The aim of this paper is to analyze risk attitudes of winegrowers in France. An inter- esting feature of French viticulture is that most of the production is carried out under an appellation regime, bearing constraints in terms of maximal yield authorized. We estimate a translog cost function under the constraint of this maximal yield. We jointly estimate cost function parameters, factor share equations and winegrowers’ attitudes to the risk function. Our estimates are based on the FADN database (2005-2014), data from the National Cultural Practices survey (2006, 2010, 2013), data from the French National Institute of Origin and Quality (INAO) and data from guides on production costs in viticulture and œnology (from 2005 to 2014). We find that pesticide demand is inelastic and all types of winegrowers are risk-averse. For the majority of them, risk aversion declines with expected profit (DARA), but for the Champagne region, it is the contrary. Expected profit is far higher than in the other regions and these winegrowers are more averse to risk when expected profit increases (IARA).
The objective of this paper is threefold. First, we estimate the causal effect of wine medals on pro- ducers’ prices. Second, we calculate the expected profit obtained by producers from participating in competitions. Third, we investigate the efficiency of wine competitions by measuring to what extent the attributed awards are good quality indicators. Our dataset combines information on transactions be- tween wine producers and wholesale traders (these data are registered by a wine broker who acts as a middleman in this market), with the records from eleven important wine competitions. Our identifica- tion strategy exploits a particularity in our data, namely that medals are not only awarded before the transaction dates but sometimes also thereafter. Under weak restrictions, a regression of price on dum- mies indicating past and future medals (plus controls) allows to uncover two interesting features: i) the difference in the respective dummy estimates identifies the causal effect of a medal, ii) the estimate of the future medal coefficient identifies the correlation between unobserved quality and medal. We find a strong medal impact: our preferred estimate indicates that producers of medaled wines can increase their price by 13%. The impact for gold is much larger than for silver and bronze, but we cannot reject that the correlation with quality is the same across the three colors. Only a minority of contests attribute medals that are significantly correlated with quality (primarily the ones founded a long time ago, and whose judges are required to evaluate relatively few wines per day). Our profit calculations show that the incentives to participate in wine contests are high.