Can a wine be feminine? Gendered wine descriptors and quality, price, and aging potential

Abstract

By analyzing more than 1,400 expert tasting notes, we assess the so-called gender profile of Bordeaux wines. We identify 329 gender-related wine descriptors, with a good balance between masculine and feminine descriptors. Some wines and vintages are described as more feminine than others, but no clear trend over time emerges. Our regression analysis further reveals that more feminine wines receive similar ratings and sell at similar prices as their more masculine counterparts, but they are perceived as having a much more limited aging potential.

“This sexy, lush, complex, perfumed Pomerol […] is fleshy, silky, and voluptuous in its elegant, feminine style.”

—Robert Parker Jr., about Hosanna 1999

I. Introduction

Gender identity is defined as the result of a social construction (Lorber, 2018) that influences both attitudes and behaviors (Zayer and Pounders, 2022). Gender differences are well documented in wine information searches (Barber, Dodd, and Kolyesnikova, 2009), visitor perceptions of a winery’s wine and service (Mitchell and Hall, 2001), and wine drinking patterns (Forbes, 2012). More recently, gender differences have been highlighted on the supply side of the wine industry, with differences in terms of careers (Bryant and Garnham, 2014), management style and leadership (Galbreath and Tisch, 2020), and performance (D’Amato, 2017). Most products, services, and brands also have a gender (Grohmann, 2009), which can be constructed when masculine or feminine cues (visual or semantic) are displayed (Spielmann, Dobscha, and Lowrey, 2021).

Gender dimensions of brands include symbolic aspects that influence a consumer response in the marketplace. Traditionally, elements considered masculine are associated with competence, while elements considered feminine are associated with warmth (Hess and Melnyk, 2016; Pogacar, et al., 2021). Schnurr (2018) shows that when pursuing hedonic (utilitarian) consumption goals, products perceived as feminine (masculine) are generally preferred over those perceived as masculine (feminine), with no differences between consumer genders. Similarly, Pogacar et al. (2021) show that feminine brands (the masculinity/femininity of a brand can be inferred by linguistic cues) are perceived as warmer, which gives them an advantage over masculine brands. This advantage is all the greater when the products associated with these brands are hedonic (vs. utilitarian).

In this article, we focus on wine, which is often associated with a particular gender. Fugate and Phillips (2010, p. 256) note that “wine, originally labelled ‘masculine’, was categorized as ‘feminine’ in this research, indicating a 180-degree shift in the gender identity of this product,” largely attributed to popular media and entertainment. Certain grape varieties and winemaking techniques produce particularly fine, delicate, and elegant wines, frequently described as feminine. Conversely, adjectives such as powerful, tannic, bold, and strong are associated with masculinity. Indeed, anthropomorphic metaphors associating wine with a person are often used in wine discourse (Normand-Marconnet and Jones, 2020). Peynaud (1980) proposed a model of wine evaluation based on three axes, including a “masculine/feminine” axis. As summarized by Lehrer and Lehrer (2008, p. 114), ‘‘wines are described as masculine or feminine, muscular or sinewy, for example, in addition to being described as heavy or light, delicate or harsh.’’

Livat and Jaffré (2022) identify two main areas of gender-related research in the wine context. One is the issue of gender and its impact on wine choice, consumption, purchase, and appreciation. The other area focuses on women and their influence in the wine value chain. In this article, we look at a third area, the gender of wine as such. It should be noted that the gender of wine, as conveyed by the wine vocabulary, is a social construct, as is the expression of terroir (Castello, 2021). We are interested in the gendered description—or gender profile—of a wine and its implications for quality and price. More specifically, by analyzing expert tasting notes, we aim to assess:

1. (i) The gender profile of wines and its evolution over time. Women are becoming increasingly important on both the supply and demand sides of the market. Therefore, one can expect wine to become more feminine.

2. (ii) The relationship between a wine’s gender profile and its score. As feminine linguistic attributes for gendering products improve consumer attitude (Pogacar et al., 2021), the gender profile of the wine can also have an impact on the wine’s rating.

3. (iii) The relationship between a wine’s gender profile and its drinking window. More masculine wines tend to have a stronger structure (more tannins and acidity) and should therefore have a longer aging potential.

4. (iv) The relationship between a wine’s gender profile and its price. Scores and aging potential correlate with wine prices (see, e.g., Ashenfelter, Ashmore, and Lalonde, 1995). Thus, if propositions (ii) and (iii) are verified, a wine’s gender profile should also correlate with its price.

Our results show that some wines are indeed described as more feminine than others, but their prevalence has not increased over time. Moreover, feminine-related wine descriptors do not significantly correlate with scores and prices, but they can be associated with wines benefiting from a shorter drinking window.

II. Background

a. Gendered wine preferences and gendered wine vocabulary

Gender differences in wine consumption behavior and attitudes have been observed across studies. Some have found that the ability to discriminate odors and tastes in wine is more acute in women than in men (Doty et al., 1985). While women are often the primary wine purchasers, they are more concerned about making the “wrong” wine decision than men (Barber, Almanza, and Donovan, 2006), seeking information from knowledgeable professionals, and leaving higher-priced or special occasion purchases to their male counterparts (Atkin, Nowak, and Garcia, 2007).

Wine preferences may differ by gender. Some researchers report a stronger preference for young and sweet wines among women, especially younger women, than among men, although both women and men prefer dry wines (Bruwer, Saliba, and Miller, 2011). Similarly, while men have a strong preference for red wines, women appreciate both red and white wines. They prefer wines with vegetal characteristics, fruit aromas, and mouthfeel, while men prefer older wines (Bruwer, Saliba, and Miller, 2011). Women also look for wines with vanilla, floral, and spicy aromas. They are less interested in earthy flavors than men. Women are also more likely to enjoy complex wines, preferring wines with low tannins and more subtle acidity (Fuhrman, 2001).

In recent years, the wine industry has seen a significant increase in the number of women occupying various positions, particularly winemakers and sommeliers (Gilbert and Gilbert, 2012; Almila, 2019). The growing importance of women in the industry may influence the production and marketing of wines, which could lead to an increase in the availability of wines with feminine characteristics. Certain aromas are commonly associated with femininity, and gender metaphors are common in wine vocabulary (Negro, 2012). For example, the presence of floral notes such as rose or violet, or adjectives such as silky or voluptuous, could contribute to the perception of wine as feminine. On this basis, we can expect wines to become increasingly feminine and make the following proposition: the use of feminine wine descriptors increases over time (Proposition 1).

b. Feminine wines, critics’ scores, and aging potential

Gilbert and Gilbert (2012) show that female winemakers are more highly regarded by experts than male winemakers in proportion to their presence in the field, particularly in California. In such a context, we can examine if wine talk or wine discourse, defined as the language employed by critics, has also developed. Wine talk is distinct from the “wine language” employed by academics in the industry (Inglis, 2020). Indeed, descriptors with gradable and evaluative concepts (Lehrer, 1975, 2009), metaphors, and personification are used liberally to convey emotions associated with wine consumption (Paradis and Eeg-Olofsson, 2013). As Vanini et al. (2010, p. 391) explain:

“Metaphors provide a colorful, convenient, and widely accepted tool to elaborate on sensuous qualities of wine and experiences of taste.”

One of these metaphors is the description of wines as either “feminine” or “masculine” (Lehrer, 1975, 1978, 2009; Negro, 2012). If wine itself is a gender-neutral commodity, the semantic relations of certain words used together can suggest a particular gender. As described by Lehrer (2009, p. 32):

“Feminine is associated with soft, smooth, light, round, perfumed, possibly sweet, and these words do have definite meanings in the wine domain. Hence, a feminine wine will be understood as one having those properties. A masculine wine is big and perhaps rough.”

Previous research has shown that male and female judges assign the same scores to the same wines (Bodington, 2017; Bodington and Malfeito-Ferreira, 2018), but the relationship between wine quality and its gendered description is uncertain. Some authors have argued that gendered wine descriptors are used to transfer societal gender distinctions to the wine sector, with lighter “feminine” wines rated as inferior and less interesting than larger “masculine” wines (Matasar, 2006; Inglis, 2020). Pogacar et al. (2021) assert that feminine names are positively associated with warmth and could provide a branding advantage and increased choice. Femininity is associated with hedonic products that create sensory, experiential, and pleasurable benefits (Schnurr, 2018). Given that taste-related words cannot be independent of a hedonistic consideration and that experts recognize the hedonistic value of wine (Brochet and Dubourdieu, 2001), we make the following propositions:

• - The more feminine the wines, the higher their tasting score (Proposition 2);

• - The more masculine the wines, the higher their aging potential (Proposition 3).

c. The price of feminine wines

Assessing wine quality is as much a sensory as a cognitive process for experts, where numerous contextual and environmental factors can influence the perception and evaluation of wine (Charters and Pettigrew, 2007; Spence, 2020). Research suggests that expert evaluations contain little private information. Most of the price of fine wine can be explained by reference to public information, such as weather data (Ashenfelter, Ashmore, and Lalonde, 1995; Ashenfelter and Jones, 2013).

However, for some influential experts, their assessment of wine ends up with a score that can have a direct impact on prices (Ali, Lecocq, and Visser, 2008). Previous research has investigated how wine reviews predict quality (Yang et al., 2022) and how tasting notes relate to price. Tasting note length is positively associated with price (Ramirez, 2010). Although previous research has unequivocally dismissed wine descriptors as “bullshit” (Quandt, 2007), the content of the tasting notes is also important. Some of these are valued by consumers (Capehart, 2021a), with different themes depending on wine region and varietal (McCannon, 2020). Capehart (2021b) shows that some “expensive” and “cheap” words are more likely to be used for high- and low-priced wines, respectively; for example, specific or elite words such as truffle, elegant, and vintage are associated with expensive wines, whereas general or accessible words such as tasty, pleasing, or harvest are associated with cheap wines. Thus, and based on the positive outcomes of feminine elements in hedonic consumption, gendered descriptors may also be of some value to consumers, and we make the following proposition: the more feminine the wine description, the higher the price (Proposition 4).

III. Method and data

To answer our questions, we used an extensive database of Robert Parker Jr.’s tasting notes and ratings on vintages from 1994 to 2013, for a total of 1,404 observations. We use Bordeaux data to make the analyses perfectly comparable and only consider en primeur ratings and prices. Focusing on a single expert and a single wine region may seem restrictive. However, it is worth noting that (i) Robert Parker Jr. is the best known and most influential taster in the world (Masset, Weisskopf, and Cossutta, 2015), (ii) he is known to have stable preferences, (iii) his scores are accompanied by tasting notes and a suggested drinking window, and (iv) Bordeaux wines represent more than 50% of the global fine wine market (see Liv-ex.com).

To determine the level of femininity of the wines in our sample, we created an indicator based on the gender of the descriptors used in the tasting notes. We extracted all individual adjectives and words. This gave us 1,183 wine descriptors in our sample, including 194 references to wine aromas (e.g., blackcurrant, floral). Each of these descriptors was then evaluated by three experts (specializing in linguistics, communication and culture, and psychology, respectively) to determine the gender with which it could traditionally be associated and classified as either masculine, neutral, or feminine. In case of disagreement, the gender was discussed within the group. If no agreement was reached, the descriptor was considered gender-neutral and not gender-related (e.g., questionable, insipid). We identified 329 gender-related descriptors (GRDs), well-balanced between masculine (e.g., muscular, solid) and feminine (e.g., elegant, soft) ones (164 and 165, respectively). The list can be found in Appendix 1. The most frequently encountered gender-related adjectives are the following: sweet (present in 45% of the tasting notes), fine (26%), pure (18%), elegant (17%), and light (16%) for female-related adjectives; rich (31%), acidic (28%), dense (25%), impressive (16%), and powerful (13%) for male-related adjectives.¹

On this basis, we calculated a ratio that measures the proportion of female-related descriptors (FRD) out of all the GRD used in each tasting note: the higher the score, the greater the degree of feminine descriptors. Table 1 provides descriptive statistics on both the independent (gender-related adjectives, upper panel) and the dependent (rating, aging potential, and price, lower panel) variables. Overall, the number of adjectives associated with the feminine and masculine genders is very close. The proportion of feminine terms is slightly higher, as some wines have a particularly large number of descriptors associated with this gender. The average rating is close to 92 points, with a standard deviation of 3.55. Some wines achieve near-perfect scores (99.5 points), while the worst performers score below 80 points. Aging potential is highly variable, and some wines were considered ready to drink when the tasting note was written (the minimum age being equal to 0). Prices are quite variable, reflecting significant château and vintage effects. Indeed, some châteaux and vintages sell for much higher prices. It will be important to take this into account in the empirical analysis by including fixed effects in the regression models.

Table 1. Descriptive statistics

Château Trotanoy 1994 is the wine with the least feminine descriptors (or the lowest FRD score), with 29% FRD in the tasting note, while Domaine de Chevalier 2000 is the wine described as the most feminine one, with 70% FRD. Overall, the average proportion of feminine descriptors within a tasting note is 30%, with a minimum of 16% for vintage 2007 and a maximum of 38% for vintage 2000 (see Figure 1). The trend over the vintages is slightly decreasing, which contradicts Proposition 1.

Figure 1. Percentage of feminine descriptors by vintage.

IV. Empirical analysis

We run a series of regressions to explain (1) Robert Parker’s rating, (2) aging potential measured at the beginning and end of the drinking window, and (3) en primeur release price. For each of these three models, we estimate a series of regressions with château and vintage fixed effects (Model I), the percentage of female-related descriptors (FRD) (Model II), the number of gender-related descriptors (GRD) (Model III), and the length of the tasting note (TN) (Model IV) as explanatory variables. In addition, the age and price equations include Parker’s rating (Model V). Aging and price equations are estimated using a double-log specification. Estimation results are presented in Tables 2, 3, and 4.

Table 2. Rating equation—estimation results

Note:

* , **, and *** denote significance at the 90%, 95%, and 99% levels, respectively.

Table 3. Age equations—estimation results

Note:

* , **, and *** denote significance at the 90%, 95%, and 99% levels, respectively.

Table 4. Price equation—estimation results

Note:

* , **, and *** denote significance at the 90%, 95%, and 99% levels, respectively.

FRD does not play a role in the evaluation (Table 2), which contradicts Proposition 2. The total number of gender-related descriptors appears to be significant (III), but this effect vanishes when controlling for the length of the tasting note (IV).

In Table 3, FRD is highly significant in each specification. It seems that wines described as feminine enter their drinking window earlier than wines described as masculine, as the coefficient associated with FRD is negative when considering age at the beginning of the drinking window. This means that wines described in a feminine way can be enjoyed much younger than those described in a masculine way. These wines also tend to have a more limited aging potential, as they are also associated with a negative coefficient when considering age at the end of the drinking window. These results are in line with Proposition 3. This result holds even when we control for the relationship between score and aging potential.

In Table 4, we obtain very high R-squared values thanks to the château and vintage fixed effects. Previous empirical literature suggests that reputation and vintage are the most important determinants of Bordeaux wine prices (see, e.g., Masset and Weisskopf, 2022). Ratings appear to play a limited but significant role. Neither the FRD nor the number of GRD are significant, suggesting that wines described as feminine sell for essentially the same price as wines described as being more “masculine.” Proposition 4 is not supported.

V. Concluding remarks

Given the increasing gender diversity in the wine industry and the extensive use of gendered metaphors and descriptors in tasting notes, we analyze the ways in which wine descriptions are gendered and how this relates to quality and price. Interestingly, despite the growing role of women in the wine industry, the description of wine has not become more feminine. Moreover, the gendered description of a wine appears to impact neither its ratings nor its price. However, wines described as more feminine both enter their drinking window and start to decline at a younger age than those described as more masculine. This finding is consistent with the premise that the attributes described are more commonly associated with hedonism, such as quick consumption. In contrast, masculine descriptors are more readily associated with the wine’s aging potential (Schnurr, 2018). So, while a feminine description of a wine may not significantly impact its price or rating, it may nevertheless have repercussions on consumer behavior.

Our results are based on data from only one expert, who is known to prefer bold wines. In addition, our dataset covers a period when consumers also tended to prefer bold wines. Further research could examine the tasting notes of multiple judges and analyze whether wine descriptors vary according to their gender. The valence of gender-related descriptors could also be considered in the analysis.