Wine Economics: Quantitative Studies and Empirical Applications
The first thing to note about this title is that this is a serious book, and, unlike many other books about wine, its primary objective is not to entertain but to inform and instruct. As a consequence, the average wine connoisseur will find it hard going, except perhaps for the introductory and concluding sections in each chapter. It is written by and for those with more than a passing acquaintance with reasonably ad- vanced topics in econometrics. The authors hail from three continents, and the editors are to be congratulated on putting together a volume of essays by highly dis- persed authors and managing the task of keeping the style and the general method- ology reasonably uniform.
My second general comment is this: the volume consists of 16 chapters, each on a different topic, and it is not possible to do justice to each and every chapter in a brief review. I want to stress that the chapters that are not discussed in detail in what follows are no less valuable or excellent than the ones that I do discuss; my selections necessarily reflect my own interests, and no implicit criticism is to be read into these selections. Overall, I can say that all the chapters are very solid and reflect the use of appropriate methodology, and those that are empirical in orientation rely on prodi- gious amounts of data.
The volume is divided into five parts, titled Alcohol Consumption and Welfare, Consumer Behavior and Prices, Wine Ranking and Financial Issues, Intermediary Markets and Strategic Decisions, and, finally, New Topics. Just to be perverse, let me start at the end. The brief final part of this volume deals with two very timely topics: wine tourism and the high and rising alcohol content of wine. To analyze the former, Françoise Bensa and Marie-Claude Pichery collect a large sample of win- eries in Burgundy and Alsace (376 in all) and analyze the impact of wine tourism ac- tivities on variables such as on-site sales. They find that impact is mostly in the long run. Much depends on the maintenance of more or less regular contacts between the growers and the tourists, and it is noteworthy that 86% of growers recontact their visitors in one way or another. Concerning the alcohol content of wine, Alston et al. find that “even a substantial rise in average temperature … would have had only modest effects on the sugar content of wine grapes” (p. 354). In any event, alcohol content has been rising, a fact that is bemoaned by many wine connoisseurs.
Chapter 1 examines the relationship between alcohol consumption in its three principal forms (beer, wine, spirits) and self-reported measures of life satisfaction. On the whole, the relationship is negative and, in some cases, even significantly so. Chapter 2, dealing with externalities and optimal taxes, considers three types of drinkers: moderate drinkers, informed abusers, and uninformed abusers. The con- sumption of alcohol entails enhanced health and accident risks, and some of the costs of those are privately borne, while others are externalities. The total social cost of alcohol ranges between 2.1% and 2.5%, of which perhaps 0.7% is in the form of externalities. Fairly standard formulas involving the relevant elasticities yield optimal tax calculations, with the result that “the current beer tax rate is broadly consistent with the base-case welfare maximizing tax rate, the current wine tax is about half the base-case welfare maximizing tax rate and the current spirits tax is about twice the base-case welfare maximizing tax rate” (p. 44). Chapter 3 examines the effect of the Evin Law in Europe, which banned the use of certain types of advertising of alcoholic beverages. A Rotterdam model is estimat- ed, and it is found that the law significantly affects demand. I do not think that this is surprising, but it is comforting to have it confirmed.
Chapter 4 examines the effect on price of a place name when the product is, in fact, not from the place in question; to wit, sparkling wine sold in the United States as “champagne” when it is obviously not from the Champagne region. Hedonic price equations are estimated, and when French and U.S. data are pooled, the variable that indicates “champagne” in the label raises the price for the French product but lowers it for the U.S. product. Chapter 5 examines the interesting question of whether bottle size has an effect on the (scaled) price. It appears that value is increasing in bottle size, and interesting charts demonstrate that well. A quibble: it should be noted that the symbols for various products are occasionally difficult to decipher―this reviewer could not tell the difference between the symbol used for Veuve Clicquot and for Dom Perignon even with the aid of a magnifying glass. Chapter 6 examines, with the aid of an ordered probit model, whether three professional panels of wine judges show any consensus about Italian wines and tests in particular how the fraction of wines in a region is related to the fraction receiving awards. On the whole, the agree- ment of the three sets of panels is weak, which underscores the conventional wisdom that people (even experts) do not have reliable taste buds. The final chapter on consum- er behavior and prices examines the effect of appellation of origin: it turns out that label information has a positive effect on the probability of purchase by the consumer, as it does on the prices that people are willing to pay.
Part II of the volume starts with an essay by Michel Balinski and Rida Laraki on the issue of how to rank wines. This is a subject that has been much written about, and the chapter is, in fact, an encapsulated version of Balinki and Laraki’s much bigger work, Majority Judgment (MIT Press, 2011; ISBN 978-0-262-01513-4). For a work of intermediate length, the curious reader may also wish to consult Michel Balinski and Rida Laraki, “Judge: Don’t Vote” (Operations Research, 62  , 483–511). Balinski and Laraki make a definitive contribution by showing that the frequently used methods of ranking the wines and then determining the “best” wine by computing the rank sums for each wine are seriously deficient in that they are “incoherent,” or, alternatively, they violate the axiom of the indepen- dence of irrelevant alternatives. If wines are assigned grades à la Parker, the results have another truly terrible feature in that they are manipulable. Balinski and Laraki’s (B&L’s) method consists of finding each wine’s “majority grade,” that is, the grade that is supported by a majority against any other grade, which, in the absence of pesky problems such as ties, turns out to be the median grade. We would then declare the wine with the highest majority grade the winner. Of course, it would be desirable that the method be “strategy proof in ranking,” that is, induce each taster to give his or her honest evaluation of each wine, but it is un- fortunately the case that no method is strategy proof in ranking, although B&L’s method is the least subject to this criticism. (Proofs are omitted in this chapter but can be found in their larger work.) It will be interesting to see how long the profes- sion of wine raters will take to adopt B&L’s methodology, which everybody should. I am not holding my breath because we are all conservative and change only slowly and because we all have much invested in existing computer programs―one hates nothing more than having to discard a program that works well and writing a new one. In any event, this is a fundamental contribution and also stands out by being the only one in the volume that is purely theoretical.
In Chapter 9, we find out about wine as a financial investment and learn that it tends to have a low correlation with other assets, which would make investment in wine a reasonable way of diversifying one’s portfolio, with the caveat that “wine is not a panacea that will always yield superb investment” (p. 196). In Chapter 10, we learn, perhaps not unsurprisingly, that the behavior of French wine companies responds to the business cycle in terms of capital management and other policies.
Part III of the volume begins with a chapter on the technical efficiency of grape growers. It estimates production frontier functions popularized in the 1970s by Dennis Aigner and finds via likelihood ratio tests that the translog production func- tion beats the Cobb-Douglas production function. A generalization of the model permits the use of panel data and accommodates time-varying technical efficiency. For the Australian data employed here, it turns out that there are significant techni- cal inefficiencies and evidence of technical change. Most interestingly, the shadow price of water indicates that farmers underutilize water. A quibble: there is an error in Figure 11-3, in that the vertical scale has separate two notches that are marked with the frequency 15. Chapter 12 deals with product assortment and efficiency and relies on the relatively rarely employed data envelopment analysis of Abraham Charnes. Chapter 13 analyzes wine bottling decisions with the aid of a sequential probit model, whereas Chapter 14 is devoted to the export intentions of wineries.
All the chapters deal with important as well as interesting issues, and all are char- acterize by high levels of technical competence. With the possible exception of Chapter 8, they will all be of interest to econometricians and scholars researching the wine industry, while Chapter 8 should be of prime interest to all who are engaged in rating or grading wines. The editors, Eric Giraud-Héraud and Marie- Claude Pichery, are to be congratulated for having rendered a great service to oenophiles.
Richard E. Quandt