the regressions are irrelevant?<br /><br />And (1) is kind of an empty statement. Health spending could easily grow faster or slower than income and still "rise with income". The latter is not a problem, which the former is.<br /><br />The entire point of the blog post is that only way you can draw the conclusion (2) is by assuming conclusion (2) — it's question begging.Jason Smithhttps://www.blogger.com/profile/12680061127040420047noreply@blogger.comtag:blogger.com,1999:blog-6837159629100463303.post-26495939984919961652020-02-21T15:52:53.142-08:002020-02-21T15:52:53.142-08:00I think it is Jason who is missing the forest for ...I think it is Jason who is missing the forest for the trees.<br /><br />You could remove the regression lines entirely and the "forest" would still be apparent: that (1) medical spending increases strongly with household income, and (2) this implies that high USA spending is not strong evidence of some unique disfunciton in the USA health system.<br /><br />And I don't know why Jason says health care spending rising faster than income is neccessarily an "enourmous global problem." Lots of categories rise faster than income, for good reasons.edhttps://www.blogger.com/profile/01150091053740909530noreply@blogger.comtag:blogger.com,1999:blog-6837159629100463303.post-19114523893724174472020-02-21T15:36:45.681-08:002020-02-21T15:36:45.681-08:00Entertainment expense are growing faster than inco...Entertainment expense are growing faster than income (despite diminishing returns in the amount of fun) and that a problem <br />You, 1900Dubynoreply@blogger.comtag:blogger.com,1999:blog-6837159629100463303.post-927119466159090622020-02-21T15:03:36.132-08:002020-02-21T15:03:36.132-08:00Overjoyed to finally see someone publish a good an...Overjoyed to finally see someone publish a good analysis of why the (long, long, *long*) RCA posts miss the forest for the trees. RCA has all the hallmarks of a programmer trying to do statistics (aka 'data science'): attractive plots, boundless confidence, and an inability to think deeply about causal structures within your data, or what precisely the math in the models you are using is actually doing.Lydia Mnoreply@blogger.com