A Simple Analysis of Healthcare Productivity

Last week the National Center for Health Statistics announced that U.S.  life expectancy had fallen slightly in 2008 to 77.8 years, versus 77.9 years in 2007.  Clearly this decline was related to the recession, in some way. Nevertheless, it struck me as odd that in the same year life expectancy declined,   employment in the private healthcare sector rose by 2.7%,  faster than the 10-year average growth rate.   

That observation made me think (again) about putting together a simple analysis of  healthcare productivity.  I understand quite well that this is a quixotic venture, since productivity is defined as output per worker and no one can agree on how to measure the output of the healthcare sector is.  But I’m going to take it step by step, for transparency. Everyone is welcome to lob tomatoes, as desired. 

 Let’s start from the beginning.  We don’t have a good measure of the output of the healthcare sector. However, population size is clearly related in some way to healthcare output (for a given level of ‘health’, we’d expect the output of the healthcare sector to rise with the population, holding demographic and income composition constant).

So here’s our first step. The chart below compares the 10-year employment growth in private health care services with overall population growth, and with growth of the 65-and-over population.

 

If productivity in the healthcare sector was rising, and the “health output” per American, however defined, was constant, then we would expect healthcare employment to rise slower than the  population.

But in fact, you can see that healthcare employment increased much more than the overall population from 1998 to 2008. (26% vs 10%). FYI, the same was true during the recession–from 2007 to 2010, healthcare employment rose by about 6%, while the population rose by slightly less than 3%).

More important, the increase in healthcare employment also far outstripped the increase in older Americans (a 12% gain). That means the big growth in healthcare employment cannot be due to the aging of the population.

So in fact, we’ve already learned something. The rapid increase in healthcare workers per capita is by itself a key reason for rising healthcare costs–separate from the cost of new drugs, the capital expense for new technology, and the aging of the population. 

(continued below the fold)

Pebbles in the Stream: Does the FDA Slow Medical Technology Innovation?

I’ve been using the metaphor of “throwing pebbles in a stream” to describe the effect of regulation on innovation.  No single regulation or regulatory activity is going to deter innovation by itself, just like no single pebble is going to affect a stream. But if you throw in enough small pebbles, you can dam up the stream. Similarly, add enough rules, regulations, and requirements, and suddenly innovation begins to look a lot less attractive.

A new study entitled  “FDA Impact on U.S. Medical Technology Innovation: A Survey of Over 200 Medical Technology Companies”  makes this point very well. The study, supported by the Medical Device Manufacturers Association and the National Venture Capital Association,  found lots of ‘pebbles’–inefficiencies and lags in the system of approval that added up to a big problem.  Survey respondents viewed

current U.S. regulatory processes for making products available to patients (the premarket process) as unpredictable and characterized by disruptions and delays…..[as well as ] inefficient and resource intensive

 Given who the survey respondents are, this view might be expected. But then the study did a comparison of U.S. and EU regulators, and it turned out that “it takes significantly longer to navigate U.S. regulatory processes than it does to complete European approvals for the same products.” For example,

For higher risk devices seeking premarket approvals (on the PMA pathway), responding companies indicated that it took an average of 54 months to work with the FDA from first communication to being approved to market the device. In Europe, it took an average of 11 months from first communication to approval.

85 percent of respondents considered EU authorities to be highly or mostly predictable, while only 22 percent gave the FDA the same ratings.

The longer and more unpredictable the approval process, the higher the hurdle rate for investment. No single regulatory request is unreasonable or a major obstacle, but the combination–the growing pile of  ‘pebbles in the stream’ –can be a massive deterrence to innovation.

Unfortunately, given the importance of innovation, an inefficient and slow approval process can have a negative impact on jobs and the economy.  As the study notes,

 Until recently, device innovation has largely been a U.S. phenomenon—the most important new technologies were invented here, and commercializing them in the sizable U.S. market was at the core most medtech company strategies. However, as medtech hurdles have climbed and available funding has declined, device companies are considering alternative strategies that are less U.S.-dependent…..It also suggests that the United States is at risk of losing its premier position at the center of the global medtech innovation ecosystem.

My bottom line: If Washington is genuinely serious about jobs and economic growth, it’s time to encourage innovation, not discourage it.

FYI:   Very soon I will be issuing a paper with the tentative title  “Biosciences and Long-Term Economic Recovery” through the Progressive Policy Institute, where I am a senior fellow. If you are interested in receiving an electronic copy of the paper when it comes out,  please drop me a note at mmandel@southmountaineconomics.com

Too Little Spent on the Human Genome Project?

I’ve got a lot of responses, pro and con, to my previous post on the Human Genome Project. I appreciate them all. But I just came across one (hat tip to Warren in the comments) that I find interesting. Mike “the mad biologist” writes in his post with a great title The Human Genome Project: What Happens When You Do Budget-Limited Science (or You Get What You Pay For):

Far too expensive. When budgets are limited, you’re forced to generate the data that is easier to get–and cheaper. So when Mandel describes the HGP as an economic flop so far–and he would be inclined to do so since he is interested “the innovation shortfall”–he fails to understand that we didn’t invest in the HGP adequately. Seriously, compare the $3 billion for the HGP to the billions in tax breaks companies get every year for R&D. Or inflation-adjust the Manhattan Project. Let’s not even talk about the Marine Corps’ Osprey program. By comparison, the HGP was done on the cheap.

Now there’s an interesting thought.  I’ve been going on the assumption that we were spending as fast as the science could absorb the money, but is it possible that we spent too little? The inflation-adjusted cost of the Manhattan project looks in the $22 billion range (Wikipedia number, which I rechecked).

The Debt Crisis and the Human Genome

My nomination for the most significant economic event of the past decade:  The failure of the Human Genome Project to  thus far deliver medically significant results.

Let me explain my thinking, and why there may be reason to be guardedly optimistic about the future.

Right now there are three depressing aspects to the current course of the U.S. economy.  First, the growth of healthcare spending, if it continues, will put a stranglehold on employers and taxpayers.  Second, the apparent inability of the private sector to generate well-paying jobs for college grads, if it continues, will put a squeeze on young workers.  Third,  the apparently inability of the U.S. to export enough to close a huge trade deficit,  if it continues, will leave the country exposed to a  dollar collapse and a sharp fall in living standards.

I could have arranged and described these differently, but that’s the outline of the negative picture.

The Human Genome Project had–and still has–the potential to be a powerful antidote to all three of these problems.  Let’s start with healthcare spending.  I’ll quote from today’s New York Times article (“A Decade Later, Gene Map Yields Few New Cures”):

..the primary goal of the $3 billion Human Genome Project — to ferret out the genetic roots of common diseases like cancer and Alzheimer’s and then generate treatments — remains largely elusive.

….At a [2000] news conference, Francis Collins, then the director of the genome agency at the National Institutes of Health, said that genetic diagnosis of diseases would be accomplished in 10 years and that treatments would start to roll out perhaps five years after that.

Cancer. Alzheimer’s. Diabetes. These are the expensive medical problems that eat up so much of our economy’s resources.  The possibility of a cure, say, for Alzheimer’s, could potentially turn the horrible economics of healthcare upside down. (see, for example, “Forecasting the Global Burden of Alzheimer’s Disease“).  Similarly, a cure or at least effective treatments for diabetes could sharply reduce healthcare outlays for diabetes, which are expected to triple over the next 25 years  from $113 billion to $336 billion (inflation-adjusted dollars).

What about jobs? Successful new innovations create new jobs–that’s what history tells us. If the Human Genome Project had led to a wave of new diagnostic test and treatments, the jobs would have followed.

Instead, what happened is that the pharma industry invested heavily in ‘genomics’ and got hit hard when it didn’t produce a flood of new diagnostics and treatments.  As a direct result, big pharma companies have been merging and laying off workers, not adding them. When’s the last time you heard someone talking about biology as a hot field for jobs?

This chart shows what happened over the past twenty years. In the 1990s,  job growth in pharma and biotech was able to keep up, more or less, with job growth in health services.  But over the past decade, just when you ‘d think that the mapping of the human genome would have created more jobs at pharma companies to take advantage of new discoveries, the opposite happened.  The drug pipeline dried up, and the big drug companies went into job-cutting mode.

This had an unfortunate domino effect. Cities around the U.S. had built their economic development strategies around attracting biotech jobs,  which looked like a great idea for getting ahead of a growth wave. But recent cuts have meant that the jobs gains have been relatively small.  Take St. Louis,  for example, which has been among the most successful areas in attracting  biotech research.  In 2006, for example, an article proclaimed “St. Louis And Its Companies Benefit From Biotech Push”:

St. Louis is coming of age as a biotech center…It has spent six years and added $500 million dollars in venture capital to build itself into a biotechnology hub. It has attracted new talent for companies already here, such as Monsanto Co. (MON), Pfizer Inc. (PFE) and Sigma-Aldrich Corp. (SIAL), and now is home to more than 15,000 employees at 400 more ventures, particularly in plant and life sciences.

Research, yes, for sure.   The number of jobs in the St. Louis area at “scientific research and development services” (including biotech) rose from  3500 in 2002 to 8300 in 2008.  That increase of +5K is  a fantastic performance, under the circumstances.

But research alone is not enough to make up for the loss of manufacturing jobs (down 28K over the same period).  You need production of real products, which require real production workers. Unfortunately,  employment in “pharmaceutical and medicine manufacturing”  in the St. Louis metro area appears to have peaked in 2006 (based on data through 2008). In November 2009 Pfizer announced that it was cutting 600 out of 1000 employees in its St. Louis research facility.

In fact, the “Biotech Strategy” used in St. Louis and elsewhere would have produced much bigger job gains if the research had been more successfully commercialized over the period.

Now let’s turn to trade. China, India, and the rest of the developing countries sell the U.S. an increasingly diverse array of goods and services. What does the U.S. provide in return? There’s the usual list of suspects, such as commercial aircraft (which is increasingly drawing on parts made outside of the country).  But they are not enough to avoid a huge trade deficit, even now.

The logical candidate for the next wave of U.S. exports should have been biotech products and knowledge. The U.S. is the acknowledged world leader; the research is expensive and lengthy; the production processes are complicated, delicate, require skilled technicians,  and cannot be easily offshored. And the category–treatments to deal with major medical problems–is something that everyone wants.

But what happened? Without compelling new biotech products, the big pharma companies were “me-tooed” to death. In fact, pharma trade went from roughly balanced to a big deficit.

This chart is simply astonishing.  Life sciences–the area where the U.S. is the clearly the world leader, where we have outspent everyone on research by a wide margin–has turned into a trade deficit.

Okay, it must feel like I’ve punished you all with negativism. I promised up top that I would be guardedly optimistic.  Here’s how I see it: The U.S, and more broadly the “advanced” countries, did what they were supposed to. They invested heavily in the cutting-edge new technology, biotech, which promised to make the biggest difference in the most important areas–health, food, energy.  The research has gone great, tremendous progress has been made. Commercialization thus far has sporadic–but the gap between research and commercialization is one which has been repeatedly bridged in the past. So I’d say that the odds are good that the Human Genome Project will have a significant economic impact over the next 5-10 years.

The big danger–that there are structural impediments in the U.S. innovation system which are slowing down commercialization. These include a lack of communication between academic scientists and pharma companies; excessive regulation by the FDA; a misguided patent system; and excessively short term thinking at pharma companies.  (You can add your own possibilities to the list).

I’m thinking about putting together a conference called “Fixing Pharma,” with the goal of identifying structural impediments to successful application of genomic knowledge. That’s just so important economically.  Anybody who wants to know more, drop me a note at mmandel@visibleeconomy.com

The Speed of Technological Progress

For an example of how difficult it is to predict the speed of technological progress, let’s take a look at an article in today’s NYT (my emphasis added):

…common diseases, like cancer, are thought to be caused by mutations in several genes, and finding the causes was the principal goal of the $3 billion human genome project. To that end, medical geneticists have invested heavily over the last eight years in an alluring shortcut.

But the shortcut was based on a premise that is turning out to be incorrect. Scientists thought the mutations that caused common diseases would themselves be common. So they first identified the common mutations in the human population in a $100 million project called the HapMap. Then they compared patients’ genomes with those of healthy genomes. The comparisons relied on ingenious devices called SNP chips, which scan just a tiny portion of the genome. (SNP, pronounced “snip,” stands for single nucleotide polymorphism.) These projects, called genome-wide association studies, each cost around $10 million or more

The results of this costly international exercise have been disappointing. About 2,000 sites on the human genome have been statistically linked with various diseases, but in many cases the sites are not inside working genes, suggesting there may be some conceptual flaw in the statistics. And in most diseases the culprit DNA was linked to only a small portion of all the cases of the disease. It seemed that natural selection has weeded out any disease-causing mutation before it becomes common.

So now scientists are adopting a new approach.

In the last few months, researchers have begun to conclude that a new approach is needed, one based on decoding the entire genome of patients.

The new reports, though involving only single-gene diseases, suggest that the whole-genome approach can be developed into a way of exploring the roots of the common multigene diseases.

<snip>

Dr. Reid said the HapMap and genomewide association studies were not a mistake but “the best we could do at the time.” But they have not yet revolutionized medicine, “which we are on the verge of doing,” he said.

Dr. Goldstein, of Duke University, said the whole-genome sequencing approach that was now possible should allow rapid progress. “I think we are finally headed where we have long wanted to go,” he said.

Sorry for the lengthy excerpt.  If I looked, I could find similar or stronger quotes from other scientists after the human genome was first sequenced, talking about an imminent revolution in medicine.

The question is–now that we know one approach doesn’t work,  have the odds of this new approach working? What are the odds that these scientists are right, and we are on the verge of a medical revolution?

I’m all in favor of technological revolutions. I’m just trying to apply the Black Swan perspective, which suggests that the space of possible scientific investigations is so big that eliminating one approach as failed doesn’t notably raise the possibility of success with a new approach.

I don’t mean this as a critique of science or scientific method.  Rather, I’m assessing this from the top-line economic perspective.  The knowledge that one approach has failed counts as new information. Does this new information raise or lower our assessment of future growth?