This text was transcribed and edited from an informal oral presentation.

It's always a pleasure to be at the Washington Legal Foundation. It's one of my favorite organizations. I can't think of another organization that manages to achieve so much given its modest size. And I say that partly because of how closely I follow the Food and Drug Administration and partly because I can assure you that the Food and Drug Administration follows the Washington Legal Foundation very closely.

So, here are a few words about what's going in the current wave of alcoholic beverage litigation, which is part of a larger trend toward regulation by litigation.

These alcohol cases focus on advertising and promotion. I don't think that's a coincidence. The plaintiff attorneys want jury trials, so it's only natural to focus on advertising and promotion because that's what juries are familiar with. Surveys generally show that most people think advertising has a powerful effect on other people. People who serve on juries seem quite prepared to be persuaded that advertising has a big effect on their fellow citizens’ behavior--especially the behavior of young people--even though they don't think advertising has much to do with their own behavior.

There is quite a bit of research on the effects of alcohol advertising. This research began, roughly speaking, in the mid-1970s. Economists have studied the effects of alcohol advertising in the United States, the United Kingdom and various European countries. There are now dozens of such studies. They have employed a large variety of econometric models and diverse data sets. In fact, econometricians sometimes use alcohol advertising data to experiment with the latest twists in econometric methods.

Invariably, empirical research finds no effect of advertising on the total amount of alcohol consumption. Yet this is precisely where the effects of advertising should be the easiest to detect. When advertising goes up by 10 percent, 50 percent, or 200 percent, as sometimes happens, you ought to be able to find at least a measurable blip in total consumption--assuming, of course, that advertising really does affect consumption. But those blips cannot be found.

This is true even in places like France and the Netherlands, which I've studied myself and which have provided remarkable natural experiments in the dynamics of advertising. In the Netherlands, for example, alcoholic beverage advertising doubled between 1990 and 1994, with no perceptible impact on consumption. In France, advertising tripled between 1977 and 1989, and then was almost completely banned by the Loi Evin in 1991. Alcohol consumption in France continued to decline at about the same annual rate throughout this entire period.

Why doesn’t advertising have the effects that so many people seem to expect it to have? I think the intuitive answer is fairly straightforward. When it comes to something as basic as the decision of whether to drink--or whether to smoke, for that matter--advertising is a very weak force. There's almost no evidence that advertising has an impact on decisions to initiate or continue those kinds of behavior. Those decisions are deeply embedded in such matters as family environment. That's one reason for the widespread failure of social marketing campaigns in which governments use advertising to persuade people not to engage in certain behaviors such as underage drinking, smoking, illicit drug use, or unprotected sex. That kind of advertising has proved extremely difficult to pull off successfully. Using the mass media to affect those kinds of decisions is a very difficult task.

What advertising usually does achieve is simply to drive market shares.  Economically speaking, there are perfectly rational explanations for this.  Individual sellers have a strong incentive to engage in advertising.  Even if it's true, as it is for alcoholic beverages, that advertising has no impact on the total market size, it does have an impact on brand shares.

Economists have been fiddling around with these ideas for a number of years. Somewhere along the line, someone formulated what is called the “fallacy of composition.” The fallacy of composition is the assumption that if a certain economic activity, such as advertising, benefits each individual participant in the market, such as manufacturers of individual brands, then all the advertising put together must benefit the entire market. This reason does not hold up in general, nor does it hold up for advertising in particular. In other words, each brand may do better when it advertises than when it doesn’t advertise, but total advertising may have no effect on total sales in the market. The brand-level effects may offset each other. This is in fact what we find in alcoholic beverage markets.

For a good example of the fallacy of composition, look at political advertising. There is no doubt that advertising has a profound effect on individual politicians’ share of the vote. But there's no evidence that political advertising increases total voter turnout.

Now let’s get back to alcohol litigation. Speaking as someone who's not a lawyer and who has not been involved in this litigation in any way, my prediction is that if this new wave of litigation is an attempt to regulate through litigation, it will fail. Or to be more precise, it will fail unless one of two things happens. One is that either state governments or the federal government become plaintiffs. Government plaintiffs make a big difference, especially when a state government sues in state courts.

The other thing that could permit this litigation to succeed would be if the courts were to permit the aggregation of plaintiffs into classes that are so large that a single jury in a single case could threaten a defendant firm with bankruptcy. That can happen when a damages award is so large that the bond required to mount an appeal exceeds the defendant’s net worth--as happened in tobacco litigation, which I will discuss shortly.

If one of those two things happen, especially the second one, then I can imagine that this litigation might succeed at least in the sense of transferring a great deal of money from defendants to state governments and other plaintiffs.

This brings me to tobacco litigation, which is obviously the closest parallel to the kind of litigation we are talking about today. It is a fact that the first wave of tobacco litigation focused on advertising for the reasons I've mentioned--mainly, that plaintiff lawyers expect juries to think that advertising may cause drinking.

The role of advertising in tobacco litigation is ironic because, parallel to what we know about the alcoholic beverage market, the economic evidence shows that advertising plays no role in how many people smoke or how much they smoke. Even in the notorious example of Joe Camel, a close look at that market, which I conducted a number of years ago, makes it obvious that Joe Camel had no discernable effect on smoking behavior beyond some modest changes in brand shares.

What we see in the case of tobacco advertising is the same thing we see with alcohol advertising.  Incentives to advertise individual brands are very strong. We observe effects from advertising at the brand level, but no effects on the overall volume of smoking. And, as far as anyone has been able to document, we find no effects from advertising on youth smoking.

Nonetheless, tobacco litigation eventually transferred immense amounts of money, and it did so only after litigation was initiated by state governments in state courts. The state litigation finally put juries in the position where they had the power to bankrupt the defendants through a single adverse decision. That was the essence of the circumstances that generated the mass settlement agreement, which before it expires will transfer several hundred billions of dollars, mainly to state governments. 

All this happened despite the fact that the premises of the tobacco litigation were highly questionable. Essentially, the tobacco litigation succeeded despite being based mainly on untenable arguments. One was that advertising causes smoking. Another was that smoking increased health care costs for state governments and private insurers; there's no evidence that that actually happened. Another allegation, which I personally think is absurd, is that the state governments and (according to the current lawsuit brought by the U.S. Department of Justice) the federal government were actually deceived into thinking that smoking really wasn't very harmful. In fact, it's perfectly obvious from the historical record, as well as from documents generated by litigation, that on the whole what governments knew about smoking and health was pretty much what the tobacco companies knew. The tobacco manufacturers were getting their information about the health effects of smoking mainly from the same sources as the public health community. The industry was engaged mainly in monitoring and replicating the publicly available medical literature. Nonetheless, the tobacco litigation has had large effects. We should pay close attention to those effects because tobacco litigation is a model for regulation by litigation generally.

This tobacco litigation directly generated large price increases on the order of 30 to 50 percent. The increased revenues were mainly handed over to state governments. The litigation and its tax-like effects also provided some of the intellectual and political foundations for even larger legislated increases in state taxes. The litigation has also greatly reduced the volume and variety of tobacco advertising. In fact, I for one hardly ever see any tobacco ads these days.

I think the tobacco litigation has also enhanced the political power of advocates seeking increasingly pervasive controls over smoking. Smoking today is a far more difficult activity to engage in than it used to be. It's harder to buy cigarettes and it’s harder to find places to smoke them. Finally, the litigation has also generated a great deal of antismoking activity, especially in the mass media, but also in the schools and elsewhere.

Nonetheless, when you look back over the last 15 years or so at what really counts--smoking behavior itself--it's amazing how little has changed. Roughly 20 to 25 percent of the adult population smokes. That's pretty much what it was in the early 1990s. This is consistent with long-run trends in developed countries around the world. In the U.S. and Western Europe, and recently in Japan, smoking prevalence declined steadily and substantially after knowledge of smoking and its harm became better established and widely disseminated. Smoking declined most rapidly in the U.S. even though the U.S. was long the home of the most prominent and creative cigarette advertising. But when the prevalence of smoking gets down to about 20 to 25 percent of adults, it usually gets stuck. That's pretty much what's happened in the U.S. as well as elsewhere. Massive litigation has had very little impact on overall smoking rates.

Youth smoking has also changed remarkably little, although here the data are softer and less stable. In fact, youth smoking actually increased during the early 1990s. In the past year or two, the antismoking community has celebrated the fact that smoking by high school students is finally back down to where it was before all this litigation began.

The bottom line is that if you see all this litigation as a tool for getting good regulations, it's hard to think of any   goal that could not have been achieved at least as well through legislation or administrative regulatory proceedings.

On the whole, the price increases and advertising restrictions obtained through litigation are ones we could have gotten through legislation. Of course, firms under attack in the courtroom may have sometimes agreed to rules that could not have withstood First Amendment scrutiny if they had been legislated, but that is another reason to prefer legislation and regulation to litigation. What counts most, however, is smoking, and that is where litigation has been a clumsy and ineffective tool.

As far as controls over smoking are concerned, most of those have not come through litigation. On the other hand, litigation has funded a variety of antismoking activities, but again, their effects seem to be very limited, and the same measures could have been created through legislation. That is clear from the breadth and variety of controls enacted in the past two or three years by city and state governments.

I think the only genuine, unique effect of the kinds of litigation we are discussing today has been the transfer of large sums of monies, partly to law firms in unprecedented amounts, but mainly to state governments. Those governments have used these revenues mainly for purposes having nothing to do with the litigation that provided them.

Well, those are one person's views on how this litigation has played out in the past and why an attempt to regulate the alcoholic beverage market through litigation is probably doomed to failure.

John E. Calfee is a resident scholar at AEI.