\documentclass[a4paper,12pt]{article}
\usepackage[a4paper,hmargin={2.5cm,1.5cm},vmargin={4cm,3cm}]{geometry}
\usepackage[english]{babel}
\usepackage{eurosym}
%\usepackage[latin1]{inputenc}
\usepackage[utf8]{inputenc}
\usepackage{color}
\usepackage{indentfirst}
\usepackage{amsmath}
\usepackage{graphicx}
\usepackage{float}
\usepackage{rotfloat}
\usepackage{rotating}
\usepackage{adjustbox}
\usepackage{multirow}
\usepackage{amssymb}
\usepackage{indentfirst}
\usepackage{dsfont}
\usepackage{textcomp}
\usepackage[T1]{fontenc}
\usepackage{lmodern}
\usepackage{pifont}
\usepackage{graphicx}
\long\def\symbolfootnote[#1]#2{\begingroup%
\def\thefootnote{\fnsymbol{footnote}}\footnote[#1]{#2}\endgroup}
% \renewcommand{\baselinestretch}{1.2}
%\makeatletter
%\renewcommand\theequation{\thesection.\arabic{equation}}
%\@addtoreset{equation}{section} \makeatother
\font\dsrom=dsrom10 scaled 1200
\def \ind{\textrm{\dsrom{1}}}
\pagestyle{plain}
\begin{document}
\title{%
%TCIMACRO{\TeXButton{Begin single space}{\begin{singlespace}}}%
%BeginExpansion
%\begin{singlespace}%
%EndExpansion
The costs and benefits of collective reputation: Who gains and who loses from generic promotion programs?\footnote{We thank Andrew Clark, Tom Coup\'{e}, Victor Ginsburgh, Garrett Glasgow, Abdul Noury and seminar participants at the University of Reims and University of Paris I for helpful comments and suggestions. All errors are ours.}%
%TCIMACRO{\TeXButton{End Single Space}{\end{singlespace}}}%
%BeginExpansion
%\end{singlespace}%
%EndExpansion
}
\author{
Olivier Gergaud\thanks{%
KEDGE Business School, email: olivier.gergaud@kedgebs.com} \\
%EndAName
\and Florine Livat\thanks{%
KEDGE Business School, email: florine.livat@kedgebs.com} \\
%EndAName
\and Bradley Rickard\thanks{%
Cornell University, email: b.rickard@cornell.edu} \\
%EndAName
\and Frederic Warzynski\thanks{%
Aarhus University, email: fwa@asb.dk}} %\\
%EndAName
\maketitle
\begin{abstract}
%TCIMACRO{\TeXButton{Begin single space}{\begin{singlespace}}}%
%BeginExpansion
%\begin{singlespace}%
%EndExpansion
In this paper we develop an original approach to evaluate the costs and benefits associated to a generic promotion program using an application to Bordeaux wines. The benefit is computed from the marginal impact of the collective reputation of the program on the individual reputation of its members. These different marginal impacts are estimated using detailed survey data about the image of Bordeaux wines in seven European countries. We find positive and significant spillover effects from the umbrella reputation (Bordeaux) that moreover increase with the individual reputation level of the wine. Controlling for the natural endogeneity of the collective reputation in this setup, we capture the important fact that this relationship is faced with marginal diminishing returns. These spillover effects, when significantly positive, vary from a minimum of 5\% to a maximum of 15\% of additional favorable quality opinions. We then show that some subregions are more likely to benefit from generic promotion programs, suggesting that fees should be established on a benefit-cost basis.
\textbf{Key Words:} Benefit-cost analysis, Individual reputation, Collective reputation, Bordeaux wines, Appellations.
\textbf{JEL Classification: }L15 - L66 - Q13 - Z13
%TCIMACRO{\TeXButton{End Single Space}{\end{singlespace}}}%
%BeginExpansion
%\end{singlespace}%
%EndExpansion
\end{abstract}
\newpage
\baselineskip1.5em
\section{Introduction}
Agricultural economists have a long tradition of evaluating the net benefits of both domestic and export promotion programs. This is typically done using time series data to estimate demand for the commodity in question as a function of prices, income, seasonality constraints, and promotion expenditures. The estimated coefficient for promotion expenditures is used to quantify the additional revenue generated by the promotion efforts. Given the fees associated with the check-off program, a benefit-cost ratio (BCR) can subsequently be calculated to show the net economic benefits for a specific promotion effort. In a review of a wide range of agricultural commodities, Kaiser (2011) reports that the median BCR for generic promotion programs in the United States has been approximately 6.0. That is, for each dollar invested in promotion, the average increase in industry-wide profits was 6.00, and in many cases it has been found that producers could have profitably invested more in promotion, not less. Examples of estimated average BCRs for major commodities include 5.7 for beef (Ward, 1996), 16.0 for pork (Davis et al., 2000), 3.4 for dairy (Kaiser, 1997), and between 2.9 and 7.0 for orange juice (Williams et al., 2004). The overwhelming bulk of empirical evidence supports the notion that generic advertising has a positive and statistically significant impact on the demand for agricultural commodities and that there are gains to producers from these programs net of costs.
In this paper, we shed new light on the benefits of the generic promotion program for selected subregions within Bordeaux. Because producers are required to fund these programs, it is important to conduct the appropriate economic analysis to better understand their net benefits for producers. Different subregions within Bordeaux pay different per unit fees towards the greater promotion effort. The variety of fees suggests that some subregions may have greater capacity to contribute, but it also suggests that some subregions earn a disproportionally greater share of the benefits from the promotion effort. %Our results would therefore tend to suggest that fees for subregions within a larger appellation should be established on a cost-benefit basis.
Also, in the case of generic advertising for Bordeaux wines, there is much greater differentiation across the products being promoted that what is typically done. We propose here an original estimation strategy to assess the marginal benefits of this specific promotion program. Indeed, we first estimate the impact of the reputation of the group – the collective reputation premium - on the reputation of its members.
Using a detailed survey about the image of Bordeaux wines in seven European countries, we show that the magnitude of this reputation premium varies positively with the individual reputation level. In this specific context, the most reputed wine appellations are those that enjoy the highest reputation returns from the collective reputation, or the Bordeaux umbrella. In a second stage, we use the estimated effects from the generic promotion to compute a measure of the annual monetary reward from promotion for producers in the subregions. The measure of the monetary reward from the generic promotion is then compared to the annual costs that producers contribute in order to assess the net economic implications of the program for producers in each subregion.
Our main theoretical inspiration comes from Tirole’s (1996) collective reputation theory, where the collective reputation emerges as an aggregate of individual reputations, and belonging to a higher reputation group generates higher rents. While his analysis focuses on the incentives effects, the aim of our empirical work is to measure asymmetric benefits from collective reputation and what it implies.
Besides Tirole, our paper is also related to the umbrella branding literature, where collective reputation effects are analyzed from the point of view of the multi-product firm. This literature is mostly concerned with brand extension, i.e. the use of an established brand name to launch a product in a new market in order to reduce introductory costs (see Tauber, 1988). A collective brand or name may also act as a quality signal through spillovers that create reputation linkages among various products or individuals (Choi et al., 1995). In this context, individual incentives are associated with those of the group, and this mechanism provides a strong commitment to maintain a high quality level for each product.
Closer to us, Winfree and McCluskey (2005) explore, both theoretically and empirically, a market situation where several producers of a differentiated product (apples) are concerned with a single collective name at the regional level (Washington State). In such a context, where a single name is used by several producers, the collective reputation becomes a public good and the incentives to provide quality decrease as the size of the group increases (free riding on quality). Indeed, it is impossible to exclude a producer from the benefits of the umbrella and there is non-rivalry in the sense that the use of the collective name from one producer does not prevent another one from using the same name at the same time. Rickard, McCluskey, and Patterson (2015) use an experiment to understand how references to French umbrella reputations by U.S. wine regions influence consumers; it is quite common to see U.S. wine regions informally compare themselves to famous French wine appellations. They find that such references have the capacity to increase consumer valuation for wines in burgeoning U.S. wine regions, and the research highlights how collective reputations can even affect individual reputations outside of the umbrella region.
In a seminal application to Bordeaux wines, Landon and Smith (1997, 1998) show that both individual and collective reputations account for a substantial fraction of price variations observed for this product. Here, the collective reputation refers to the appellation name and individual reputations at the firm level are proxied by the average ratings the wines have received from a popular wine guide. In the Californian wine industry, Costanigro et al. (2010) show that consumers are willing to pay for more information to form accurate quality expectations on specific names when prices (i.e. opportunity costs) are high, while they accept to use aggregated names for inexpensive products. For Mosel Valley wines, Frick (2010) finds statistically significant non-linear returns for individual reputation as well as significant returns for collective reputation. However, none of these studies carefully look at the interaction between individual and collective reputation, and this is the main contribution of our paper.
The rest of the paper is organized as follows. Section 2 is a brief survey of the various evaluation methods used to evaluate generic promotion programs. Section 3 describes the empirical model, while section 4 introduces our survey data. Section 5 discusses our empirical strategy and main results on the interaction between individual and collective reputation. Section 6 details our costs and benefits analysis. Section 7 concludes.
\section{Economic evaluation of generic promotion programs}
The primary purpose for generic promotion programs is to generate net benefits to those that fund the efforts, which in most cases in a group of producers that pay a check-off or a promotion fee (Alston et al., 2007; Kinnucan and Myrland, 2008; Zhao, Anderson, and Wittwer, 2003). Generic (or umbrella) advertising is a cooperative effort among producers of a nearly homogeneous product to disseminate information about the underlying attributes of the product to existing and potential consumers for the purpose of strengthening demand for the commodity (Forker and Ward, 1993). These programs have evolved from relying on voluntary contributions to requiring mandatory participation. The reason for the switch is that voluntary programs, while generally successful immediately after the programs are established, have been plagued by free-rider problems over time (Messer, Kaiser, and Schulze, 2008). Such promotion or check-off programs exist for a wide range of agricultural commodities in the EU, in the United States, and elsewhere (Carman and Alston, 2005). Assessments are typically applied per unit of output and therefore larger firms contribute a larger share of the total promotion budget. Larger agricultural firms may also use branded advertising efforts to promote their products, and as a result there have been a number of controversial legal cases in the United States where large firms have requested to leave the mandatory generic program (Crespi, 2003).
It is difficult to evaluate the effects of promotion for Bordeaux wines following the approach that has been employed in the agricultural economics literature given the data constraints. It is complicated because the quantity of wines produced in each subregion in Bordeaux each year is relatively constant, and therefore it is difficult to directly estimate the effect of the promotion expenditures on demand (i.e., the promotion elasticity measure).
Instead, we propose a novel approach to examine the net returns to wine producers from the generic promotion effort. We start by describing the profits to a producer of wine in appellation $i$ as $\Pi_{i}=P_{i}Q_{i}-C_{i}-fee_{i}$, where revenue is equal to the product of the wine's price, denoted $P_{i}$, and quantity, denoted $Q_{i}$. The term $C_{i}$ describes costs for wine $i$ including all production and marketing costs and $fee_{i}$ refers to the additional costs used to support the generic promotion program. For our purposes, we assume that $Q_{i}$ and $C_{i}$ are fixed and outside the scope of this analysis, and we focus on $P_{i}$ and $fee_{i}$.
Following Landon and Smith (1998), Costanigro, McCluskey and Goemans (2010) and Rickard, McCluskey and Patterson (2015), we model the differentiated wine products in Bordeaux in a hedonic price framework to value bundled product attributes that are not marketed individually. The model is in the form $P_{i}=\Psi(q_{i},z_{i})$ , where % $P_{i}$ is the price of product from subregion $i$
$\Psi$ is the hedonic function relating product prices and attributes, $q_{i}$ is a measure of quality that exists and is known to consumers, and $z_{i}$ is a vector of other product attributes. For experience goods, consumers approximate $q_{i}$ with reputation associated with a particular product. The quality expectation is partly driven by the reputation of subregion i, denoted as $r_{i}$, and the collective reputation associated with the $k$-th region of production is denoted as $R_{k}$. Introducing a time dimension by using the subscript $t$, we express the price as a function of subregional reputation, the collective reputation, and product attributes as $P_{i}=\Psi(r_{it},R_{kt},z_{i})$. Adding a vector of parameters, $\beta$ , and an independently and identically distributed stochastic error term, we can express the equilibrium hedonic price as $P_{ikt}=\Psi(r_{it},R_{kt},z_{i};\beta)+\epsilon_{it}$.
It is widely agreed that agents form quality expectations based on past performance and signals of past performance (e.g., Shapiro 1982; Winfree and McCluskey 2005), yet the exact relationship linking quality performance, reputation and prices remains unknown. Here we estimate the contribution of the reputation (the joint effect of subregional reputation and the collective reputation) to the average price of wines that producers receive in each subregion. This benefit to producers due to the reputation is compared to the costs that producers pay for the generic promotion efforts, and we provide a benefit-cost ratio across the subregions that we study.
\section{Empirical Reputation Model}
Denoting $h$ as an index for individual survey respondents, $i=1,...,n$ as an index for the various appellations and $g$ as a group index (which in our case is the Bordeaux region), we can write the perceived quality
of the group and each sub-appellation $i$ by individual $h$ ($q_{g}^{h}$ and $q_{i}^{h}$) as:
\bigskip
\begin{center}
$\left\{
\begin{array}{l}
q_{g}^{h}=X_{g}^{h}\beta _{g}+\overset{n}{\underset{i=1}{\sum }}%
q_{i}^{h}\gamma _{i}+\varepsilon _{g}^{h}\ \ \ \ \ (0) \\
q_{1}^{h}=X_{1}^{h}\beta _{1}+\delta _{1}q_{g}^{h}+\varepsilon
_{1}^{h}\qquad \ \ \ (1) \\
\multicolumn{1}{c}{...} \\
q_{n}^{h}=X_{n}^{h}\beta _{n}+\delta _{n}q_{g}^{h}+\varepsilon
_{n}^{h}\qquad \ \ \ (n)%
\end{array}%
\right. $
\bigskip
\end{center}
where $X_{g}^{h}$ and $X_{i}^{h}$ are vectors of exogenous variables including the characteristics of individual $h$ like his/her self-assessed degree of knowledge of wine, his/her region of origin and their socio-professional category (upper, medium, lower incomes). These two vectors also contain information on past consumption (whether consumer $h$ experienced $i$ or $g$ at least once in the past 12 months or not) and a dummy variable which informs us whether consumer $h$ is familiar with the wine or not.
The parameters $\delta _{1}...\delta _{n}$ capture the average impact of $%
q_{g}^{h},$ the collective reputation, on the various individual reputations
($q_{i}^{h}$). The parameters $\gamma _{1}...\gamma _{n}$ measure the
contribution of each individual reputation to $q_{g}^{h}$.
By construction, $q_{g}^{h}$ and $q_{i}^{h}$ are endogenous variables. This means that $\varepsilon _{g}^{h}$ is potentially correlated to every $q_{i}^{h}$ and
$\varepsilon _{i}^{h}$ is not independent of $q_{g}^{h}$. We would therefore
need valid instruments for $q_{g}^{h}$ and each $q_{i}^{h}$.
To simplify the problem, we focus exclusively on the estimation of equations
(1) to (n) for which we only require instruments for $q_{g}^{h}$.
We use as instrument ($Z_{g}^{h}$) the answer to what surveyed consumers
think about the quality of other famous French appellations such as Alsace ($q_{Al}^{h}$),
Beaujolais ($q_{Be}^{h}$), Burgundy ($q_{Bu}^{h}$), C\^{o}tes du Rh\^{o}ne ($q_{Cr}^{h}$),
Languedoc-Roussillon ($q_{Lr}^{h}$) and Loire Valley ($q_{Lv}^{h}$) (see Map 1).
These appellations are umbrella brands in the same way as $q_{g}^{h}$.
The intuition for the validity of these instruments is that wine consumers
imagine the quality of a Bordeaux wine by comparing it with the quality of some
of its closest competitors. Indeed, it appears reasonable to assume that
these opinions on Bordeaux wines will be based, among other
things, on a sort of ranking of the main wines produced in France. On the
other hand, while it makes sense to believe that wine consumers will compare
a Bordeaux with a Beaujolais for instance (which are two regional
appellations), they will not compare so naturally (i.e. frequently) a C\^{o}%
tes-de-Bourg which is a sub-appellation in the Bordeaux region with a
regional appellation such as Burgundy. The main reason for this intuition is
that C\^{o}tes-de-Bourg and Burgundy are not at the same level in the French
wine classification system which is based on two types of appellations:
regional (Burgundy, Bordeaux, etc.) and local/village (C\^{o}tes de Bourg,
Margaux, etc.). For all of these reasons we expect these variables ($%
q_{Al}^{h}$, $q_{Be}^{h}$, $q_{Bu}^{h}$, $q_{Cr}^{h}$, $q_{Lr}^{h}$, $%
q_{Lv}^{h}$) to be correlated to $q_{g}^{h}$ and independent of every $%
q_{i}^{h}$.
\section{Data}
Survey data\footnote{In this survey, wine consumers were selected only if they drank wine at least once a quarter.} were collected in seven European countries: Belgium (1,028 wine consumers), Denmark (613 wine consumers), Germany (1,133 wine consumers), France (819 wine consumers), the Netherlands (1,258 wine consumers), Switzerland (584 wine consumers), United Kingdom (959 wine consumers). The survey was conducted in 2001 by \textit{Sociovision} on behalf of the \textit{Comit\'{e} Interprofessionnel des vins de Bordeaux} and includes information from 6,394 respondents. Although the data come from a survey conducted in 2001, we do not anticipate that consumers' evaluations for wines across French subregions have changed susbstantially in more recent years. Furthermore, the data are only available in 2001 and the survey contains a rich amount of detail concerning wine preferences for a large number of consumers.
Respondents were on average 46 years old and 51\% of them were women. All respondents had purchased wine within 90 days of taking the survey; nearly
one-third (32\%) of the sample participants perceive themselves as wine
connoisseurs, while 66\% estimate that they are not highly knowledgeable in wine. People were first invited to give their opinion on
French wines in general (Alsace, Beaujolais, Bordeaux, Bourgogne, C\^{o}tes
du Rh\^{o}ne, Languedoc-Roussillon, etc.) then on 9\ Bordeaux
sub-appellations: Bordeaux Sup\'{e}rieur (BSUP), C\^{o}tes de Bourg (CBG),
Entre-deux-Mers (E2M), Graves (GR), Margaux (MGX), M\'{e}doc (MDC), Premi%
\`{e}res C\^{o}tes de Bordeaux (PCB), Saint-Emilion (SEM) and Sauternes
(SAU).
Table 1 shows the share of favorable opinions for each wine ($q_{i}^{h}$ and
$q_{g}^{h}$) including the instrumental variables ($q_{Al}^{h}$, $q_{Be}^{h}$%
, $q_{Bu}^{h}$, $q_{Cr}^{h}$, $q_{Lr}^{h}$, $q_{Lv}^{h}$).\ This informs us
about the way the quality of these wines is perceived on average among the subjects in our sample. %in these countries.
With a level of agreement on quality higher than 50\%, Bordeaux is clearly
the most appreciated French wine appellation in all Western Europe, followed
by Saint-Emilion, Bordeaux Sup\'{e}rieur, Sauternes, M\'{e}doc. These
appellations have a score of more than 20\% of favorable quality opinions.
The other wine regions appear to be far less well reputed, with their
reputation level not exceeding 20\%. With the exception of Beaujolais
(17.91\%), the other wines produced in France have reputation levels lower
than 10\%.
\section{Estimation Procedure and Results}
\subsection{2-Stage Least Squares}
We first estimate a series of recursive models using a simple 2SLS estimation procedure\footnote{%
For simplicity, we assume that there is no image spillovers between the
different individual reputations and focus on the relationship with the group reputation.}:
\begin{center}
\bigskip $\left\{
\begin{array}{l}
q_{i}^{h}=X_{i}^{h}\beta _{i}+\delta _{i}q_{g}^{h}+\varepsilon
_{i}^{h}\qquad (i) \\
q_{g}^{h}=X_{g}^{h}\beta _{g}+Z_{g}^{h}\theta _{i}+\varepsilon _{g}^{h}\ \ \
\ (0)%
\end{array}%
\right. $
\end{center}
In this setup, $q_{g}^{h}$ is regressed in the collective reputation
equation (0) against $X_{g}^{h}$ and the instruments $Z_{g}^{h}$ ; whereas $%
q_{i}^{h}$ is regressed against $X_{i}^{h}$ and $q_{g}^{h}$ in the
individual reputation equation ($i$). A system like this has been estimated
for each of the 9 appellations beneath the Bordeaux umbrella ($i=1,...,9$).
The results are listed within Table 2 along with those of a battery of tests for
the endogeneity of $q_{g}^{h}$, the validity or weakness of the instruments
(Hansen's J, Stock and Yogo) in Table 3. Whenever necessary we tested the
exogeneity of one or more questionable instruments using the
"Difference-in-Sargan" statistic also known as the C-Statistic. Figure 2
summarizes the results of the 2SLS procedure which does not allow us to
control for the fact that the quality variables are of the binary type.
The instruments $Z_{g}^{h}$ turned out to be reasonable predictors of what
people think about Bordeaux as a generic appellation. Among these,
Beaujolais and Languedoc-Roussillon turned out to be the most predictive
(significant).\footnote{%
The results for the first step equations are available from the authors upon
request.} The highest relative bias that we get (20-30\%) concerns only one
regression (C\^{o}tes-de-Bourg) in which $q_{g}^{h}$ were not found
endogenous. In the other regressions, the relative bias potentially induced
by the weakness of the instruments is quite acceptable (between 10\% and
20\% in two regressions and lower than 10\% in the others). The results of
the various Hansen's overidentification tests failed to reject the
hypothesis that the instruments are exogenous in every regression.
\subsection{Robustness check}
As a robustness check, we ran a second series of regressions using a
Recursive Bivariate Probit (RBP) procedure which is more appropriate given
that both $q_{i}^{h}$ and $q_{g}^{h}$ are of the binary type. The RBP
results (Table 4) are then compared to those obtained after a regular ML
probit estimation procedure (Table 5) which ignores the potential
endogeneity of $q_{g}^{h}$ in each equation.% $i$ (see Figure 3 for ease of comparison).
The results are striking. Notably, $q_{g}^{h}$ came out endogenous in most systems we estimated. Indeed, in most regressions the exogeneity tests rejected the
hypothesis that $q_{g}^{h}$ is exogenous. The exceptions are C\^{o}%
tes-de-Bourg and Entre-deux-Mers, the two less well reputed appellations
(7.57\% and 7.65\% respectively). Not controlling for this endogeneity
pitfall results in a downward bias in the estimated returns to collective
reputation. From Figure 3 we observe that ML\ Probit tend to systematically
underestimate the various impacts compared with those obtained from an
appropriate RBP estimation procedure. Moreover, it fails to capture the fact
that this relationship exhibits marginal diminishing returns (concave shape
with RBP versus more linear shape with ML\ Probit). In other words, the
marginal impact of the Bordeaux reputation (the "umbrella brand") actually tends to decrease to zero (and not to increase in a linear way) as the reputation level of its entities goes up.
We get positive and significant spillover effects from the umbrella
reputation for 8\ individual appellations out of 9. Highly-reputed
appellations are found to enjoy larger umbrella impacts than less-reputed
appellations. These image spillover effects when positive vary from a
minimum of 5\%\ to a maximum of 15\% of additional favorable quality
opinions (see Figure 4 which reproduces on the vertical axis the marginal
effects in percentage points for the RBP estimates).
In this group, only the leaders take a significant advantage from the high
level of reputation of Bordeaux. For the followers, there is no advantage in
being part of this group as they are not clearly associated to Bordeaux in
the consumer's mind. This is particularly true for Entre-deux-Mers and C\^{o}%
tes de Bourg which do not enjoy any benefits from the fact
that they fall in the Bordeaux region.
\section{Cost and Benefit Analysis and Industry implications}
\subsection{The Cost Dimension: How are Fees Determined}
The Bordeaux Wine Council (CIVB in French), founded in 1948, represents the three entities of the Bordeaux wine industry: winegrowers, wine merchants and brokers. The missions are the following ones (Source: CIVB):
\begin{itemize}
\item Marketing: develop the reputation of Bordeaux wines, in France and abroad, through advertising campaigns, digital communications, public and press relations, and training.
\item Economic: acquiring data and improving knowledge relating to the production, the markets and the sale of Bordeaux wines throughout the world.
\item Technical: improve the industry's understanding of various technical issues relating to the production and quality of Bordeaux wines and anticipate new environment - and health-related requirements.
\end{itemize}
These missions are all costly activities that are funded through appellation fees. The fees are revised every four years, and historically the largest share of the fees are used to support marketing efforts and to fund promotion programs for Bordeaux wines. The amount of this fee varies substantially from \euro{5.65} per hectoliter in Entre-deux-Mers up to \euro{12.43} per hectoliter for wines produced in the Margaux area (see Table 6 for further details). The fee is positively associated with the level of reputation of the wines produced in each subregion (Pearson correlation coefficient = 0.51) but also with market prices (Pearson correlation coefficient = 0.35).
We use the econometric results estimated using our survey data that were reported in section 5 to calculate an average measure of the benefits from promotion in euros per hectoliter (Hl) for producers in each subregion. The measure calculates a measure that describes the proportion of the average prices for wines in the subregions that can be attributed to that subregion's reputation. We compare this measure of the average benefits of promotion to the average cost in fees for the generic promotion paid by producers (also in euros per Hl) to calculate the average BCR by subregion.
Table 6 summarizes the information we used to calculate the average BCR for wine producers in the various selected subregions in Bordeaux. The first column labelled "Individual reputation" is a critical piece in the calculation of the average BCR. The values sown in this column show the average quality score earned by subregions across all respondents in the survey; individual respondents simply chose whether the subregion was associated with a high quality wine (a quality score of 1) or not (a quality score of 0). We use this average quality score as a proxy for the link between reputation and price, and implicitly assume that percentage changes in the quality scores are fully incorporated into changes in the producer (bulk) price of the wine. Altering this assumption will affect our numerical results, but it will not affect the rank ordering of our results.
The information listed in columns 2 through 6 outline the parameters needed to calculate the "Net flow" shown in column 7 and the average BCR in column 8. The second column provides the average benefit of the collective reputation on the reputation of the subregion, and is calculated as the marginal effect of the estimated coefficients for the subregions shown in Table 4. The third column provides the statistical significance levels for the econometric estimates shown in the previous column. The fourth column reports the fees paid by Bordeaux producers during the 2011-2014 period and the fifth column lists the average market prices for bulk wines produced in each subregion. The sixth column shows the net benefit of the promotion efforts to producers in the various subregions. Here we take the product of the "Individual reputation" score and the effect of the collective reputation on the subregion to generate an aggregate reputation score by subregion; we then multiply the aggregate reputation score by the average market price for wine in each subregion to calculate the benefits of the promotion efforts (in euros per Hl). The difference between the calculated value of benefits and the fees is shown as the "Net flow" in column 7, and the average BCR is shown in the final column.
\subsection{Findings and policy implications}
Here we present the results for our measure of the net benefits from the generic promotion efforts for Bordeaux wines across the nine subregions. The final column in Table 6 shows the calculated average BCRs by subregion; this measure is the ratio of the benefits reported in the sixth column to the fees shown in the fourth column. For the following discussion we assume that the benefits from promotion in the Côtes de Bourg and Entre-deux Mers subregions are zero given that the reputation estimates were not statistically significant for these subregions. The bulk of our discussion below focuses on the BCR calculations for the remaining seven subregions.
As shown in Table 6, the fees range between \euro{5.65} and \euro{12.43} (taxes included) per Hl, and the average market prices of the wines range between \euro{111} and \euro{842} per Hl; in general, the subregions with higher fees also experience higher average market prices, but in the middle group with fees of \euro{9.32} per Hl we see a wide range of average market prices. Because there is no estimated effect from the umbrella advertising program for Côtes de Bourg and in Entre-Deux Mers, these regions have no calculated benefits and the BCR is zero in both cases. For the next four regions in the list - Premieres Côtes de Bordeaux, Bordeaux Supérieur, Graves, and Médoc - we calculate an average benefit of between \euro{1.26} and \euro{4.70} per Hl that is associated with the umbrella promotion program; however, in all four cases the costs outweigh the benefits and each sub-appellation shows a BCR less than one. Within this group, the Bordeaux Supérieur sub-appellation has the highest BCR (at 0.63) and this is largely driven by its higher regional reputation effect (probably due to a favorable name that closely associates itself to the umbrella name). The final three regions in the list - Sauternes, Saint-Emilion, and Margaux - have higher average market prices and subsequently higher calculated benefits per Hl. Two of these subregions also have fees of \euro{9.32} per Hl and Margaux has a slightly higher fee of \euro{12.43} per Hl, and all three regions have BCRs that exceed 1.0. The highest BCR is for Sauternes given the estimate for its sub-regional reputation and its relatively low fee structure.
Results in Table 6 highlight some interesting differences concerning the net benefits of the promotion program across the subregions. In addition, our findings may have important implications for wine producers and the other stakeholders that manage the structure of the funds collected and used to promote Bordeaux wines. First, the non-weighted average BCR across all nine sub-appellations is 0.7 (it is 0.9 for the seven regions with strictly positive BCRs) indicating that the promotion program may not be generating net benefits to the region overall. Second, some of the subregions, namely those with relatively lower fees compared to average market prices, are the ones receiving the net benefits from the promotion program. Our calculation of the BCR indicates that only three of the nine subregions are benefiting from the program; the other six regions that appear to be cross-subsidizing the promotion program may be better off if they left the program. Third, our results give us some sense for how the fees might be adjusted across regions as a way to realign the fee structure with the benefits of the promotion efforts. In particular, the sixth column in Table 6 (labeled Marginal Benefit) outlines the maximum fee that producers should be willing to pay per Hl in each subregion. However, the fees that are charged in each subregion appear to be sticky as they have not changed for any subregion in the most recent fee schedule for the period 2014 to 2017\footnote{See $https://info.agriculture.gouv.fr/gedei/site/bo-agri/document_administratif-987ec417-5c6d-4b7c-85c1-59faa6dd7606/telechargement$}. Although we expect that there would be much resistance to any changes in the fee structure, our findings suggest that the current arrangement may not be optimal for individual wineries or for the Bordeaux region more generally.
\section{Conclusion}
The success and stability of generic promotion programs depend to a large extent on their effectiveness and cost to participants. In this paper, we measured the influence of Bordeaux as a brand on a series of 9 appellations beneath its umbrella to assess the average benefit of this generic promotion program and then compare it with their respective marginal cost. Controlling for the fact that both types of reputation are released simultaneously, we get significant positive
spillover effects from the umbrella, the magnitude of which depends positively on the individual reputation level of the wine under the umbrella. The reputation of this prestigious wine appellation would thus also act as a positive quality signal among a significant fraction of surveyed people in Western Europe.
While collective reputation generally provides some benefits to individual group members, we find that these gains do not always compensate the costs of membership. Our results therefore suggest that promotion programs should take this cost-benefit consideration into account, and possibly better target their potential customers in various markets on an individual basis.
\newpage
\begin{thebibliography}{99}
\bibitem{} Alston, J.M., J.M. Crespi, H.M. Kaiser, and R.J. Sexton, 2007, 'An evaluation of California’s mandated commodity promotion programs', \textit{Review of Agricultural Economics} 29, 40-63.
%\bibitem{} Andersson, F., 2002, 'Pooling Reputations', \textit{International Journal of Industrial Organization} 20, pp. 715-730.
%\bibitem{} Bar-Isaac, E. and Tadelis, S., 2008, 'Seller Reputation', \textit{Foundations and Trends in Microeconomics} 4, 273-351.
%\bibitem{} Borenstein, S. and Zimmerman, M. B., 1988, 'Market Incentives for Safe Commercial Airline Operation', \textit{American Economic Review} 78, pp. 913-935.
%\bibitem{} Cabral, L., 2000, 'Stretching Firm and Brand Reputation', \textit{Rand Journal of Economics} 31, pp. 658-673.
\bibitem{} Carman, H.F., and Alston, J.M. (2005), 'California’s mandated commodity programs'. In H.M. Kaiser, J.M. Alston, J.M. Crespi, and R.J. Sexton (Eds.), The Economics of Commodity Promotion Programs: Lessons from California. New York City: Peter Lang Publishing, Inc.
\bibitem{} Choi, C.J., Lee, S.H. and Oh, D., 1995, 'The strategy of grouping and reputation linkage in clubs and multi-products firms', \textit{European Journal of Political Economy}, 11, pp. 521-533.
%\bibitem{} Choi, J.P., 1998, 'Brand Extension as Informational Leverage', \textit{Review of Economic Studies}, October, 655-670.
%\bibitem{} Choi, J.P., 2003, 'Bundling New Products with Old to Signal Quality, with Application to the Sequencing of New Products', \textit{International Journal of Industrial Organization} 21, 1179-1200.
%\bibitem{} Choi, J.P. and Jeon, D.S., 2007, 'A Leverage Theory of Reputation Building with Co-Branding: Complementarities in Reputation Building', \textit{Working Paper}, Michigan State University.
\bibitem{} Costanigro, M., J.J. McCluskey, and C. Goemans, 2010, 'The Economics of Nested Names: Name Specificity, Reputations, and Price Premia', \textit{American Journal of Agricultural Economics} 92, 1339-1350.
\bibitem{} Crespi, J.M., 2003, 'The Generic Advertising Controversy: How Did We Get Here and Where Are We Going?', \textit{Review of Agricultural Economics} 25, 294-315.
%\bibitem{} Damiano, E., Hao, L. and Suen W., 2010, 'First in a Village or Second in Rome?', \textit{International Economic Review} 51, 263-288.
\bibitem{} Davis, G.C., Capps, Jr., O., Bessler, D.A., Leigh, J.H., Nichols, J.P., and Goddard, E., 2000, An economic evaluation of the pork checkoff program. College Station: Texas A\&M University.
%\bibitem{} Erdem, T., 1998, 'An Empirical Analysis of Umbrella Branding', \textit{Journal of Marketing Research}, August, pp. 339-351.
\bibitem{} Forker, O.D., and Ward, R.W., 1993, Commodity advertising: The economics and measurement of generic programs. New York: Lexington Books.
\bibitem{} Frick, B., 2010, 'The Impact of Individual and Collective
Reputation on Wine Prices: Empirical Evidence from the Mosel Valley',
\textit{Working paper}, University of Paderborn.
%\bibitem{} Fudenberg, D., Tirole J., 1989, 'Noncooperative Game Theory for Industrial Organization: An Introduction and Overview', in Schmalensee, R., and Willig, R. (eds.), \textit{Handbook of Industrial Organization}, \textit{vol. 1}, pp. 259-327 (North Holland).
%\bibitem{} Hayek, F.A., 1948, \textit{Individualism and Economic Order}, University of Chicago Press, Chicago.
%\bibitem{} Jarrell, G. and Peltzman, S., 1985, 'The Impact of Products Recalls on the Wealth of Sellers', \textit{Journal of Political Economy} 93, 512-536.
\bibitem{} Kaiser, H.M., 1997, 'Impact of national dairy advertising on dairy markets, 1984–95', \textit{Journal of Agricultural and Applied Economics} 29, 303–314.
\bibitem{} Kaiser, H.M.,2011, 'Effects of generic advertising on food demand'. In J.L. Lusk, J. Roosen, and J. Shogren (Eds.), Oxford Handbook of the Economics of Food Consumption and Policy. Oxford, UK: Oxford University Press.
\bibitem{} Kinnucan, H.W., and Ø. Myrland, 2008, 'On Generic Versus Brand Promotion of Farm Products in Foreign Markets', \textit{Applied Economics} 40, 673-684.
%\bibitem{} Klein, B. and Leffler, K.B., 1981, 'The Role of Market Forces in Assuring Contractual Performance', \textit{Journal of Political Economy} 89, 615-41.
\bibitem{} Landon, S. and Smith C.E., 1997, 'The Use of Quality and
Reputation Indicators by Consumers : The Case of Bordeaux Wine', \textit{Journal of Consumer Policy} 20, 289-323.
\bibitem{} Landon, S. and Smith C.E., 1998, 'Quality Expectations,
Reputation and Price', \textit{Southern Economic Journal} 64, 628-647.
%\bibitem{} Levin, J., 2001, 'Career Concerns and Collective Reputation', \textit{Mimeo}, Stanford University.
%\bibitem{} Mailath, G.J. and Samuelson, L., 2002, 'Who Wants a Good Reputation?', \textit{Review of Economic Studies} 68, 415-441.
%\bibitem{} Marshall, A., 1949, \textit{Principles of Economics: An Introductory Volume}, Macmillan, New York, 8th edition.
\bibitem{} Messer, K., Kaiser, H.M., and Schulze, W., 2008, 'The problem with generic advertising: parallelism and possible solutions from the lab', \textit{American Journal of Agricultural Economics} 90, 540-552.
%\bibitem{} Rickard, B.J., Liaukonyte, J., Kaiser, H.M. and Richards, T.J., 2011, 'Consumer response to commodity-specific and broad-based promotion programs for fruits and vegetables', \textit{American Journal of Agricultural Economics} 93, 1312-1327.
\bibitem{} Rickard, B.J., McCluskey, J.J., and Patterson, R.W., 2015. 'Reputation tapping', \textit{European Review of Agricultural Economics} 42, 675-701.
\bibitem{} Shapiro, C., 1982, 'Consumer Information, Product Quality, and Seller Reputation', \textit{Bell Journal of Economics} 13, 20-35.
%\bibitem{} Shapiro, C., 1983, 'Premiums for High Quality Products as Returns to Reputations', \textit{Quarterly Journal of Economics}, 98, pp. 659-679.
%\bibitem{} Sullivan, M., 1990, 'Measuring Image Spillovers in Umbrella-Branded Products', \textit{Journal of Business} 63, pp. 309-329.
%\bibitem{} Tadelis, S., 2002, 'The Market for Reputations as an Incentive Mechanism', \textit{Journal of Political Economy,} 92, 854-882.
\bibitem{} Tauber, E.M., 1988, 'Brand leverage: Strategy for Growth in a
Cost-Control World', \textit{Journal of Advertising Research},
August/September, 26-30.
\bibitem{} Tirole, J., 1996, 'A Theory of Collective Reputations (with
Applications to the Persistence of Corruption and to Firm Quality)', \textit{The Review of Economic Studies} 63, 1-22.
\bibitem{} Ward, R.W., 1996, 'The beef checkoff impact on U.S. beef demand', Department of Agricultural Economics, University of Florida.
\bibitem{} Winfree, J.A, and McCluskey, J.J., 2005, 'Collective Reputation and Quality', American Journal of Agricultural Economics, 87(1), 206-213.
%\bibitem{} Wernerfelt, B., 1988, 'Umbrella Branding as a Signal of New Product Quality: An Example of Signalling by Posting a Bond', \textit{Rand Journal of Economics}, 19, pp. 458-466.
\bibitem{} Williams, G.W., Capps, Jr., O., and Bessler, D.A., 2004, 'Florida orange grower returns from orange juice advertising', Department of Agricultural Economics, Texas A\&M University.
%\bibitem{} Zeithaml, V.A., 1988, 'Consumer Perceptions of Price, Quality and Value: A Means-End Model and Synthesis of Evidence', \textit{Journal of Marketing} 53, 2-22.
\bibitem{} Zhao, X., K. Anderson, and G. Wittwer, 2003, 'Who gains from Australian generic wine promotion and R\&D?', \textit{Australian Journal of Agricultural and Resource Economics} 47, 181–209.
\end{thebibliography}
\newpage
%TCIMACRO{%
%\TeXButton{TeX field}{\renewcommand{\baselinestretch}{0.9}\small\normalsize}}%
%BeginExpansion
\renewcommand{\baselinestretch}{0.9}\small\normalsize%
%EndExpansion
%TCIMACRO{%
%\TeXButton{Table 1}{\begin{tabular}{lc}
%
%\multicolumn{ 2}{c}{{\bf Table 1: Reputation levels*}} \\
%
%{\bf } & {\bf } \\
%\hline
%\hline
%{\bf } & {\bf \ } \\
%
%
%{\it Bordeaux (umbrella)} & 50.08 \\
%
%& \\
%
%{\it Sub-Appellations within the Bordeaux Region:} & \\
%
%Bordeaux Sup�rieur & 25.21 \\
%
%Entre-deux-Mers & 7.65 \\
%
%Margaux & 19.21 \\
%
%M�doc & 21.14 \\
%
%Saint-Emilion & 25.6 \\
%
%C�tes de Bourg & 7.57 \\
%
%Graves & 19.32 \\
%
%Premi�res C�tes de Bordeaux & 13.85 \\
%
%Sauternes & 23.02 \\
%
%& \\
%
%{\it French wine regions (Instruments)}: & \\
%
%Alsace & 3.17 \\
%
%Beaujolais & 17.91 \\
%
%Bourgogne & 6.99 \\
%
%Languedoc-Roussillon & 8.52 \\
%
%C�tes du Rh�ne & 0.2 \\
%
%Loire & 6.05 \\
%
%& \\
%\hline
%\hline
%\multicolumn{ 2}{l}{* Average levels of agreement on quality (percentages)} \\
%
%\end{tabular}}}%
%BeginExpansion
\begin{table}
\begin{center}
\protect\caption{Summary statistics about reputation levels from the survey data*}
\begin{tabular}{lc}
%\multicolumn{ 2}{c}{{\bf Table 1: Reputation levels*}} \\
{\bf } & {\bf } \\
\hline
\hline
{\bf } & {\bf \ } \\
{\it Bordeaux (umbrella)} & 50.08 \\
& \\
{\it Sub-Appellations within the Bordeaux Region:} & \\
Bordeaux Supérieur & 25.21 \\
Entre-deux-Mers & 7.65 \\
Margaux & 19.21 \\
Médoc & 21.14 \\
Saint-Emilion & 25.6 \\
Côtes de Bourg & 7.57 \\
Graves & 19.32 \\
Premières Côtes de Bordeaux & 13.85 \\
Sauternes & 23.02 \\
& \\
{\it French wine regions (Instruments)}: & \\
Alsace & 3.17 \\
Beaujolais & 17.91 \\
Bourgogne & 6.99 \\
Languedoc-Roussillon & 8.52 \\
Côtes du Rhône & 0.2 \\
Loire & 6.05 \\
& \\
\hline
\hline
\multicolumn{ 2}{l}{* Average levels of agreement on quality (percentages)} \\
\end{tabular}%
\end{center}
\end{table}
%EndExpansion
%TCIMACRO{%
%\TeXButton{TeX field}{\renewcommand{\baselinestretch}{1.7}\small\normalsize}}%
%BeginExpansion
%\renewcommand{\baselinestretch}{1.7}\small\normalsize%
%EndExpansion
\begin{center}
\begin{sidewaystable}
\protect\caption{Main determinants of the individual reputation level (Two-Stage Least Squares)}
\begin{tabular}{ l l l l l l l l l l }
\hline
& Saint & Bordeaux & Médoc & Margaux & Entre-deux & Côtes de & Sauternes & Prem. Côtes & Graves \\
& Emilion & Supérieur & & & -mers & Bourg & & de Bordeaux & \\ \hline
Opinion on Bordeaux Wines & 0.555*** & 0.293*** & 0.346*** & 0.547*** & 0.0920** & -0.0777 & 0.393*** & 0.341*** & 0.362** \\
(Good = 1 ; Bad = 0) & (0.0918) & (0.0740) & (0.0743) & (0.0728) & (0.0439) & (0.105) & (0.130) & (0.0884) & (0.141) \\
& & & & & & & & & \\
Past consumption : & 0.148*** & -0.0290 & 0.112*** & 0.176*** & 0.0298* & 0.0575* & 0.125*** & -0.00747 & 0.0616 \\
Has consumed wine $i$ last year & (0.0202) & (0.0232) & (0.0184) & (0.0303) & (0.0153) & (0.0323) & (0.0326) & (0.0495) & (0.0419) \\
& & & & & & & & & \\
Recognition: & 0.0509** & 0.0515 & 0.0700*** & 0.222*** & 0.0142 & 0.0374 & 0.104** & 0.0684 & 0.125* \\
Respondent knows wine $i$ & (0.0249) & (0.0372) & (0.0207) & (0.0381) & (0.0252) & (0.0600) & (0.0502) & (0.0814) & (0.0661) \\
& & & & & & & & & \\
Constant & -0.0683 & 0.0746** & -0.0428 & -0.123*** & 0.0237 & 0.160** & 0.0843 & -0.0360 & 0.105 \\
& (0.0443) & (0.0376) & (0.0369) & (0.0377) & (0.0235) & (0.0804) & (0.0946) & (0.0399) & (0.0713) \\ \hline
Instruments: & & & & & & & & & \\
Alsace & Yes & Yes & No & No & Yes & Yes & Yes & Yes & Yes \\
Beaujolais & Yes & Yes & No & Yes & Yes & Yes & Yes & Yes & Yes \\
Bourgogne & Yes & Yes & Yes & Yes & Yes & Yes & Yes & Yes & Yes \\
C. du Rhône & Yes & Yes & Yes & Yes & No & Yes & Yes & Yes & Yes \\
Languedoc & No & Yes & Yes & Yes & Yes & Yes & Yes & Yes & Yes \\
Loire & Yes & No & Yes & No & Yes & Yes & Yes & Yes & Yes \\ \hline
Excluded instruments (C-stat): & 7.662 & 8.740 & & & 4.083 & - & - & - & - \\
C-stat. & (0.0056) & (0.0031) & & & (0.0433) & - & - & - & - \\ \hline
Observations & 6,307 & 6,307 & 6,307 & 6,307 & 6,307 & 819 & 1,382 & 3,667 & 1,258 \\
F stat & 20.77 & 11.51 & 16.88 & 17.46 & 6.90 & 1.60 & 8.44 & 6.67 & 2.88 \\
P-value & (0.0000) & (0.0000) & (0.0000) & (0.0000) & (0.0000) & (0.0500) & (0.0000) & (0.0000) & (0.0000) \\ \hline
\end{tabular}
{\small{}Exogenous controls include age, gender, region of origin of the respondent, socio-economic category and level of self-assessed wine-knowledge. Robust standard errors in parentheses ; *** p<0.01, ** p<0.05, * p<0.1}
\end{sidewaystable}
\end{center}
\begin{center}
\begin{sidewaystable}
\protect\caption{Exogeneity Tests and Weak Identification Tests}
\begin{tabular}{ l l l l l l l l l l }
\hline
& Saint & Bordeaux & Médoc & Margaux & Entre-deux & Côtes de & Sauternes & Prem. Côtes & Graves \\
& Emilion & Supérieur & & & -mers & Bourg & & de Bordeaux & \\ \hline
Exogeneity test (Umbrella) & 33.866 & 3.668 & 14.457 & 52.612 & 2.352 & 1.496 & 7.666 & 8.193 & 3.723 \\
& (0.0000) & (0.0555) & (0.0001) & (0.0000) & (0.1251) & (0.2213) & (0.0056) & (0.0042) & (0.0537) \\
Weak identification test: & & & & & & & & & \\
Kleibergen-Paap stat. & 18.071 & 21.509 & 23.137 & 28.261 & 23.850 & 6.454 & 7.642 & 11.119 & 7.410 \\
Relative bias & 5-10\% & 0 & 0 & 0 & 0 & 20-30\% & 10-20\% & 10-20\% & 10-20\% \\
Hansen's overid. test & 5.107 & 7.011 & 5.465 & 5.649 & 8.050 & 2.118 & 1.128 & 6.105 & 0.321 \\
& (0.5302) & (0.3198) & (0.3618) & (0.3419) & (0.2345) & (0.7140) & (0.9803) & (0.4115) & (0.9884) \\
Observations & 6,307 & 6,307 & 6,307 & 6,307 & 6,307 & 819 & 1,382 & 3,667 & 1,258 \\
F stat & 20.77 & 11.51 & 16.88 & 17.46 & 6.90 & 1.60 & 8.44 & 6.67 & 2.88 \\
P-value & (0.0000) & (0.0000) & (0.0000) & (0.0000) & (0.0000) & (0.0500) & (0.0000) & (0.0000) & (0.0000) \\ \hline
\end{tabular}
{\small{} Robust standard errors in parentheses ; *** p<0.01, ** p<0.05, * p<0.1
Tests based on 2SLS estimates}
\end{sidewaystable}
\newpage
\begin{sidewaystable}
\protect\caption{Main determinants of the individual reputation level (Recursive Bivariate Probit)}
\begin{tabular}{ l l l l l l l l l l }
\hline
& Saint & Bordeaux & Médoc & Margaux & Entre-deux & Côtes de & Sauternes & Prem. Côtes & Graves \\
& Emilion & Supérieur & & & -mers & Bourg & & de Bordeaux & \\ \hline
Opinion on Bordeaux Wines & 0.13*** & 0.1173*** & 0.1013*** & 0.1152*** & 0.0248 & -0.1072 & 0.1314** & 0.082*** & 0.1522** \\
(Good = 1 ; Bad = 0) & (0.0216 ) & (0.0356 ) & (0.0237) & (0.017) & (0.016) & (0.1533) & (0.0517) & (0.0289) & (0.0769) \\
& & & & & & & & & \\
Past consumption : & 0.0652*** & -0.008 & 0.0509*** & 0.0648*** & 0.0173** & 0.053* & 0.0581*** & 0.0069 & 0.0286 \\
Has consumed wine $i$ last year & (0.0071 ) & (0.0094) & (0.007) & (0.0098) & (0.0077) & (0.0308) & (0.0141) & (0.0167) & (0.0178) \\
& & & & & & & & & \\
Recognition : & 0.0336*** & 0.0227 & 0.0278*** & 0.0719*** & 0.0056 & 0.0303 & 0.0518** & 0.0116 & 0.0585** \\
Respondent knows wine $i$ & (0.0078) & (0.0174) & (0.0078) & (0.0129) & (0.0102) & (0.0516) & (0.0204) & (0.0277) & (0.0283) \\
Instruments : & & & & & & & & & \\
Alsace & Yes & Yes & No & No & Yes & Yes & Yes & Yes & Yes \\
Beaujolais & Yes & Yes & No & Yes & Yes & Yes & Yes & Yes & Yes \\
Bourgogne & Yes & Yes & Yes & Yes & Yes & Yes & Yes & Yes & Yes \\
C. du Rhône & Yes & Yes & Yes & Yes & No & Yes & Yes & Yes & Yes \\
Languedoc & No & Yes & Yes & Yes & Yes & Yes & Yes & Yes & Yes \\
Loire & Yes & No & Yes & No & Yes & Yes & Yes & Yes & Yes \\
& & & & & & & & & \\
Observations & 6,307 & 6,307 & 6,307 & 6,307 & 6,307 & 819 & 1,382 & 3,667 & 1,258 \\
Wald test of rho = 0 & 33.6415 & 2.97198 & 16.342 & 61.8815 & 2.3219 & 0.4397 & 7.0031 & 11.9136 & 3.0959 \\
P-value & (0.0000) & (0.0847) & (0.0001) & (0.0000) & (0.1276) & (0.5073) & (0.0081) & (0.0006) & (0.0785) \\
Wald Chi-Sq. & 2454.82 & 1393.72 & 1860.17 & 35057.09 & 34525.19 & 140.02 & 491.23 & 30687.25 & 1881.09 \\
P-value & (0.0000) & (0.0000) & (0.0000) & (0.0000) & (0.0000) & (0.0000) & (0.0000) & (0.0000) & (0.0000) \\ \hline
\end{tabular}
{\small{}Note: See table 2. Coefficients correspond to marginal impacts}
\end{sidewaystable}
\newpage
\begin{sidewaystable}
\protect\caption{Main determinants of the individual reputation level (ML Probit)\label{tab:table5}}
\begin{tabular}{ l l l l l l l l l l }
\hline
& Saint & Bordeaux & Médoc & Margaux & Entre-deux & Côtes de & Sauternes & Prem. Côtes & Graves \\
& Emilion & Supérieur & & & -mers & Bourg & & de Bordeaux & \\ \hline
& & & & & & & & & \\ \hline
Opinion on Bordeaux Wines & 0.0734*** & 0.1467*** & 0.0697*** & 0.0668*** & 0.0156** & 0.0092 & 0.0713*** & 0.0768*** & 0.1339*** \\
(Good = 1 ; Bad = 0) & (0.0114) & (0.0127) & (0.0111) & (0.0105) & (0.0075) & (0.023) & (0.0271) & (0.0147) & (0.0325) \\
& & & & & & & & & \\
Past consumption : & 0.1668*** & -0.011 & 0 .1194 & 0.1670*** & 0.0389** & 0.0548 & 0.1425*** & 0.0224 & 0.0782* \\
Has consumed wine $i$ last year & (0.0162) & (0.0194 ) & (0.0154) & (0.0243) & (0.0167 ) & (0.0349) & (0.0312) & (0.0406) & (0.0454) \\
& & & & & & & & & \\
Recognition & 0.0817*** & 0.0451 & 0.0616*** & 0.1807*** & 0.0126 & 0.0321 & 0.1212** & 0.0265 & 0.1526** \\
Respondent knows wine $i$ & (0.0188) & (0.0362) & (0.0174) & (0.0327) & (0.022) & (0.0559) & (0.0471) & (0.0659) & (0.0700) \\ \hline
Observations & 6,307 & 6,307 & 6,307 & 6,307 & 6,307 & 819 & 1,382 & 3,667 & 1,258 \\
Wald Chi-Sq. & 962.47 & 563.30 & 737.73 & 807.04 & 203.71 & 29.82 & 146.68 & 213.44 & 66.63 \\
P-value & (0.0000) & (0.0000) & (0.0000) & (0.0000) & (0.0000) & (0.0541) & (0.0000) & (0.0000) & (0.0000) \\ \hline
\end{tabular}
{\small{}Note: See table 4.}
\end{sidewaystable}
\newpage
\begin{sidewaystable}
\protect\caption{Calculating a measure of the average net benefits of promotion by subregion\label{tab:table6}}
\includegraphics{table6new.png}
%{\small{}Note: the fees included in the table were computed based on an average for the years 2011-2014}
\end{sidewaystable}
\end{center}
\newpage
\begin{figure}
\centering
\protect\caption{Wine Appellations, France}
\includegraphics{fig1.png}
\end{figure}
\newpage
\begin{figure}
\centering
\protect\caption{Delta coefficients (2SLS)}
\includegraphics{fig2.png}
\end{figure}
\newpage
\begin{figure}
\centering
\protect\caption{Delta coefficients (RBP vs. ML Probit)}
\includegraphics[width=6in]{fig3.png}
\end{figure}
\newpage
\begin{figure}
\centering
\protect\caption{Marginal impacts (RBP))}
\includegraphics[width=6in]{fig4.png}
\end{figure}
\end{document}
