HN user
http://randomcriticalanalysis.com
Outcomes are determined by more than health system inputs alone (indeed, it has little to do with health care in the developed world), and costs have much to do with the intensity of health care. Rich countries spend more because they can, not because they need to.
This is my blog. I'll check comments later on and respond if anyone has serious questions or comments.
Your findings are not mutually exclusive with other theories about the underlying causes of high U.S. healthcare expenditure, such as a chronic lack of price transparency.
My findings strongly agitate against the notion that high US health spending is a product of idiosyncratic features of our system. Presumably most of these critics believe we'd spend much less if only our system looked more like other countries, but my evidence indicates we'd spend very similar amounts in the long run regardless. Further, we'd likely have similar outcomes and many other similar healthcare attributes (prices, intensity, health worker density, etc). Most of the things about US healthcare people believe to be important and unique (i.e., not explained by income) just aren't.
The more money is in the pot, the greater the incentives to pilfer it, and, in lieu of adequate controls, the more it will be pilfered
I wouldn't argue there's no "pilfering" or that more money doesn't create more opportunity for this, but the high income elasticity likely has little to do with pilfering. Where are these ill-gotten gains going? I certainly don't think you'd have much success in showing this if you look at, say, the growth in physician incomes, pharma/biotech industry profits, and so on. The data are much more consistent with mundane explanations like rising technological sophistication, higher intensity, and so on (ultimately much of this being driven by some combination of patient/family demand and providers' "spare no expense" approach to caring). I mean, if you look at the economic data it's quite obvious most of the increase can be arithmetically attributed to a swelling of health workers (density or share of workforce) and that most of these workers have lower-to-middle income levels (especially on the margin).
https://twitter.com/RCAFDM/status/1193949748841111552
I surmise that the accelerating nature of health expenditures... is due to the rapid inflation of disposable income relative to overall income at the higher levels.
I'm not sure what you mean by this exactly, but a better, more parsimonous way to understand high income elasticity is that higher income countries are usually inherently more productive countries. We can spend substantially smaller shares of our income on food, clothing, shelter, and other "necessities" because we are able to produce these things so much more efficiently (or otherwise procure on the market) than we did decades earlier or than OECD countries of more significantly more modest income levels while still consuming more of these things in real terms. This frees up resources to be spent on higher order wants like health, education, recreation, culture, and so on. Many of these growth areas, meanwhile, are inherently subject to less productivity growth, meaning prices tend not to fall relative to incomes nearly as quickly as we observe in other sectors.
https://randomcriticalanalysis.com/2019/12/03/no-means-no-th...
I say this with no irony: a linear regression between income and amounts extorted during kidnapping, controlling for other variables, would show similar results.
I doubt that's true, though we're talking about the total spending (kidnapping ransom) per capita here. It's pretty clear the price per transaction (as in, health inflation) explains very little in US time series or cross-sectionally. Now maybe if kidnapping started to become a high amenity affair in developed countries (presuming this sort of thing happened with measurable frequency here) and 25% of the population worked delivering these services.....
We spend 17.4% of GDP on healthcare. The OECD average is 9.5%
Health spending is almost entirley explained by income levels, especially in the long-run. The US spends much more because the US is much richer than most and because health spending is highly elastic at a national level.
# TL;DR https://i0.wp.com/randomcriticalanalysis.com/wp-content/uplo...
# Long explanation
https://randomcriticalanalysis.com/2018/11/19/why-everything...
and we don't have better outcomes to show for it
Norway and Luxembourg also spend 2x Spain and Italy and don't have more to show for it either despite the fact that they're also much richer, have larger welfare states, etc.
https://i1.wp.com/randomcriticalanalysis.com/wp-content/uplo...
Countries increase health spending because they can, not necessarily because they need to. Evidence strongly suggests returns to health spending are falling everywhere and the US isn't particularly unique in this regard.
https://randomcriticalanalysis.com/2019/11/07/a-tale-of-two-...
Obesity is not an independent variable
Nothing is perfect, but most experts believe this has little to do with healthcare today because healthcare interventions tend not to be effective causes of long-run weight loss and most countries aren't doing enough of the stuff likely to have large effects (e.g., surgical interventions) to explain much of the variance. Even if you could argue it might explain something, say 0.5 mean BMI points, other factors are clearly highly important. Cultural * and genetic factors are likely to play a significant role amongst high-income countries. Further, obesity rates rise with time and income levels despite higher health spending.
https://i0.wp.com/randomcriticalanalysis.com/wp-content/uplo...
socialized healthcare does a better job of controlling obesity with preventative health measures
evidence?
As it is now in the U.S., patients only go to medical professionals when there is a problem
The US spends more on preventive medicine than almost any other country, though preventative medicine generally has very-small-to-modest effects on outcomes and rarely, if ever, saves money (usually quite the other way around)
~ RCA
note: * some of these "cultural" factors may be residual economic influences... the US escaped the malthusian trap long before almost all other high-income countries and this may have latent effects on attitudes towards food, diet, etc)
The obesity explanation doesn't pass the sniff test. Canada is 12% less obese than the U.S., but it spends 6% less of its GDP on healthcare.
It's my blog (RCA). My argument is that obesity substantially explains US health outcomes in relation to other countries. I never claimed obesity is the cause of high national health spending (as in, "inputs"). To the contrary, I have consistently argued US health spending is well explained by its wealth (technically income levels).
https://randomcriticalanalysis.com/2018/11/19/why-everything...
To a first approximation, national health spending is entirely explained by the average house income level in the long run. While time, healthcare technology, and other factors are assocatied with rising spending, these changes are ultimately very well explained by changing income levels. Amongst high-income countries, a 1% increase in income is robustly associated with a long run increase of about 1.8% (it's highly elastic).
https://i0.wp.com/randomcriticalanalysis.com/wp-content/uplo...
The US spends more than Canada because it's still a much richer country (which isn't to say Canada isn't a nice place!).
That is beyond the realm of believability, even if I introduce the other population-induced causal factors which you implied but didn't specify.
Again, I never said this, but other population health risk factors such as age structure, disease rates, and the like are of negligible significance when it comes to long run aggregate spending. Such factors may be highly predictive within countries and may have some say in the short run (within budgetary constraints), but in the long run national picture the evidence suggests these factors amount to little more than noise. National household income levels trumps everything.
Additionally, the government would be more invested in the population's health under a single-payer model.
US government programs, namely Medicare and Medicaid, spend more on healthcare than most other high-income countries do in total (even more so comparing public-to-public). Just how much more incentive do we need before these magical effects kick in? Higher health spending predicts higher obesity rates in time series and cross-sectionally (though this is likely ultimately mediated by long-run income levels and by time).
https://i0.wp.com/randomcriticalanalysis.com/wp-content/uplo...
Where is the evidence that these programs have large, sustained effects and are cost effective? Most data indicate these programs have negligible effects in the long run and they almost always cost more than they save (which isn't to say we shouldn't necessarily do it, but the economic rationale is v. weak).
~ RCA
a) many (perhaps the majority, don't have time to source now) of the bankruptcies in the US are due to medical issues and spending
The vast majority of medical bankruptcies have nothing to do with the cost of medical care, but the disruption to career/income flow imposed by illness. This is clearly a problem in other countries as well.
https://www.nejm.org/doi/pdf/10.1056/NEJMp1716604
The rate of growth of US spending on healthcare....far outstrips the rate of growth of money
This is true in other OECD countries too. Further, this doesn't mean what you think it does. As our productivity rises, the share spent on consumption categories with high productivity growth (increasingly low relative prices) can decline, which frees up spending to be spent on health and other areas subject to less productivity growth (the majority of the expenditure growth corresponds to rising real health consumption tho)
https://randomcriticalanalysis.com/2019/12/03/no-means-no-th...
The same medicines by the same manufacturers and often the same production lines ... often cost 10-100 times more in the US than they do in other places
One might be able to find outliers of this sort, but that clearly doesn't reflect anything close to central tendencies (mean, median, mode, etc), especially when compared (accurately) to other high-income countries. Richer countries, like the US, generally pay relatively higher prices.
The US may pay a somewhat higher premium, but there are tradeoffs here vis-a-vis incentivizing innovation in the long run. It's also not widely appreciated that the US pays markedly less for generics....
I'm not a language zealot (most languages/frameworks come with pros and cons) and I use R and the tidyverse quite regularly, but "performant" is not a word I'd associate with the tidyverse. I'd be surprised if it wasn't slower than most alternatives (can't claim to have systematically benchmarked it, still....), even if it's often easier to use and usually "good enough."
Yet 81 percent of the leading health economists agreed with the statement, “The primary reason for the increase in the health sector’s share of GDP over the past 30 years is technological change in medicine"
I've highlighted the role of technological change on my blog before, but technological change is a major proximate cause. The root cause is income growth. Countries are chasing these technological advancements in direct proportion to their income and the degree to which they can afford them (as prices fall, countries with lower real incomes are more able to afford technologies the US and other rich countries had long before, but the frontier has long since moved on....). If you throw year fixed-effects or a time trend into analysis the coefficient on time (a proxy for tech chg) is very modest and the income effects are virtually identical.
Thanks. I haven't seriously considered that before, but I may put one up!
1. You didn’t.
https://i0.wp.com/randomcriticalanalysis.com/wp-content/uplo...
2. Asymptotic growth towards 100% certainly sounds unsustainable.
"Sounds" isn't an argument and I'm explicitly arguing the slope is likely to flatten, eventually. The point the income elasticity of health expenditure can be (is) well north of one and we can (and do) consume more of everything else at the same time.
https://randomcriticalanalysis.com/2019/12/03/no-means-no-th...
Log-log overfitting
That isn't a thing. You can argue this specification minimizes the residuals at the high end if you want, but it's very likely to the correct modeling decision, it's bog standard in economics, and the US residual ~= 0 (certainly not notably high). It's also pretty obvious health expenditures are increasing in % terms and that failing to log-transform results in particularly poor model performance with constant slope out of sample.
https://i0.wp.com/randomcriticalanalysis.com/wp-content/uplo...
So if you have data about healthcare prices not being out of order, that seems a lot more relevant
https://randomcriticalanalysis.com/2018/01/06/its-not-the-pr...
https://randomcriticalanalysis.com/2017/07/27/health-care-pr...
The lower third of this country can’t afford essential care.
The lower third of the country consumes approximately the same amount of care as the rich, as in other high-income countries, and the socioeconomic gaps in other countries are likely comparable to even larger (depending on how measured).
https://randomcriticalanalysis.com/2017/04/15/some-useful-da...
https://twitter.com/RCAFDM/status/1203715358152167424
To the extent there are real and ultimately consequential issues with affordability for some small segment of our population, these aren't likely to be explained by aggregate costs or prices so much as by narrow details that we can tweak, i.e., without requiring massive change, should the political desire exist to do so.
seems a lot more relevant— and less cherry-picked — than the history of household spending on food consumption.
Pardon me, but I was engaging with someone that was arguing this expenditure growth implied starvation and you're making very similar (wrong) arguments. Whether you appreciate it or not, the role of general increases in productivity, the source of real income growth in the long run, and differences in the rate of productivity growth in different sectors, which we are clearly reflected in prices, is very much on point. Food production is simply a way to make this concrete for people that struggle with abstractions like price indexes and relative prices.
I (RCA) have been meaning to dedicate a blog post to this topic exclusively and quantify the root causes in considerable detail.
In short, yes, I believe globalization and, specifically, multinational corporations (affiliates of foreign owned corporations) are the primary cause. However, in regression terms, this has less to do with net income flows into the US than net income flows out of several small high GDP/person countries. The US is a very large, very rich country whereas many of these countries are much smaller so it doesn't take much to have outsize affects on them. Even if 100% of this ultimately accrues to the benefit of US households, (which I don't think is quite accurate) the effect in % terms is vastly different.
Although some of this can be directly observed as primary income flows in the current year national accounts (dividends, rent, etc), some of it shows up as gross savings or disposable income in the foreign affliates, i.e., it's equivalent to retained earnings. Over the past decade or two non-financial corporate savings have increased massively and these savings are largely uncorrelated with domestic (capital) investment, i.e., it's almost entirely financial and it's largely leaving these countries in the form of net lending. These things also influence the calculation of (GDP) PPPs and cause other headaches.
Of course, there are also other reasons why GDP misleads. For example, Luxembourg has a very large non-resident workforce. Something like 50% of their workers live in neighboring countries (varies year to year), meaning ~50% of aggregate employee compensation goes home (cross border) to Germany and the like. Then there are petro-states like Norway whose income flows are inherently temporary (finite amount of natural resources to extract) and highly volatile, meaning they can't consume out of their measured GDP like most other countries. They need to practice massive consumption smoothing if they don't want their standard of living to crash in the not too distant future.
Long story short, GDP was never intended to be an indicator of material wellbeing and the household perspective (consumption, disposable income) are better measured and more reliable. One might try to throw a bunch of variables in to counteract the many issues imposed by GDP as a proxy for the household perspective, but why bother?
~ RCA (sorry for typos, grammatical errors, etc... limited time to comment and would rather focus more effort on blog)
The point is that almost all of what people have termed "medical bankruptcy" has approximately nothing to do with healthcare costs. It has to do with income/career disruption associated with their health, so these issues play out to a similar degree in other countries for similar reasons.
While I can believe there is somewhat higher variance in the US system in terms of bills/prices/etc, which sometimes affects people meaningfully, it's also worth pointing out that out of pocket costs aren't unusually high in the United States. I mean, we might have a slightly larger problem with these, but that's much more of an implementation detail in terms of who pays and how much than the sort of top-down policy proposals some people seem to believe are necessary (because they've badly misdiagnosed the situation or are ideologically blinkered)
One can come up with a narrative to support almost any argument. The question is, do you have data to support it, is it credible, and how much can it actually explain? The truth is richer countries tend to be significantly fatter countries and countries are getting fatter, even though these same countries also spend significantly (increasingly) more money on healthcare and a variety of other social services. These are empirically verifiable points that actually explain things and make quite a lot of sense theoretically.
The high average caloric consumption in the US is likely substantially explained by our income levels. Humans are not genetically adapted to living in an age of cheap, plentiful, readily available, and highly palatable food. Above and beyond high average income levels and particularly affordable food prices, countries also differ in a number of other dimensions. For example, it's quite clear the US escaped the malthusian trap long before most of Europe because of the ample amount of fertile land settlers had at their disposal. Americans were taller and fatter than Europeans in Europe long before the development of the welfare state. This is also verifiable. It's quite possible such differences lingering effects on dietary preferences, serving sizes, and other variables that impact food consumption.
Further, there are clearly large spatial differences in the United States and they're unlikely to be well explained by the sorts of policy you mentioned. Parts of the US settled by people that migrated from periphery of the UK (Scotland, N. Ireland, Wales, and Northern England) suffer from these issues much more than most Americans and more than can be explained by observables like income, education, and the like. This residual also seems to play out in the UK today
https://www.bmj.com/content/342/bmj.d508.full.pdf+html
I've done my own analysis and found this maps well to genetic population structure (not yet published), but whether this is genetic or some sort of subtle cultural phenomenon is a different question.
While I'm quite skeptical that existing non-medical programs of that sort have significant effects (evidence?) or even vary nearly as much as you seem to believe (US affluence has implications for purchasing power of all households, even allowing for somewhat higher inqequality, and size of welfare state often underestimated), I think you need to explain how it is that these higher income, higher spending countries with larger welfare states achieve results no better (probably somewhat worse) than much poorer countries.
https://i1.wp.com/randomcriticalanalysis.com/wp-content/uplo...
Also, you should probably also be able to explain why it is that socioeconomic gaps in life expectancy and other health outcomes are probably at least as large in Finland, Norway, and the like.
1, I can fit the US on a linear trend amongst high income countries.
2, there’s no necessary reason why increasing health share with rising real income is unsustainable. We can and have increased share spent on health while increasing real expenditures across the board.
3. I touch on some reasons why this may curve up and then eventually flatten out.
4. No, not everything fits on log-log slopes and us is very close to the trend.
https://randomcriticalanalysis.com/2019/12/03/no-means-no-th...
I’m not sure what you’re getting at. Obviously, healthcare must be paid for somehow and it’s a fair assumption workers bear most of the incidence. However, the role of real incomes and relative prices are generally under appreciated here.ie. It’s not so much that health benefits are something extra as part of total remuneration (ultimately a reflection of rising productivity elsewhere which frees up spending to be allocated to health and other services)
You absolutely can in real terms.
https://randomcriticalanalysis.com/2019/12/03/no-means-no-th...
Also while I agree that 100% of a GDP is a reasonable approximation of an upper limit in practice, that's not necessarily strictly true either. GDP measures domestic production, not the income available to residents. The income available to residents can certainly be much less than GDP (e.g., Ireland, Luxembourg, etc), so the reverse can be true in principle too.
My analysis strongly suggests costs strongly track with income levels. Relative Prices rise with income, but purchasing power also rises, so it's not necessarily less affordable relative to incomes actually earned domestically.
https://randomcriticalanalysis.com/2019/12/03/no-means-no-th...
Some people may prefer a regime that offers more insulation (socialization) from the true costs (generally overstated IMO), but that's something of a different problem with different potential solutions than the popular notion that US prices are inexplicably high or outcomes inexplicably bad where we're (presumably) likely to win-win-win if only we copy the regimes found elsewhere.
Yes, I'm the author of the blog (RCA). No, I haven't studied this narrow question, though it's somewhat tangential to the arguments I've advanced (outcomes; prices; aggregate costs; etc). I have, however, seen several studies that suggest "medical bankruptcies" have very little to do with medical debt and much to do with the effects of their health on their employment/earned income (also similar patterns in Canada and some other comps in the available data...)
US also has "programs" and generally spends more on preventative medicine. Further, actual medical evidence suggests medicine doesn't have good treatments for obesity (getting people to lose much weight and keep it off), save the surgical inventions (which aren't used often enough to explain much). It's also clear that obesity rates are rising throughout the developed world despite rising spending.
https://i1.wp.com/randomcriticalanalysis.com/wp-content/uplo...
The exact numbers depend on the source/methodology, but all credible estimates show large differences and these differences imply large effects for mortality rates [in the broader population](https://randomcriticalanalysis.com/2019/11/07/a-tale-of-two-...).
No, but this is addressed in my post. Australia's obesity rate is substantially lower, plus they suffer from lower homicide, drug abuse, and the like. More generally, the diminishing returns to health spending imply there's very little to believe the higher expenditure the US makes is likely to purchase appreciably better outcomes (even with different healthcare regimes).
My point is that there's very little to suggest US healthcare is uniquely ineffective. There rapidly diminishing returns to health spending and US outcomes are badly handicapped by factors like obesity and homicide.
Maybe try actually reading it though?
That argument is either wrong or tautological.
There are two distinct claims here.
(1) National health expenditures are overwhelmingly determined by the resources available to households, as measured by consumption or disposable income (not individual income), and these same measures effectively explain why the US spends much more than other countries.
(2) The increase in health spending in the US, as in other countries, has overwhelmingly been driven by an increase in the volume or quantity of healthcare consumed. While healthcare prices have risen they have generally not risen much faster than incomes have in the long run (especially not overall, but even when disaggregated somewhat), and thus cannot substantially explain why the share of income or consumption allocated to healthcare has risen overtime (indeed, the indices published by CMS and BEA indicate that had the quantity of health per capita remained constant the share would have actually fallen over time). This is also mirrored by related indicators, such as the vast rise in the health workforce, slow growing wages in healthcare (average healthcare wage approximates national average and hasn't increased for decades), and so on.
The US spends more on health care for worse outcomes
So-called health "outcomes" are substantially endogenous (lifestyle, social influences, etc) and these factors vary systematically between countries. Moreover, health expenditure is almost certainly subject to rapidly diminishing returns and there is little to nothing to suggest the broad outcomes people typically cite (e.g., life expectancy) improve systematically with expenditures amongst countries in the developed world.
https://i2.wp.com/pbs.twimg.com/media/DUwPBxoXUAEMCxy.jpg
The pattern of rising expenditure with little or nothing to show for it on the margins in these broad measures may be regrettable, but my point is largely being driven by increasing disposable income, as opposed to idiosyncratic features of the US healthcare system. It sure seems like people really don't like to cut spending in practice and ultimately reject aggressive rationing conditional on income levels.
"Using average income, a clearly bad statistic, can't be justified by saying you could not find a good statistic."
As lliamander already mentioned, I addressed this on my blog. Several quick points:
1) International comparisons of health spending overwhelmingly relate directly to mean health expenditures (e.g., per capita, % of GDP, etc). Even if you know nothing of this area it seems most reasonable to compare mean health spending to mean material living conditions (disposable income or consumption)!
2) Individual income and health expenditures on behalf of individuals (i.e., not just OOP) are effectively uncorrelated within countries in the developed world (including the United States). There are clearly large national level effects independent on one's place in the national income distribution, which are exceptionally well correlated with mean income (r^2>0.9), and we shouldn't expect to income of the median person to be terribly informative.
3) We know healthcare to be heavily socialized in the United States and throughout the developed world. Mean income is a much better indicator of overall financial wherewithal than median income. The median just doesn't make much sense from a theoretical point of view. Moreover, estimates of the income distribution and the medians are surely subject to greater measurement error and issues of comparability between countries, so there are practical issues with this as well.
4) More practically speaking, I've run many different regressions using medians, various indicators of income inequality, etc and found little to nothing to suggest the income distribution is an important independent predictor. The r-squared and various goodness of fit indicators suggest these metrics perform less well than the National Accounts-based means I have used for disposable income and consumption. In multiple regression on these necessarily smaller samples (which are easy to over-fit) the coefficient/effect size of these distribution-related indicators are rarely significant, don't significantly improve model fit, and would almost certainly be rejected in lasso or the equivalent.
https://randomcriticalanalysis.com/2018/01/20/on-the-relevan...
5) With other aggregate consumption/expenditure statistics, such as food, housing, transport, entertainment, etc the fit is also (unsurprisingly) much better with my statistics than cross-sectional comparisons of median income, so I'm not sure why you would expect one of the most socialized and most elastic categories of expenditure at a national level (~1.4 as a function of GDP) to be any more favorable for the distributional perspective.
https://rpubs.com/random_critical_analysis/oecd_2014_consump...
First, we can assess the degree to which the US is unusual by inspecting the scatter plot. If US household income was somehow unusually inflated by health spending, then it its residual should be much larger than it is (especially in better models that incorporate some of the non-linearity observed in other countries).
Second, just where is this extra income supposed to come from? More likely households bear almost all of the incidence of this spending, meaning each dollar spent on health means approximately one dollar less available to households to use elsewhere. This is certainly what the literature on employer health insurance benefits tends to suggest.
Third, the timing of this is all wrong. It’s clear the arrow of causation goes overwhelmingly from income to health spending because changes in health spending clearly follow changes in income. Over the past few decades it takes an average of about 2-3 years for changes income to be fully reflected in changes in health spending due to the large role played by 3rd party payers, employers, etc (the current year elasticity is about 0.2 whereas the long-term elasticity is north of 1.6).
~RCA
P.S.: It’s my blog.
It’s really not in this context. The median household vis-a-vis income isn’t necessarily the median household vis-a-vis health consumption. More generally, health spending is massively subsidized by society writ large. Society’s willingness and capacity to pay is much better indicated by average income than the income of the median household (never mind issues pertaining to measurement of the income distribution within countries!). Furthermore, the vast majority of health expenditures in most high-income countries are spent on a small fraction of the population in any given year (those that are very sick, badly injured, etc.)
But if you look carefully you will find most of this behavior plays out similarly in those few countries with somewhat more comparable wealth (e.g., Norway). The issues that many people think of as being particularly "American" problems are problems that mostly have to do with our high material living conditions (as in, diminishing marginal utility attached to other forms of consumption and increasingly high values attached to human life). Overall health expenditures are increasing equally rapidly in these countries as a function of AIC, disposable income, etc and there is no evidence they are getting any more value for their money (having a healthier populations helps naive comparisons of life expectancy, IMR, etc a ton).
Here are some papers that touch on some of these more fundamental theoretical issues:
Except the rich and poor spend pretty similarly in the US and elsewhere. The income elasticity > 1 occurs at a national level....
http://mason.gmu.edu/~rhanson/EC496/SOURCES/LUXURY.PDF
And the better data on price suggests US actually isn't unusually expensive for a high income country and so on and so forth. Much data points to higher volumes of health goods and services as we chase diminishing returns as a society.