The Agenda

Credentialism Creep as a Source of Stagnation

I see that Sarah Kliff has beaten me to the punch of writing about a new (quite short) article by Robert Kocher and Nikhil Sahni on labor costs in the health sector. I’ll focus on a few different aspects of the article:


[H]ealth care labor is becoming more expensive more quickly than other types of labor. Even through the recession, when wages fell in other sectors, health care wages grew at a compounded annual rate of 3.4% from 2005 to 2010.

Complicating this picture is the expansion of health insurance coverage to 34 million additional people over the next 10 years under the Affordable Care Act (ACA).1 This increase in the population of insured Americans will expand demand and the need for labor — potentially to the point where labor becomes scarce and therefore even more expensive. If we add these new beneficiaries to the health system and expand the workforce proportionally while retaining today’s labor structure, total health care costs will increase by $112 billion, or 13%. Therefore, to be successful, any effort to slow the rate of growth of health care spending will require a change to the labor structure.

Recently, we discussed what a change in labor structure might entail, e.g., a shift to more use of PAs working in collaboration with hospitalists and away from heavy reliance on medical residents. I should note that Razib Khan has a less sanguine view of our trajectory in this regard:


I applaud the wider distribution of medical services outside of the licensing monopoly of M.D.s. As an empirical matter I think there was a practical reason for the professionalization of medicine in the 20th century and the emergence of degree holding as necessary. To be frank about it for most of human history doctors were frauds or butchers. Modern medicine in the 20th century was a major revolution in that sense (though doctors are only part of it, the rise of an effective pharmaceutical industry is probably just as important if not more so). But the arrow of history does not always move in one direction, and we live in an “information age.” Doctors are human, and therefore fallible. They need the aid of both their patients and various other medical professionals to optimize health outcomes. The paternalistic model is just not viable in the long run, especially as the median educational qualifications of their patients keeps rising.

But notice that in this case we’re seeing greater and greater credentialism in fields which were traditionally perceived to be auxiliary to medical doctors. This is not a good sign. Instead of challenging the unquestioned prominence of medical doctors in domains where nurses are sufficient and more cost effective, the nursing profession is “fighting fire with fire.” This is not going to end well. Having to pile on education removes productive years in the work force. This is justifiable when education results in gains in productivity, but just as in education, I suspect that all the extra years for physical therapists and nurses is not doing anything but signalling, and further tightening up labor supply as the number of patients keeps on increasing because of the aging of the population. [Emphasis added]

Kocher and Sahni touch on this issue:

If the health care sector is to achieve even the average improvement in labor productivity seen in the overall U.S. economy, we will need to redesign the care delivery model much more fundamentally to use a different quantity and mix of workers engaging in a much higher value set of activities. (Although some activities, such as feeding patients and tending to their hygiene, may be impossible to accelerate, productivity is improved when these activities are performed by lower-cost but capable labor. Approaches that encourage delegation of tasks from physicians and nurses to other workers — for instance, transferring postsurgical care from surgeons to physician assistants — provide opportunities for additional savings and increased productivity.) This solution implies eliminating myriad time-wasting, low-value activities; increasing our use of technology, data, evidence, and teams; increasing standardization to avoid rework; and relying on evidence-based personalized care to avert complications. [Emphasis added]

In light of an emerging meme (“anyone who thinks that regulation impacts employment levels is a right-wing lunatic”), I was pleased to the following:

A large obstacle to such a wholesale redesign is the complexity of the federal and state reimbursement rules and requirements for scope of practice, licensure, and staffing ratios. One example of the current inflexibility is the requirement that all imaging centers have a physician on hand at all times if intravenous contrast may be administered, owing to the 0.1% probability that a patient will have a severe, life-threatening allergic reaction.3,4 Surely, other health care professionals could be trained to respond effectively to such an allergic reaction, which would liberate these physicians to fill higher-productivity roles. In addition, though providers are integrating new technology into their systems, they have no incentive in fee-for-service reimbursement for improving productivity by converting inpatient encounters to virtual visits, incorporating remote monitoring, or managing treatments with lower-cost care coordinators. 

The authors have done all of us a service for bringing attention to a critically important issue. And they also offer a sophisticated discussion of economy-wide productivity laggards:

Over the past 20 years, our real gross domestic product (GDP) grew 2.5% annually, with total employment contributing 0.7% and labor productivity the remaining 1.8%. In comparison, the real “value added” of the health care sector, measured as the contribution of the industry’s labor and capital to its gross output and to overall gross GDP, grew at 2.3%, with total health care employment contributing 2.9% while labor productivity actually decreased by 0.6% annually. …

Aside from health care, only education and defense showed no aggregate gains in productivity.

To be sure, we don’t have very good metrics for understanding output in education and defense and other sectors dominated by the public sector, as government spending is evaluated in terms of inputs. My suspicion is that a rigorous market-based assessment of the value of public services would actually lead us to an even grimmer conclusion regarding how well resources in that space are deployed, certainly in contrast to the public sector in Singapore and northern European states like Denmark, Norway, and Finland.


The Latest