The Cost Disease: Why Computers Get Cheaper and Health Care Doesn't Hardcover – Sep 3 2012
Customers Who Bought This Item Also Bought
No Kindle device required. Download one of the Free Kindle apps to start reading Kindle books on your smartphone, tablet, and computer.
Getting the download link through email is temporarily not available. Please check back later.
To get the free app, enter your mobile phone number.
About the Author
Most Helpful Customer Reviews on Amazon.com (beta)
Baumol projects healthcare spending to rise from 15% of US GDP in 2005 to 62% in 2105. But, not to worry because overall we will be over 8 times wealthier as our GDP per capita will rise from $41,800 to $343,000. Given that, it will be so easy to spend nearly 2/3 of every dollar on healthcare.
However, Baumol's extrapolations 100 years down the road are meaningless if not completely wrong. To understand how he derived his 2105 projections you have to read carefully note 13 on page 187.
For healthcare spending as a % of GDP, Baumol observed that they grew by 1.41% per year over the 1995-2005 decade. So, here is how he got the healthcare spending of 62% of GDP: 15%(1 + 1.41%)^100 = 60.8% (I got a different figure because of decimal figures). This same logic suggests that by 2140 or just 35 years later, healthcare will account for 100% of GDP. This does not make any sense. The 62% by 2105 does not make any more sense than the 100% by 2140. Taxes, housing, other consumptions of goods and services, business investments, Government spending can't so readily be squeezed into our remaining 38 cents on the dollar (1 - 62% allocated to healthcare).
When it comes to real GDP per capita, Baumol took the 2005 level of $41,800 and used the 2.13% average annual growth rate in this measure over the 1950 to 2001 period. His calculation: $41,800(1 + 2.13%)^100 = $343,000. Now do you believe that in 2105 we could possibly be over 8 times wealthier than we are currently? This entails that our prospective economy will be far greater than 2 times the current World economy (when you factor US population growth). If we grasp what that entails in terms of World's resources and industrial capacity, this extrapolation does not make sense. So, what went wrong? It is simple. Baumol used a 2.13% GDP per capita growth rate that captured the post WWII economic boom. And, this growth rate does not resemble at all the relevant current trend. When looking at this same measure over the 1990 - 2011 period, this growth rate plummets to 1.38%. If we focus on the 2000 - 2011 period it further drops to only 0.70%. Using this last growth rate and current GDP per capita level ($42,906 in 2011), you get a GDP per capita of $82,658 in 2105 only 24% of Baumol's level. Instead of predicting we will be over 8 times wealthier; this more realistic projection suggests we will be less than twice as wealthy as currently. And, even this level may be taken with a grain of salt. What will it truly mean in terms of resource constraint, living standard, and purchasing power parity with other countries?
Where Baumol's Cost Disease theory go astray? His basic rational is that labor productivity in the "progressive sectors" (industries with fast labor productivity such as high tech) drive overall wage earnings including the ones within the "stagnant sectors" (sectors with no labor productivity increase such as healthcare and education). This perfectly explains "why computers get cheaper and healthcare doesn't" (subtitle of the book). But, this certainly does not mean that one specific single stagnant sector (healthcare) will inevitably take over the economy. This proposition is absurd in itself given that there are so many other stagnant sectors to begin with. For healthcare to take over, it would need to squeeze out not only all the progressive sectors but also all the other stagnant sectors into this minuscule 38 cents on the dollar slice of the economy.
There is another reason that even all stagnant sectors combined will not take over the economy relative to the progressive ones. That reason is the famous Jevons Paradox. The latter states that increase in efficiency do not lead to increase in savings; they lead instead to increase consumption.
The Jevons Paradox contradicts the Cost Disease theory. Computers indeed got cheaper as the Cost Disease suggests. But, spending on computers and related hi tech appliances has gone way up (think not only of PCs but laptop, tablets, smartphones that did not exist just a few years ago) as Jevons Paradox suggests. So contrary to what Baumol thinks there is no reason for the stagnant sectors to gain share of the economy relative to the progressive sectors. In fact, the opposite is not unlikely.
The Cost Disease theory embeds many other contradictions besides the Jevons Paradox. Regarding healthcare spending, how could our wages be depressed by the rising cost of healthcare benefits since it is our own wage rate increases that supposedly set the cost increase of such healthcare prices?
Baumol's proposition that we can readily afford rapidly rising healthcare cost is laughable. Those costs are bankrupting the US fragile fiscal position as any CBO projections show. They are much reducing the competitiveness of US domestic manufacturers and have lead to off shoring manufacturing capacity. Municipalities, corporations, and household budgets have all felt the crippling impact of rising in healthcare costs.
The rising costs of healthcare and college education are indeed problems. But, it is not so much because of the Cost Disease (which primarily addresses the rising wages of employees within those industries). And, Baumol's own data proves that.
Regarding healthcare, Baumol discloses a graph (pg. 13) that shows that medical employees (doctors, nurses) salaries have grown at nearly exactly the same pace as inflation (3% to 4%). But hospital costs have grown far faster by 8% a year (graph pg. 7). Thus, the staggering rise in healthcare costs is related to factors outside the Cost Disease domain.
Turning to college education, Baumol shows that college tuition and fees have risen at 7% a year or far faster than inflation (graph on page 8). Meanwhile, wages of professors and other employees have risen a lot slower than inflation (graph pg. 13). Thus, college education costs have risen very fast for reasons outside the Cost Disease.
Given that his analytical framework is so off, Baumol's policy recommendations are of little interest. The chapters written by his coauthors in part 2 of the book regarding how to reduce the growth of healthcare costs and other stagnant services are reasonably good and interesting. But, they are lost in Baumol's book whose main thesis is wrong.
Within the stagnant sector are "high touch" service industries, including health care, education, legal services, and the arts. As Baumol and Bowen noted a half century ago, it still takes five musicians the same amount of time to play a string quintet as it did in the 18th century: technology has not changed that. (In fact, as Baumol notes in the current book, technology has led to some productivity increases even for musicians; for example, by reducing the amount of time it takes them to travel to their concert sites. But any productivity gains in the stagnant sector are very small compared to those in the progressive sector.) The cost to attend a musical concert has thus grown at a greater rate than the cost of, say computers: the former has grown faster than inflation overall, while the latter has grown more slowly. And, as with concerts, so with health care, college tuition, and so on.
This much was understood 50 years ago, and has continued to be borne out by the economic facts. What Baumol adds in his new book is a corollary to the cost disease, namely, a claim that despite the out-of-proportion growth in the their costs, services are and will remain affordable at a society-wide level. In some sense, the reason for this is clear: while services cost more in real terms, products from the progressive sector cost more, and so all that is needed is to redirect spending. In fact, because productivity increases have happened everywhere, just more slowly in some sectors than others, the society as a whole has more wealth, and thus can afford more of everything.
Here's an example of the underlying math. As you read it, remember that it's just an abstraction to show how the cost disease works and is not meant in any way to provide realistic numbers. After all, autoworkers produce many more than 5 cars a year and professors teach many more than 5 students; the price of cars must take into account the cost of material and profit for the auto company, not just the salary of the autoworkers; autoworkers and college professors don't typically compete for the same jobs; etc. None of this matters to the core argument.
So: Suppose that at a given time--let's call it Time A--an autoworker can produce 5 cars a year, with each car selling for $10,000. Then each worker can be paid $50,000. And similarly suppose that a professor can educate 5 students a year, and that tuition is $10,000. By the same logic, the professor can be paid $50,000. Now assume that at Time B, as a result of technological improvements, the autoworker can build 6 cars a year, and so can now be paid $60,000. As already noted, competitive forces tend to cause wages to rise consistently across the economy, so now the professor also receives $60,000. But technology has not made the professor more efficient: bigger classes still provide a lower-quality experience to the student. Thus, we assume that the professor can still only teach 5 students and so tuition must rise to $12,000, to compensate.
What's happened? On the one hand, it looks like tuition has gotten more expensive. But note that wages have also grown: everyone has more money to spend: $60,000 at Time B versus $50,000 at Time A. In fact, at time A, after paying for tuition, a worker (autoworker or professor) has enough money left over to buy 4 cars. At time, a worker (autoworker or professor) has enough left over the buy 4.8 automobiles. Not only do the increases in productivity mean that the output of the service section--in this example, education--remains affordable: in fact, it's even "more affordable" in the sense that there's even more money left to spend on other things.
If this seems like economic sleight of hand, it's worth doing the calculations step by step, because it's anything but! Another way to get a feel for what's going on is to consider how long the average worker has to work to purchase various products. Baumol gives lots of examples: in 1940, you had to work (on average) for 27 minutes to purchase a Big Mac; today you have to work 9 minutes. In 1910, you had to work 553 hours to buy a clothes washer, but by 1997 it was only 26 hours. A Ford Model T required 4,696 hours of labor in 1908, versus 1,365 for a Ford Taurus in 1997. And on and on. As Baumol notes repeatedly, "We can surely afford it--cars and computers, as well as health care and education. The quantity and quality of the cost-disease affected services we obtain in the future will depend on how we order our priorities. . . Society does have a choice, but if we fail to take steps to exercise this, our economy could continue to drift toward a world in which material goods are abundant, but many things we consider primary requisites for a high quality of life are scarce, particularly for the poor (pp. 54-55)."
This last point is especially important: because wealth is not distributed equally, Baumol's conclusion that we can afford it all is true society-wide, but not necessarily for every individual in the society. Hence, he argues, it is society as a whole that must commit to shifting some resources from goods produced by the progressive sector to services produced by the stagnant sector. And that, he admits, is a huge political challenge, although a critically important one.
"The Cost Disease" is an important book, definitely worth reading. If you find it interesting, you may also want to read the very lucid "Why Does College Cost So Much?," by Archibald and Feldman, for an extended discussion of how the Cost Disease has played out in the higher education realm.
Most of the arguments tend to be US-centric, and very early on in the book, the authors do acknowledge that policy may be attributable to the significant increases in US healthcare than other countries (not withstanding an attempt at re-framing the discussion to "rate of increase" - and arguing Japan has a higher rate than US - ignoring the population shifts to elderly may have contributed to that anomaly in Japan). The definition of productivity in healthcare is also a tad misleading when the assertion that - an increase in number of patients seen will inevitably lead to a decline in quality. The issue in healthcare is not how many patients are seen - but what their clinical outcomes are. That disconnect cannot be fully captured in arguments centered on productivity metrics based on number of patients seen. Plus, it mostly ignores the possibilities of technology and increasing use in analytics to help in risk stratification and personalization of treatment plans (though case studies in the second part of the book partly addresses the role of IT). Perhaps, effectiveness may be more important than productivity in healthcare...Same goes for education. A 2% increase in teacher wages (authors example) doesn't have to yield an increase in productivity - the expectation (and the right metric to measure) is did it improve student engagement and result in better test scores.
Despite that premise that leaves more-than-generous latitude for disagreement (and strong rebuttal perhaps) and examples that beg for a counter-argument, the author(s) does provide some interesting insights on cost inflation in human-centric processes primarily through their arguments centered on productivity growth and its impact on cost trends - in both "stagnant" and "progressive" sectors - and the impact on affordability. Their arguments also center on the relative inelastic nature of healthcare demand. But the general argument that our percentage of our total costs spent on healthcare (it is going to go up) should be viewed in the context of decreasing costs in other categories (manufacturing, food, etc - an assumption) is an intriguing one. In fact, the chapter that expands on this argument the associated caveats is the most thought-provoking part of the book and well-worth the investment.
The last part of the book - contributed by the co-authors - provide a good snapshot of opportunities of cost reduction. None of it would seem particularly new ideas or ones that can scale quick enough to make a meaningful change in cost inflation. With a thought provoking hypotheses, forty pages of notes and citations, and some guest authors adding to the 'story' with different perspectives, this is a very informative read.
Cost disease, consequently, is an important and timely topic. Cost disease is not widely known and a good general book about cost disease would be useful in helping the public and many policy makers understand the genesis of rapidly rising costs in important areas like health care and education. This book, unfortunately, is not that volume. Part of the book is written by Baumol, parts are by some of his collaborators, and some chapters are co-written by Baumol and a collaborator. The book has a cobbled together feel, the quality of writing is uneven, and the organization is poor. In a book about productivity, for example, the discussion of what productivity is and how to measure it is left to the middle of the book. A couple of chapters are tangentially related to the topic. Some chapters, even those written by Baumol, contain factual errors. Yale University Press produced this book cheaply; the graphs are poor, which really impairs readability at several key points. In addition to these general problems, there are at least 3 major problems with the arguments presented by Baumol and his colleagues.
Baumol has a somewhat contradictory discussion of productivity. In his chapter on this topic, Baumol discusses the different ways of assessing productivity, something relatively difficult in service-oriented industries. He comes out with an assessment of productivity which appears to dismiss assessment of productivity in terms of benefits and focuses on costs-output measures in a way that looks primarily like an arbitrary accounting device. This seems like an odd argument from an economist. If prices in a market economy don't have something to do with the value of the product, then what do prices mean? Is Baumol describing a fairly spectacular example of market failure? Baumol and one of his colleagues appear to implicitly contradict his argument about cost-based accounting of productivity in a chapter discussing business services. These would appear to be an example of a stagnant sector because of their need for expensive labor and personal services. But Baumol and his colleague argue not, because business services are actually "inputs" to economic sectors with dynamic productivity. This argument appears to accept the concept, previously dismissed by Baumol, that productivity shouldn't be measured by cost-output measures but by benefits. This argument would apply, for example, to education, as a well-trained labor force is certainly crucial to "dynamic" sectors of the economy.
A lot of this book is devoted to health care costs, particularly US health care costs. On this topic, Baumol and his colleagues are actually misleading. Baumol makes the surprising assertion that health care costs are rising at approximately the same rate in the US and other developed nations. Many analysts, however, suggest that over the last 40 years, US health care costs have risen at a faster rate than those of other industrialized nations. This can be seen in Figure 1.5 and is supported by a simple statistical analysis. If the rate of increase in US health care costs is an outlier, and cost disease is a universal phenomenon (which I think is correct), the large discrepancy between rising US health care costs and those of other nations must be due to other phenomena. This fact has 2 important consequences for Baumol's argument about health care costs. The first is positive. To the extent that rising US costs are not due to cost disease, we should have good opportunities for cost reduction. In principle, cost containment/reductions should be obtainable without great penalties, as other developed nations provide universal care with aggregate better results. The negative implication is that Robinson's corollary argument will not apply to a large fraction of rising US costs and that rising health care costs are a large economic threat.
Finally, Baumol appears to miss an important connection between "stagnant" and productive sectors of the economy. Baumol's argument stresses that cost disease is a product of rising productivity in dynamic sectors. Baumol indicates that productivity gains derive from 2 sources; large oligopolistic firms with big R&D commitments (think IBM or Intel) and individual entrepeneurs. This is an imcomplete picture as it omits 2 key sources of productivity. One is the crucial state-sponsored R&D efforts which support not just basic research but also a good deal of technology development. The second is the education sector. Large oligopolistic corporations do not, by and large, train their scientists and engineers (though IBM and a few others have research centers with post-doctoral opportunities) and individual entrepeneurs do not just spring from the ground. The dynamic sectors of the economy depend on the "stagnant" sectors for essential intellectual and human capital. Its not just that we can afford the rising costs of the "stagnant" centers, we have to pay these costs if we want the type of productivity growth that Baumol sees as crucial.
For years, I have been advising school boards and administrations about Baumol's "Cost Disease". Those labor-intensive institutions, with few means of productivity improvement, will have costs that inevitably outstrip growth in the Consumer Price Index (CPI). This happens because they have to offer compensation that matches growth in the more productive sectors (such as manufacturing). Otherwise, we will lose workers in vital low-productivity growth sectors such as education, health care, the criminal justice system at all stages of the process, and more. This is Baumol's thesis, and he should be the best to expound it. However, he fails to pursue its consequences in a manner that survives careful examination, with the character of the projections he makes. This is sad, because those arguments could have been made effectively.
Consider that many school districts are restrained by "tax caps", where revenue growth without a referendum is limited to CPI growth. They have an inevitable collision with that. In manufacturing, however, workers can be given compensation increases above the CPI increase, because of the productivity from capital investment, technology, training, better methods, and more. This Baumol describes as the difference between the "stagnant" and the "progressive" sectors, with a "hybrid" sector in between. Some of the "stagnant", personal service industries die off (doctors do not make house calls anymore, there are no milkmen, and there are butlers and cooks only on Downton Abbey). Others will never vanish, because they are necessary to society: Health Care, Education, Criminal Justice, and more. Even barbers and hairdressers must survive.
So I needed a book I could give to schools, and one from the original source would be excellent. This book fails miserably. It is not an academic book by any means (nor should it be, but at times it tries, schizophrenically). It attempts to be a popularization of economic thought, and guru on policy advice, but it does not succeed. Instead, it is a mishmash of different authors and of styles (even within Baumol's own chapters, where he ranges from the academic, to attempts at a breezy style). He fails in the key Chapter 4, "Yes, We Can Afford It". I will have more on that later.
As another reviewer remarked, some of the graphs are poorly rendered, making them almost useless. The XY Charts use data markers that are so small that they are indistinguishable among the multiple data sets. That could have easily been fixed with distinctive line styles (still not requiring color printing), but then the book is poorly edited overall. The most significant editing problem is for coherent content on the modeling in Chapter 4, which would get an MBA student flunked. Even Baumol, in his extremely lengthy Endnote 13 to that chapter, knows that something is wrong. His own colleagues must have warned him. One has to go to the endnotes to ferret out the details, on this pivotal chapter.
He assumes that health care costs, as a percent of GDP, will continue to grow at a compounding rate of 1.41% per year, with no leveling off, ever. Reviewer Gaetan Lion points out that such an assumption will eventually exceed 100% of the economy. Baumol seems to recognize that in Endnote 14, and then blithely ignores it. In fact, in his lengthy Endnote 13, he says he is not doing a "forecast", but a "projection". That seems to be a distinction without a difference, unless "projection" now means "meaningless extrapolation for 100 years, without any underpinnings and logic, leading to an impossibility." Yet he refers to his own use of logic in the endnote. Redo your homework, sir.
Health care in his projection ends up constituting 62% of the economy in one hundred years. As other reviewers have pointed out, this crowds out even the other "stagnant" industries, such as education and criminal justice. We're still going to need barbers and hairdressers also. It has reduced the "progressive" sector to only 38%. Yet this is what is supposed to be paying for everything, so his 2.13% real productivity growth rate across the economy is unsustainable, and its contribution is being steadily eroded. No one seems to have pointed that out -- he has some means to pay for things, but not the amount he assumes, and it would be shrinking every year. This is what results when someone does a back-of-the-envelope calculation, without actually trying to flesh it out in a moderately detailed spreadsheet, where the absurdities and inconsistencies would become apparent.
With regard to some of these considerations, he remarks that a "nonlinear" model would probably not materially change the "linear" model he uses. All I can figure out is that he is rejecting a model that would have certain things grow to some asymptotic limits, such as an "S Curve", or "Logistic Curve". Yes--that is exactly what would make his projections plausible, and perhaps help him make his point - that we will be able to afford all this in some way. I agree with him that no one can fine-tune such predictions. So run half a dozen different plausible scenarios, which actually might make some sense, demonstrate the range of possibilities, and show how things might still work under a variety of circumstances.
As a final critique, I will note that some things noted as "Cost Disease" are actually "Mission Creeps/Mission Enhancements". On page 21, he refers to negative productivity in education (not just stagnation), referring to changing student/teacher ratios in recent years. Much of that was due to Federal and State mandates re Special Ed, and increasing classification of children as Special Ed, where class sizes are often only half as large. Likewise, other reviewers have pointed out that compensation of doctors, nurses, and other medical professionals does not explain much of health care expense growth. Stats from the Dept of Education and Dept of Labor show that Baumol's Cost Disease per se only accounted for less than a third of the slippage above the CPI over a twelve-year period in Cost per Student. The rest was Mission Creep (think Special Ed, Technology, Security...), or possibly waste. This book attempts to address a significant set of issues, but does so superficially.
This book requires such a major rework that I think the author will not bother. Yale University Press did a poor vetting of its content, perhaps based on the distinguished past of its chief author. It cannot be used for giving to policy makers, since it can be easily attacked for its absurdities. That need not have been the case, since a proper exposition could have solidly made many of his points, and those are important ones.
Look for similar items by category
- Books > Business & Investing > Economics > Economic Conditions
- Books > Business & Investing > Economics > Inflation
- Books > Business & Investing > Economics > Theory
- Books > Business & Investing > Industries & Professions > Service
- Books > Business & Investing > International > Economic Conditions
- Books > Education & Reference > Higher & Continuing Education
- Books > Medical Books > Administration & Medicine Economics > Health Care Delivery
- Books > Professional & Technical > Accounting & Finance > Economics > Economic Conditions
- Books > Professional & Technical > Accounting & Finance > Economics > Theory
- Books > Professional & Technical > Accounting & Finance > Finance > Inflation
- Books > Science & Math > Medicine > Administration & Policy > Health Care Delivery