@lymanstoneky @Noahpinion @mileskimball Thank you. My argument is high (disposable) incomes lead to high quantities per capita and higher real prices (Baumol effect), but it's mostly quantities. Reliable domestic and intl price indices suggest prices don't rise that much faster than incomes. 
@lymanstoneky @Noahpinion @mileskimball The prices people cite for international comparisons are invariably sourced from the International Federation of Health Plans (IFHP), a payer trade group. They're fairly opaque on the particulars, but the details they do provide makes it apparent their methods are unreliable.
@lymanstoneky @Noahpinion @mileskimball Many of the (proxy) statistics use to compare international quantities of care, such of drs office visits, are very weak today. Yes, they fail to explain US health spending, but they also fail to systematically explain health spending in other high-income countries too.
@lymanstoneky @Noahpinion @mileskimball I argue the reason why US spends more and the reason why spending increases strongly with incomes in the long run has little to do with prices, people getting sicker, or higher propensity to seek care; it's mostly a product of higher intensity of care.
@lymanstoneky @Noahpinion @mileskimball As incomes rise, societies become increasingly willing to pay for more involved, more cutting edge healthcare. This involves not just spending more on equipment (capex), drugs, etc. It also involves a lot of more people to provide this care, to manage higher complexity, etc.
@lymanstoneky @Noahpinion @mileskimball Though OECD, WHO, and company don't provide many good indicators to compare cuttng edge care directly, we see far more relationship between spending levels and healthcare technology (MRIs, CT scans, etc).
@lymanstoneky @Noahpinion @mileskimball We also see evidence of this in cutting edge procedures (e.g., novel transplants), orphan drugs, and the like.
@lymanstoneky @Noahpinion @mileskimball We also see evidence for higher intensity in proxies for overall inputs, like health worker density or share of the workforce. US health workforce much larger, proportionally speaking, than much lower-income countries.
@lymanstoneky @Noahpinion @mileskimball Indeed, in one were to simply divide aggregates of health spending from SNA or SHA we'd find overall expenditure, value-add or, more narrowly, employee compensation in healthcare well explained by the growth of the workforce.
@lymanstoneky @Noahpinion @mileskimball That is, while real healthcare compensation is higher in higher-income countries, overall compensation or spending per employee isn't rising. Clearly, most of this growth isn't well explained by rising physician density or higher remuneration for physicians.
@lymanstoneky @Noahpinion @mileskimball Relative to the average wage in other parts of the economy healthcare isn't remarkable in the US or in other high-income countries (decreasingly so as healthcare expands to include more and more people).
@lymanstoneky @Noahpinion @mileskimball To return to prices briefly, while OECD health PPP and related research finds a rather strong relationship between income levels and price of procedures (generally, "hospital services"). None of the procedure prices IFHP purports to show correlate as expected.
@lymanstoneky @Noahpinion @mileskimball (Even with the US excluded that is!) IFHP's efforts, tho *widely* cited, are too trivial for the task and they make several weird choices in how they present the data (e.g., using exchange rates rather than PPPs).
@lymanstoneky @Noahpinion @mileskimball They also make weird choices that affect the US specifically, like comparing "prices" for US DRG (diagnostic) codes, which include a bunch of ancillary services, to ICD-10 procedures, which clearly don't. It's far from an "apples-to-apples" comparison, esp. for US.
@lymanstoneky @Noahpinion @mileskimball Making apples-to-apples (measurement invariant) price comparisons cross country is non-trivial. So far as I am concerned, OECD health PPPs and related research efforts are the only reliable health prices that have been published publically.
@lymanstoneky @Noahpinion @mileskimball They have provided prices (quasi-prices) for specific procedures w/ relevant patient attributes (vignettes). Results are consistent with prices observed for their hospital services PPPs (also indicates US prices aren't inexplicably high)
i2.wp.com/randomcritical…
i2.wp.com/randomcritical…
@lymanstoneky @Noahpinion @mileskimball The earliest health PPPs were likely too crude to be useful, but the latest benchmark reports and research reports have reasonable methods and produce credible results IMO.
oecd.org/sdd/prices-ppp…
oecd.org/sdd/prices-ppp…
@lymanstoneky @Noahpinion @mileskimball A fairly recent summary report if you're interested in a quick overview. Though I strongly recommend reading the health PPP manual and the working papers around it if you want to better understand the issues and the details.
oecd.org/health/health-…
oecd.org/health/health-…
