On Wednesday evening I attended a talk given by Dean Baker at the Middlebury Institute of International Studies (MIIS). You might think it’s in Middlebury, Vermont. It’s not. It’s in Monterey, California. Dean is co-founder of the Center for Economic Policy Research. His topic was “Health Care and the Future of the National Deficit.” In his talk he showed that if we had German-level per capita spending on health care, our federal budget in the future would move from deficit to surplus over a number of years. (I’ve forgotten how many. I took detailed notes but somehow forgot to write down dates.) I don’t know if he was saying we would also need German-level increases in per capita spending and, if so, I don’t know where he got his data on projected German-level increases.
A large part of his talk was about how to get to lower health care spending and lower growth of spending. In each case, though, he didn’t use the word “spending” although that’s what he was clearly talking about. He used the word “cost.” There’s an overlap, but they are different. He advocated a number of policies:
. Letting in more doctors from other countries.
. Allowing nurse practitioners to do more procedures on which doctors now have a monopoly.
. Having the government negotiate drug prices, just as governments in other countries do. (He didn’t say whether he would have price controls but I think he used that terminology at one point.) He would also end patents on drugs.
He also argued that even though it would be hard to get Medicare For All past the Supreme Court, it would be desirable. He compared administrative costs for Medicare (about 2% of the Medicare budget) with administrative costs for private insurers (about 19% of premiums.) He didn’t mention that the amount of fraud in Medicare is huge and that having more spent on administration would almost certainly reduce fraud.
How would we get more drugs if we didn’t have patents? He at least recognized the problem. His solution would be to have the feds spend $10 billion to $15 billion a year on research. Whatever they came up with would be patent free and would be revealed on line so that anyone could manufacture it. He didn’t say why you could trust the government to choose the right things to spend on and the right people to fund. He seemed to think that would be straightforward, but I don’t why he thinks that.
In Q&A, I said that I agreed with him on reducing barriers to immigrating doctors, but that he had left out another major barrier. I pointed out that the major cause of the huge time it takes to get a drug to market and the cause of many hundreds of millions of dollars spent is due to the 1962 Kefauver Amendment (I had forgotten that it is actually the Kefauver Harris Amendment) to the Federal Food, Drug, and Cosmetic Act. That Amendment requires that the drug company show proof of efficacy; previously the company had had to show only proof of safety. (Ironically, the incident that triggered the Amendment was the thalidomide tragedy, which was not at all about efficacy but about safety.) I asked him if he would also favor repealing that law. He said he recognized the point but wouldn’t want to repeal it as long as we have drug patents, because drug companies have incentives to ignore safety and would aggressively market their drugs in spite of safety considerations. He gave Vioxx as an example. He said, though, that if we got rid of patents, he would be fine with repealing the Kefauver Harris Amendment. That’s a victory of sorts, I guess.
Seeing him caused me to dig up my review of his 2010 book, Taking Economics Seriously. I think my review, “Taking Dean Baker Seriously,” stands up well. (And I have no idea why FDA shows up, each time, as fda.) A lot of what he said in the book was in his talk.
Here’s one of my opening paragraphs in that review:
I take Dean Baker seriously. Why? In part, because many on the left do. But mainly because he is a thoughtful economist who has flashes of wisdom and often an independent take on policy issues. Taking Economics Seriously shows some of this wisdom. Some of his proposals for health care, for example, are refreshingly pro-free-market, and he backs them up well. It also, however, shows his tendency to set up policy issues by excluding certain free-market options. The result is that, with some exceptions, he plays economics between the 40-yard lines.
Two more thoughts.
First, one thing that was off-putting was the introduction of Dean Baker by MIIS economics professor Jason Scorse. Scorse’s introduction of Dean was fine but in discussing Dean’s topic, he said there are two major political parties in the United States and one of them, which he didn’t need to name, advocates dealing with future cost increases by taking away people’s health care. That made me wonder how he would treat students in his class who have different views; would he mischaracterize their views as badly as he mischaracterized the Republican Party’s views?
Second, I had always wondered why Dean, in various blog posts over the years, thinks that mainstream reporters on economic issues are very conservative. That point came up Wednesday night in the discussion between Baker and Scorse. I hadn’t realized, until I checked Wikipedia, that Dean wrote his Ph.D. dissertation under Marxist economics professor W. H. Locke Anderson. I think I understand Dean’s baseline better now.
There were some new facts in his talk, though. Some were really interesting points about health care; one was about loss of jobs due to trade with China. He also slammed generational accounting. I’ll deal with these in a subsequent post.
Final mention. Before the talk, I went up and reintroduced myself to Dean, reminding him that he, Michael Boskin, and I had discussed Social Security reform in this 1998 episode of Peter Robinson’s TV show, Uncommon Knowledge.
READER COMMENTS
Steve Fritzinger
Mar 1 2019 at 9:43pm
When discussing Medicare’s administrative costs did he mention that Medicare outsources its billing to the IRS?
You don’t get to ignore those costs just because they are imposed by another agency.
David Henderson
Mar 1 2019 at 10:10pm
You wrote:
Good point. And with a typical deadweight loss being 30% of the revenue raised, this makes those costs much higher.
Mark Z
Mar 1 2019 at 10:43pm
I assume he also ignored the fact that medicare isn’t an insurance company. A lot of people inexplicably do this. Insurance companies mainly generate profit (as I understand it) by reinvesting the money the collect through premiums, and devote lots of resources to reserving. In the naive comparison between medicare and private insurance, all reserving and investment-related costs, are written off as waste (administrative costs). They’re not. It shouldn’t be difficult to convince an economist that investment is a productive economic activity. In any case, insurance companies aren’t just insurance companies, they’re financial companies as well. The Medicare-private insurance administrative cost comparison is comparing apples to coconuts.
Lastly, Baker’s numbers are wrong; Medicare administrative costs are more like 6%; while for private insurance the number is 11-14%, not 19%. (https://www.washingtonpost.com/news/fact-checker/wp/2017/09/19/medicare-private-insurance-and-administrative-costs-a-democratic-talking-point/?noredirect=on&utm_term=.1ade85dcaf0f). CMS regulatory filings for individual plans in 2014 actually had admin costs of about 7%. (https://www.forbes.com/sites/theapothecary/2017/09/20/medicare-for-all-would-increase-not-save-administrative-costs/#3133b7bf60ba)
David Henderson
Mar 2 2019 at 9:32am
Good points.
What do you mean by “reserving?”
Mark Z
Mar 2 2019 at 12:09pm
Thanks!
Reserving is (I’m not an actuary so this isn’t quite expert testimony) the practice of determining how much income from premiums needs to be retained in expectation of payment on claims, vs. how much can be reinvested. I think, much like with reserves banks keep, it’s a mixture of 1) making sure you have enough money on hand to pay for claims over all reasonably likely scenarios regarding how many claims you might get in the near future, while not reserving “too much” because you want to invest as much as you can, and 2) complying with complex regulations that say how much you have to insulate yourself regarding incoming claims.
From what actuaries tell me, there may be considerable deadweight loss from regulatory compliance costs, but I don’t think that helps Baker’s point since that’s a state-induced cost rather than a market-induced one.
David Henderson
Mar 2 2019 at 2:18pm
Thanks, Mark Z.
Dylan
Mar 2 2019 at 8:03am
I’m curious where he gets the $10-$15B a year that he thinks government should spend on drug development? That seems a lot lower than the $70+ billion that the industry reportedly spent on R&D in 2017. Other things being equal, this suggests a lot fewer drugs being developed, even if we make the assumption that the government would be just as productive at the development as private companies.
But I also do not understand your recommendation on removing the need to prove efficacy. Safety is kind of a meaningless concept without efficacy, no drugs are completely safe, safety is relative to the benefit they provide. The best cancer drugs are exceedingly not safe, I have two friends that have lost parents to the initial cancer treatment in the last couple of years, without that treatment they likely would have lived for months or possibly years longer. We accept that risk because we know that the drugs, on average, work to extend life a little bit, and for some people they work to extend it by a lot. Most of us would not want to take that risk for a drug that got rid of migraines, and we certainly wouldn’t want to take it for a drug that didn’t work at all. I think you would be surprised by the number of drugs that have a solid scientific basis for working in some indication, appear safe in PhI trials, and then in later trials we find out not only that it doesn’t work, but that it actually makes things worse.
To the best of my knowledge, the only way we really know to find that information out is RCTs in lots of people. I’ve tried to think through a few different alternatives for how you could bring them to market based on just that early PhI safety data…but all of them seem to leave us worse off in overall health and knowledge. Would love to see some alternative that I’m not considering, because I don’t think developing drugs is long term viable the way it is going now.
David Henderson
Mar 2 2019 at 9:35am
You write:
You’re right, and to his credit, Dean Baker did make that point. He was basically saying that you should start with this and still let Pharma carry on its research because we aren’t quickly going to get rid of patents.
You write:
You’re absolutely correct. The dose makes the poison. My sometime co-author Charley Hooper is close to finishing his book on the FDA. So stay tuned for his policy ideas.
Dylan
Mar 2 2019 at 11:10am
Thanks for the clarification. This makes a bit more sense, particularly if he is focused on just increasing the NIH budget that already goes to the earliest stage research right now, and not actually having them try to discover new drugs. That’s a defensible position, especially if you think that one of the issues with drug research productivity is that we have a lack of understanding of the basic science. Unfortunately, I suspect that, like many people who don’t know the details, he thinks that the majority of the work that is done for developing a new drug is already done using NIH funds, and that pharma companies just swoop in for some minor finishing touches, and to reap the big profits.
Look forward to seeing this.
Mark Bahner
Mar 2 2019 at 11:51am
This is way outside my expertise–which never prevents me from commenting–but there are a couple ways that simply having to prove safety, without having to prove efficacy, could be a real help:
There are more new drugs (for cancer, for example) that are targeted to very narrow groups of people, based on those people’s genetic makeup. In the historical method of clinical trials, they might give treatment to 200 people, and 190 show no response, but 10 are completely cured. So by “efficacy” standards, the drug appears to be a failure. And then it’s only considerably after the trial that researchers figure out why the 10 people were completely cured. If the standard is “safety” rather than “efficacy,” the drug can still be introduced.
There are drugs that could fail in efficacy trials for one medical condition, but are then later found to be effective for another medical condition. If the drug was already approved based on safety, it could immediately be used for the second medical condition
I think that even better than proving safety versus efficacy would be to strip the FDA completely of the power to prevent drugs from coming on the market, and to only give FDA the power to advertise their approval of drugs…such as Underwriters Laboratories does. But just proving safety would be a step in the right direction.
Mark Z
Mar 2 2019 at 12:26pm
You beat me to it. This is precisely the case.
It’s worth noting that it’s not even just a matter of whether the drug is effective in clinical trials: drugs can get denied approval even if (as determined by phase III trials) they are effective, but less effective than the standard treatment.
But this is problematic for a few reasons. It’s not even just about drugs possibly being re-purposed for a different disease. A drug may be less effective than the standard treatment for a disease, but more effective for a subgroup, maybe for one gender or another, or people with a particular genotype, even if less effective for the general population (or rather, the population of volunteers for clinical trials). Additionally, the sample sizes for clinical trials – over 100 for phase II and 300 for phase III – can still be small enough that there’s a reasonable possibility of sufficiently effective drugs not performing well enough to meet approval by chance.
David Henderson
Mar 2 2019 at 2:23pm
Thanks, guys.
Just so you know, I wasn’t backing down on the “proof” (quotation marks because it’s not literally proof) of safety vs. “proof” of efficacy point. I have a habit (some people think it’s bad; I think it’s good or I wouldn’t do it) of responding to the narrow question raised, which is what I did with Dylan. But Sam Peltzman’s seminal study in the early 1970s found a dramatic reduction in the number of new drugs after the 1962 Amendment. We had had 24 years of experience with just “proof” of safety.
Dylan
Mar 2 2019 at 3:40pm
These are good points, and this is part of the rationale for making changes to the way clinical trials are run, doing smaller trials in pre-specified subpopulations, adaptive trials, etc. However, one way or another I think we’re still going to want to show that the drug works, and that’s going to take real resources. At least until we get to a point where our understanding of human biology has progressed further than where it is now. It doesn’t have to be the FDA that has the power of approval, but I have trouble thinking of another method that doesn’t come with even worse problems.
As an example, it is very common when a clinical trial fails, to take a second look at the data and find that it seems to work in one subpopulation. This is almost certainly due to random chance more than anything real, yet the company is almost universally able to convince themselves that there is a rationale for why it should work in that group, and then they go out and convince investors to give them millions more to run a second, third, sometimes even forth trial on a failed drug based on this hope. FDA, doctors, and patients all go along with it too because, hey, it’s not their money, and there’s a really small chance it might work.
In a world without some kind of gatekeeper, I worry that if we even get that first failed trial, we probably wouldn’t get the second, but instead it would be pretty easy to convince patients and doctors without much hope to try this drug just based on that subgroup analysis, and patients would be spending money on something that wouldn’t help them, and might hurt them…and because they wouldn’t be doing this in the context of a trial, we’d never get the knowledge out of this use to know it definitively doesn’t work.
Ideally the insurance companies would be the gate keepers, and insist on a certain standard of evidence before paying for a drug. But we see all the time insurance companies fail to do that now, where they even will pay very expensive prices for drugs that have been shown not to work. Maybe that would change in a world where any drug could just come to the market, but I’m having trouble seeing how. I think if anything it would be the opposite, and insurers would compete by offering access to whatever was popular, and we’d see things like acupuncture and reiki being covered.
Alan Goldhammer
Mar 2 2019 at 2:08pm
Dean and I have been in frequent correspondence over the years on his proposal to eliminate pharmaceutical patents. His proposal to let NIH fund drug development is fundamental flawed in that they do not have any incentive to stop bad projects. A pharma company puts its business on the line with every go/no go decision on developing a drug and they are quick to pull the plug if the Phase 2 data are not promising. NIH would not have this threat and furthermore, I envision significant Congressional meddling about which diseases to study and constant questioning of decision-making.
Inspection and qualification of manufacturing facilities will still be required under Dean’s scenario and those costs will have to be born by the manufacturer. If there is uncertainty about profitability, Dean’s approach may not work as designed.
David Henderson
Mar 2 2019 at 2:24pm
Good points. Thanks, Alan.
David Henderson
Mar 2 2019 at 7:14pm
Dylan above writes:
I don’t know how old you are but if you’re over 40, my guess is that you don’t worry much.
Why? Two words: off-label uses. A huge percent of drugs are now prescribed by U.S. doctors without any gatekeeper assuring efficacy. They’re prescribed for off-label uses: the drugs have passed the safety test and the efficacy test for dealing with malady A and doctors routinely prescribe them for malady B. The most famous example is Zoloft, which was approved for dealing with depression and which doctors started prescribing for pre-menstrual syndrome. That was so successful that Eli Lilly went and did efficacy tests for PMS after the fact and then got FDA approval. But there were a few years there when it was prescribed to hundreds of thousands of women for PMS without any approval for efficacy.
Alan Goldhammer
Mar 3 2019 at 8:46am
It’s important to remember that off-label use is often based on published articles in the medical literature that ‘documents’ the drug’s putative effectiveness. This is most often the case with oncology drugs that might be approved for a narrow indication(s) but use by physicians expands the utility. The big question is whether insurers will reimburse for such uses. This is the main reason why the companies will seek FDA approval for these additional indications. FDA approval is regarded as the “gold standard” and as such it is difficult for insurers to decline to reimburse as the treatment is no longer “experimental.” The follow on indications also do not require large data packages for FDA review.
Depending on the price of the drug and other treatments for the medical indication, insurers can place expensive drugs in a different co-payment tier or “force” the patient to try less expensive therapies before approving the use of the expensive drug (most often the case for new biotech drugs which can cost $2-3K/month.
Comments are closed.