Direct to Consumer Advertising revisited

I'd quickly reviewed the literature on direct-to-consumer pharmaceutical advertising after a rather ill-considered call for a ban on it in New Zealand. I concluded that the best case you could make against DTC advertising would have a harder time applying to New Zealand given Pharmac, and that even that best case was empirically questionable given the reasonable existing evidence of consumer benefits.

In short, there's a pretty big literature on the actual effects of direct-to-consumer advertising of drugs. It's far from obvious that banning it is a good idea, and, in particular, that it would be a good idea here given the structure of our pharmaceutical market.

Tyler Cowen today points to a new paper by Dhaval Dave reviewing the literature on DTC advertising. Here's the abstract, via Marginal Revolution:

This review discusses the role of consumer-directed and physician-directed promotion in the pharmaceutical market, based on the classic conceptual framework of whether such promotion is “persuasive” and/or “informative”. Implications for public health and welfare partly depend on whether, and to what extent, advertising: 1) raises “selective” or brand-specific demand versus “primary” or industry-wide demand; 2) impacts drug costs; and 3) impacts competition. Empirical evidence from the literature bearing on these effects is surveyed. These studies show that pharmaceutical promotion has both informative and persuasive elements. Consumer advertising is more effective at enlarging the market, educating consumers, inducing physician contact, expanding drug treatment, and promoting adherence among existing users. Physician advertising is primarily persuasive in nature, effectively increasing selective brand demand. There is no strong evidence the drug promotion deters entry, and there is some suggestive evidence that it may even be mildly pro-competitive. There is also no strong evidence that either consumer- or provider-directed promotion substantially raises retail-level prices. While all of these effects point to welfare improvements as a result of pharmaceutical promotion, there is also evidence that consumer ads may induce overuse and overtreatment in certain cases. Market expansion, overtreatment and shifting brands for non-therapeutic reasons further raise the concern of a sub-optimal patient-drug match at least for some marginal patients. A comprehensive evaluation of the welfare effects of pharmaceutical promotion requires a balanced assessment of these benefits and costs.

The bar for banning things should be high.

DTC prescription drug advertising

John Pickering at SciBlogs' Kidney Punch recommends banning direct-to-consumer drug advertising in New Zealand. He writes:

New Zealand is one of the few countries in the world where pharmaceutical companies are allowed to advertise to consumers directly.  I believe this is detrimental to our health and I call on the government to ban the practice immediately.

I expect that the case against DTC advertising is weakest in New Zealand, precisely because of Pharmac. But let's walk through it.

Let's begin with the best case against DTC advertising. Brekke and Kuhn, 2006, argue in the Journal of Health Economics that DTC encourages manufacturers to provide greater detailing (lobbying physicians about their drug's particularistic benefits) and so better to segment the market; this allows manufacturers to increase the price of their drug with less harm to their drug's relative demand. Where consumers are not price sensitive, this can lower consumer welfare - if somebody else is paying for the drug, and the one on the ad seems shiny, why worry about the price?

Now let's think about the New Zealand pharmaceutical market. Pharmac is the government's drug-buying agency. They develop a schedule of drugs worthy of funding based on clinical effectiveness and cost; people can buy non-funded drugs, but it comes out of their own pocket unless they have private health insurance that covers non-funded medications. The cost jump in moving from scheduled to non-scheduled pharmaceuticals then is substantial.

And, physicians seem reluctant even to mention treatments that could be off-schedule. Based here on personal experience, we almost had to pry out of our physician a list of non-funded vaccinations that she thought worthwhile but that weren't on the funded schedule. Had we seen ads for a chicken pox vaccine, we would have been made (expectationally) better off; we didn't know one existed until we heard about it from friends in the States.

Consumers here will not be price sensitive among scheduled drugs, but they should be when going off-schedule. And all the drugs on the schedule have already been vetted for cost-effectiveness. Maybe there could be some second-order distortions where DTC could lead a consumer to ask for a scheduled drug for an off-label use for which it's less effective than some other scheduled drug. If we expect doctors here would fail to provide alternative advice, either because the kinds of patients who do all their research online before coming to the visit are a hassle to deal with, or because they've been lobbied by the manufacturers of DTC-advertised products already, then this could yield some losses. And maybe this could hold for some on-schedule drugs still here under patent, but losses will remain lower than in regimes without the kind of price controls that New Zealand has.

So, if there is a case for banning DTC advertising, it would seem weaker here than elsewhere.

It's also not completely clear that DTC advertising is all that bad. Calfee, Winston and Stempski (2002) find that DTC advertising of statins did not result in any increase in statin prescriptions but seemed instead to have increased the proportion of those on statins whose treatment was successful; they interpret this as that patients on statins were more likely to comply with doctor's orders on their statin prescriptions due to all the happy statin people on the ads. Further, de Frutos, Ornaghi and Siotis (2013) find that better drugs are advertised more heavily; likelihood of flipping to lower-quality drugs then is lower. Jayawardhana (2012) finds that DTC advertising, at least in cholesterol drugs, increases consumer welfare through its informative role: patients who hadn't realised that they ought to see a doctor about a condition turn up at the doctor's office.

In short, there's a pretty big literature on the actual effects of direct-to-consumer advertising of drugs. It's far from obvious that banning it is a good idea, and, in particular, that it would be a good idea here given the structure of our pharmaceutical market. A post on SciBlogs calling for such a ban could profit from some interaction with the literature; one online ad for a drug that could be less effective than some alternatives hardly seems sufficient basis for calls for such a broad policy change.

Advertising, brands, and prices

Advertising both persuades and informs. Informative advertising makes markets more competitive, reducing prices faced by consumers; persuasive advertising that reinforces brand loyalty segments markets and increases prices paid by consumers.

Paul Walker points to a new study looking at which effect dominates. From the VoxEU write-up:

I first show that the taxation of advertising is indeed a powerful instrument to restrict advertising expenditures of firms. I also show that advertising increased consumer prices in some industries such as alcohol, tobacco and transportation, in which the persuasive effect dominates. But it also decreased consumer prices in other industries such as food. I use data from existing marketing studies which make it possible to relate different responses of market prices to characteristics of advertisements in industries. I can indeed show that those industries which exhibit the informative price include more information in their advertisements, consistent with the interpretation of informational and persuasive forces of advertising.

I expect what's going on with persuasive advertising is mostly brand-loyalty reinforcement. The test for that would be whether the gap in price between branded and unbranded soft-drinks decreased with higher advertising taxes; that test isn't reported in the paper, and I'm not sure if it's feasible with their data. But it's at least a falsifiable hypothesis.

Most tobacco advertising is illegal; New Zealand is debating mandating plain tobacco packaging.* The tobacco industry's been pretty angry about plain packaging. If brand labelling mostly works to reduce competition across brands and to help maintain customer loyalty, we'd expect that plain packaging mandates will lead to a shift towards discount brands and lower prices absent further excise hikes. And that's also what Clarke and Prentice expect.** You can simultaneously have a policy anger the tobacco industry while increasing smoking if it pushes current smokers to discount brands, reducing the average price they pay and consequently increasing consumption while decreasing industry profits. The policy that's the enemy of your enemy isn't necessarily your friend.

The paper's overall result - that Klein-esque bans on advertising would really hurt consumers - likely won't change the no-logo crowd's views.

*Or, rather, the anti-tobacco groups were shouting a lot about the need for it, perhaps encouraged by MoH. Sometimes-lobbyist Carrick Graham points out the NGOs went quiet in preparation for the TPP talks' coming to town, implying that they're mostly client groups for the Ministry of Health and that MoH was leaned on a bit by Trade. It's an interesting hypothesis.

** And see discussion here.