Anyone who defends the pharmaceutical industry has to be ready to hear, over and over and over, about how much it spends on sales and marketing versus R&D. This is thought to be a telling point about where the priorities really are. I've addressed this one several times, and my best response is to point out that sales and marketing are actually supposed to bring in more money than you spend on them, and do so more reliably than R&D in the short term.
There's now a very useful paper in Nature Reviews Drug Discovery looking at just this issue. The authors (from three universities in the US and Israel) are looking into the general question of which is the better use of money: put it into R&D for the long term, or promote existing products for the short term? I should make clear at the outset that those two options do line up in that way. R&D expenditures take years to pay off, if ever, given the amount of time that drug development takes. And marketing of a current product had better start paying off in a shorter time frame, because every patented drug is a wasting asset, constantly being eaten into by competition and by its time to patent expiration.
So which makes more financial sense? The authors' numbers are from the Wharton databases on publicly traded drug companies, looking at those with more than $50 million in sales. Using the company stock prices as a measure of value (J. Finance LVI(6), 2431–2456 (2001), I'm giving you references here), they found, in general, that R&D investments have a net positive effect, while increased promotion has a negative effect. (See also Rev. Account Stud. 7, 355–382 (2002), another journal I don't reference much). Both effects are larger for smaller companies, as you might expect, but they held up across the industry. The effect also holds up if you factor out the compensation packages of the top five executives of each company (which is a nice control to run, I have to say). And yes, since you ask, there is a negative effect on stock price that correlates to higher executive compensation, and I'm willing to bet that this effect holds for more than just the drug industry.
Since we're talking about stock prices, which are generally forward-looking, the way to interpret these results is probably that investors expect R&D expenditures to pay off in the long term, but actually expect sales and marketing expenditures to reduce long-term value. If that's so, then why spend money on marketing? The reason the authors propose is just what I'd been talking about: short-term reliability. Drug discovery and development is inherently risky, and promotion of existing products is (at least comparatively) more of a sure thing. Companies engage in a mix of the two to try to even the cash flow out. (And as the authors note, if executive compensation is tied more to short-term performance, then there's an incentive to go with the short-term gains).
In general, though, you'd figure that companies should invest more in R&D. And here's the real kicker: that's exactly what's been happening. As this graph from the paper shows, over the last thirty years expenditures in the Sales, General, and Administrative area have risen only slightly as a percent of sales. The Cost of Goods Sold category (materials, physical plant, manufacturing facilities, etc.) has gone proportionally down, with an interesting excursion in the mid-1990s. (Note also that this used to be the leading category). And R&D expenditures (again, as a percent of sales) rose in the 1980s, were flat in the 1990s, and have risen since then. Overall, since 1975, the proportion of money spent on R&D has more than tripled, from 5% to 17%.
This, I hardly need point out, does not fit the narrative of some of the e-mails and comments I get. Some perceptions of the drug industry have us, Back In the Old Days, as spending our money on R&D, only to slimily slide into becoming pure marketing businesses as time has passed, with our recent years being especially disgusting and rapacious. According to these figures, this is at the very least not accurate, and comes close to being the opposite of the truth. Comments are welcome - most welcome, indeed.