Code Red: Two Economists Examine the U.S. Healthcare System

June 16, 2014

What we do in our spare time (i.e. we have a new paper)

Filed under: Uncategorized — David Dranove and Craig Garthwaite (from Oct 11, 2013) @ 6:06 am

It seems that you can’t spend more than five minutes socializing with health economists these days without the topic of Sovaldi and the prices of new drugs coming up (we economists are a wild and crazy bunch).  As we blogged a couple of weeks ago, this new cure for hepatitis C is far more effective than existing treatments and has fewer side effects but is also quite expensive (though it should be noted that it is not actually more expensive than many of the previous treatments, it is just more expensive per day and the lower number of side effects has caused a greater number of patients to seek out treatment).  Calls are coming from all quarters to force lower prices on Sovaldi’s manufacturer Gilead.

As we discussed, constraints on the profitability of drugs once they are released, such as price controls, are concerning because they limit future innovation.  A host of previous economic studies have shown that increases in expected profits have caused greater research activity by pharmaceutical firms as measured by new products, clinical trials, or both

What has been generally absent from this existing literature is whether these increased investments result in products that are socially valuable or if they are simply rent seeking by firms.  In general, new products can increase welfare by offering more value than existing products or lowering prices and increasing the quantity sold.  However, if demand is relatively inelastic (as is the case with many pharmaceuticals) lower prices do little to change demand and just represent business stealing by firms. In this setting, new products are only welfare increasing if they are sufficiently differentiated from existing offerings.     

Critics of the pharmaceutical industry often claim that there is little true innovation and most new products are simple modifications of existing blockbusters.  If this were true, most research by pharmaceutical firms is just a very expensive form of business stealing.  We would be remiss not to note that these critics have been noticeably silent in the case of Solvadi, which represents real innovation. 

Previous studies examining the link between profits and research investments for pharmaceutical companies have not directly confronted the question of the social value of these investments. Ultimately the social value of innovation after changes in marginal profits is an open empirical question and one that we hope to help answer.  To that end, we (along with our co-author Manuel Hermosilla) just released a new NBER working paper [ungated] (yes we do something other than blogging or teaching Kellogg’s MBAs). 

In this new paper we examine the social value of changes in research investments by biotechnology firms into drugs targeting the elderly after the creation of Medicare Part D – the pharmaceutical insurance program for elderly Americans.  These biotechnology firms have traditionally been associated with innovative activity.  For example, they disproportionately produce biologic products and the firms in our sample have a history of focusing on products targeting unmet medical needs.  As a result, the products they develop are less likely to be simple modifications of existing products than those from traditional pharmaceutical firms.

Similar to the previous literature, we find a marked increase in clinical trial activity for drugs targeting conditions with a large share of patients that are on Medicare.  This appears to have been caused by the insurance expansion, i.e. prior to Part D there was no noticeable difference in the number of clinical trials based on the elderly patient share.

Given that the increase in clinical trials occurred among biotech firms, they are likely for products that represent a scientific advancement of some type. The question is the degree to which they provide an increase in treatment availability – a key indicator of their contribution to social value.  We therefore examine two indicators of this outcome:  (1) the number of existing treatments for the targeted condition and (2) regulatory indicators of the product targeting an unmet need. 

First, we categorize the conditions in our data based on the number of available pharmaceutical treatments.  We then estimate the response to Part D based on the number of treatments for the targeted conditions and find that the increase in clinical trial activity was primarily for diseases with five or more existing treatments.  Importantly, we find that there was no change in the number of clinical trials for conditions with one or fewer treatments.  While the products aimed at conditions with five or more treatments may be more effective than existing treatments, our results show that marginal changes in profits do not spur investments in products targeting conditions that currently lack treatments.

We recognize that products aimed at conditions with existing treatments may still improve welfare improving the standard of care.  So we next turn our attention to three FDA indicators of products that address an unmet need: priority review, fast track and orphan drug status.  We estimate no increase in the prevalence of products receiving these designations as a result of the creation of Part D.

To understand the mechanism driving our results, we should consider the investment decision facing biotechnology firms.  Regardless of firm size or origin, products make their way to market based on their profit potential.  Truly novel products earn large profits and are there pursued by firms in all time periods.  However, those products offering simply another treatment option in the arsenal or providing price competition for existing products might generate limited profits.  As a result, there may be a large number of products for these conditions which were not sufficiently profitable in expectation prior to an insurance expansion but become profitable after a small change in market size. If this were the case, we would expect a differential response to the creation of Part D based on the number of existing conditions.  This is what we observe.

So this brings us back to the question of price controls and more specifically the case of Sovaldi.  Our results demonstrate that marginal changes in profits do not appear to increase investments in products that serve an unmet need.  However, the question is how large of a change is required to see a change in the incentives to develop truly novel products?  Unfortunately, we (and by that we mean all economists) simply don’t know the answer to that question. What we do know, is that the easiest path is to ignore the negative dynamic incentives for innovation and impose strict price controls on new drugs offering valuable benefits, i.e. the very types of drugs we have been asking the pharmaceutical sector to develop for years.  For example, Doctors without Borders is pushing for Sovaldi to be made available for $250 in the developing world—which is their estimate of the marginal cost.  Certainly this policy would dampen incentives to develop more products that would be socially valuable.

While we must avoid the temptation of draconian price controls, our results provide more nuance to the argument against any action affecting the prices of blockbuster drugs.  Modest efforts at controlling expenditures on these new medications do not appear to dampen incentives to invest in socially valuable products.  Recently, struggling to afford the cost of this new medication, some state agencies are rationing access.  For example, the Oregon Medicaid System is limiting access to the treatment to only its sickest residents with hepatitis C.  The hope is that over time, new products from AbbVie and potentially Merck will emerge.  Importantly, these new drugs may be even better than Sovaldi – as we would expect from the incentives provided by the high prices.  At a minimum, these new products should lower prices and transfer resources from pharmaceutical firms to patients (either through lower future premiums or taxes).  Our results demonstrate that small changes such as the rationing of access based on the progression of the disease might be a reasonable way for health plans to adjust to the high cost of new treatments without foreclosing the development of new treatments. 

Blog at

%d bloggers like this: