Monday, February 28, 2022

Global wine consumption has been declining for a long time

Given current concerns about possibly declining wine consumption (e.g. Declining wine consumption a global trend; Global wine consumption declines), it is worthwhile to look at long-term trends, to put the current data into context. We could, for example, look at particular countries (see Worldwide patterns of wine consumption); but in this post I will look at global tends, which are thus averaged (or summed) across all countries.

When working out national consumption of any consumable product, it is of interest to work out the amount per person, rather than just the total amount for each country. The calculation is a simple one (national consumption / population size), but a possible complication for alcoholic beverages is that not all persons are legally permitted to consume alcohol.


An obvious solution is to divide by the population of adults, rather than the total population size. Even this is only an approximation, though, as we then need to define adult, specifically in terms of the legal age for drinking alcohol — adult is often defined as 15 years old, whereas alcohol consumption typically involves an older age. More to the point, we do not really have good past data for adult population size for most countries. So, calculations per capita are the data most commonly available.

For wine consumption, we have data going back to 1835, available in the Annual Database of Global Wine Markets, compiled by Kym Anderson and Vicente Pinilla. In the first graph, I have plotted the per capita beverage wine consumption (in liters of alcohol), starting way back in 1862. Even a quick glance suggests that wine consumption per capita has been declining since its peak a century ago.

Wine consumption per capita since 1860

There is clearly a lot of variation in the data prior to World War II, presumably due to approximate data estimates for many countries. However, the general trend up until the War is reasonably stable, varying around an average of c. one liter of alcohol per person per year.

Equally clearly, there was a decline during the War (no surprise there), followed by an increase for the first decade after the War. From there, the average remained at c. 0.85 liters of alcohol per person, for the next two decades.

However, since the mid 1970s there has been a continual decline in consumption per capita. There was a steady decline for the first decade (in pink on the graph), followed by a precipitous decline during the next decade, and a steady decline again since 1990 (also in pink on the graph). The graphed data are current to 2018. The decline from 2013 to 2018 was c. 0.005 (liters of alcohol per person) per year— this is about 5% of a bottle per year.

So, nothing seems to be particularly unusual just recently. The initial decline beginning in the 1970s was associated with the worldwide financial downturn (1973–1974 stock market crash) plus the oil embargo (1973 oil crisis). Subsequently, there was a general increase in wine prices that has gone way beyond any possible increase due solely to inflation (The outrageous prices of modern high-quality wines). For my own personal take on prices back then, see: Exponential increase in wine prices (1980-2000).

The Global Wine Markets database also contains data for wine consumption per adult (rather than per capita) starting from 1950. We can thus compare this to the per capita consumption, which reveals an almost perfect correlation. So, the numbers change somewhat (obviously, average consumption is slightly higher per adult than per person), but the patterns through time are almost identical.

Alcohol consumption per adult since 1960

If we want to compare wine consumption to that of beer and spirits, then we can do this per adult from 1961 until 2015. The three sets of data are shown in the second graph.

As noted above, the pattern for wine is almost identical to that in the first graph, with a steady decline in consumption since 1990. Beer consumption, on the other hand, rose steadily until the mid 1990s, and has been relatively steady since then. Interestingly, it shows a decrease since 2012, just as occurred for wine.

Clearly, given that almost all beers contain less alcohol than the equivalent volume of wine, beer consumption far exceeds wine consumption, on average per adult. People get c. 2.5 times as much of their alcohol from beer compared to wine, and wine contains 23 times as much percent alcohol as does beer. So, per adult beer consumption is c. 7 times that of wine.

Consumption of spirits rose until the mid 1970s, then declined for a decade, and has remained relatively steady since then. The same as for wine and beer, it shows a decrease since 2012.

Given that a bottle of spirits contains much more alcohol than either wine or beer, it is not necessarily surprising that people get more of their alcohol that way, on average. So, it is the trend in spirit consumption that dominates total intake of alcohol, which has remained relatively steady since the mid 1980s, having reached its peak in the mid 1970s.

Conclusion

So, if the wine industry is going to concern itself about recently declining wine consumption, it first needs to take into account declining consumption of alcohol, in general, since 2012. Only after that, should the long-term decline from the mid 1980s be addressed, since this has not occurred for either beer or spirits.

10 comments:

  1. What happened dating back to 2012 that diminished alcohol consumption? For us here in the States, we hit bottom on The Great Recession.

    (Around that time, unemployment and underemployment peaked at 21% in the state of California. Higher -- much higher -- than the current coronavirus pandemic-caused recession.)

    Alcohol consumption ran headlong into declining personal incomes.

    Drinking became an elective, "discretionary income" purchase that could be foregone, in order to meet more pressing concerns in one's household budget (e.g., food, shelter, ground transportation, communications -- smartphone data plan, household Internet connection).

    ReplyDelete
  2. As the SVB report points out “While only 28 percent of the boomer population is nonwhite, 45 percent of the millennial population — and almost half of Gen Z — is nonwhite,” Increasing the U.S. population is made up of minorities from areas that are genetically predisposed to bitterness sensitivity (www.tastescience.com). If your genetically from Africa, most of Asia or Latin America you will tend to be three to four times as likely to be sensitive to bitterness. That means you are less likely to consume beverage alcohol and far more likely to consume sweeter wines than the typical wine consuming baby boomer. When you add in a student debt load of 1.6 trillion you have two headwinds pushing against the younger generations becoming wine consumers.

    ReplyDelete
    Replies
    1. Indeed, genetic factors play a large role in choice of beverage. Changing demographic makeup thus has a big effect.

      Delete
  3. The “framing effect” has a significant impact on current and potential wine consumers.
    The framing effect is the cognitive bias that occurs when a person is given a choice from a set of options. The choice is influenced more by how the information is presented (framed) than by the information itself. Danny Kahneman’s research on framing demonstrates that people are more influenced to avoid a loss than the prospect of an equivalent gain (Tversky & Kahneman, 1981).
    Research indicates we tend avoid options that are framed negatively. Some of the ongoing negative framing in the current dominant narrative that will inhibit adoption of wine include.
    If you consume sweet wines, you are immature. If you consume sweet wines, you are unsophisticated. If you do not order red wine when you order a steak (or other red meat) you are making a mistake and you will have a negative experience. You will learn to like big tannic red wines as you mature and become more sophisticated.
    We tend to choose options that have higher numbers, our bias tends to believe that higher is better (100-point scale, higher price equals higher quality).

    ReplyDelete
  4. There is always the possibility that there is less wine consumption because the stuff just isn't that tasty of a beverage?!?

    ReplyDelete
    Replies
    1. One can never exclude that possibility! However, I suspect that there's more to it than that. At any rate, a lot of today's stuff seems pretty good to me.

      Delete
  5. Genetics truly does play a pivotal role. Genes determine that no two people taste the same wine the same way — https://wineindustryinsight.com/?p=116368

    ReplyDelete
  6. You are talking about the whole world consumption. Sure Europe consumption has dropped but the gallons per person of wine in the the U.S. has increased https://wineinstitute.org/our-industry/statistics/us-wine-consumption/

    ReplyDelete
    Replies
    1. My post makes it clear that global consumption is the subject here. I also make clear that I have already written a post about individual countries.

      Delete
  7. The genetic propensity for bitterness sensitivity is an adaptation that improves a human’s chance of survival in areas where there are poisonous fruits, nuts, and vegetables that are not acutely bitter. Adults (30+) who prefer or only consume sweet wines are far less likely to consume poisons because poisons tend to be bitter. At some point the level of bitterness hits the “threshold of rejection” and your gag reflex kicks in. If you ask a sweet wine drinker what a scotch neat tastes like? They often respond with “gasoline.” I doubt it is possible to learn to like the taste of straight gasoline. The U.S. is approaching a point where half the population will be to some degree bitter sensitive, far more than the current boomer population.

    ReplyDelete