Tim James: Is there comparative value in wine?

After some recent devotion to what one commenter rather nicely called “musing”, I thought it time to turn to facts. Though I’ve always found that facts are no less contentious than musings, as they tend to need contextualisation and interpretation before they are usefully digestible. But this is always the time of year to look at some facts (raw statistics, at least) about the South African wine industry, because it is in chilly mid-winter – May or June ­– that Sawis (South African Wine Industry Information and Systems) brings out its compendium of a selection of the stats that have  been gathered over the previous year.

I was pushed in its direction by a blog from “The Wine Economist”, Mike Veseth, an American academic economist who writes widely and interestingly about wine. This recent article looked at the idea of wine value (compared with that of other alcoholic drinks), and wondered if hard-pressed consumers were looking not so much (or not only) at lower prices but also at better value.

He argues that it’s reasonable for consumers to come to this conclusion, as all the studies he’s seen suggest that “wine is more expensive per serving than either beer or spirits using average price data”. Further (and I presume he’s talking specifically about the American market and I don’t know whether it applies here), Veseth says that “indeed. consumers have for several years moved away from inexpensive wines, calculating perhaps that they are not worth even the low price charged. The premiumization trend has plateaued, too, suggesting that perhaps higher price doesn’t always mean better value.”

I remember being told by someone seemingly knowledgeable that a large segment of extremely price-conscious South African drinkers have no loyalty to any particular drink, but shop around for whatever form of alcohol (spirits, wine, cider, illegal shebeen concoctions) offers the biggest bang for their buck. I don’t know if this segment is large enough to be statistically meaningful, but I went to the relevant pages in the Sawis 2023 statistics publication (No. 48), to see if could enlighten me with a broad-brush picture.

Firstly, I consulted a table showing total liquor consumption, in litres, for the years 2002–2023, broken down into categories: total wine (there are sub-categories too, but not of interest here), spirits, RTDs (ready-to-drink blends), and beer. No cider, which surprised me; it seems to be included with RTDs. There are, in fact, numerous annual discrepancies, arguably sugesting that perhaps there are shortish-term category-jumping drinkers, but mostly there are pretty steady trends, Covid years excepted.

Total wine consumption has risen by about 20% in the two decades reflected in the table and, unsurprisingly given population increase, all the categories have grown, with RTDs the biggest grower – to the point that there is now a substantially greater volume of them than of wine drunk: over 629 million litres compared with nearly 464 million litres of wine, and 143 million litres of spirits. And vastly, vastly more beer, of course.

Perhaps more useful to understanding changing patterns of drinking is a table showing per-adult consumption over those years. The categories mentioned above are used, except that “spirits” is broken down into brandy, whisky and “other”. As well as following basic trends, the figures for each category do still go up and down over the years. Wine is overall down a bit, though it’s pretty stable. Beer is up, though not hugely and it’s also reasonably stable overall. RTDs are up hugely – by some 60%. Whisky’s proportion has stayed much the same (though it had a quite big bulge from about 2006–2016 for whatever reason), but consumption of “other spirits” has increased significantly.

The big loser is brandy, coming down relentlessly from 0.6 litres consumed per adult in 2001 to just 0.26 litres today (despite my own personal best efforts). The reasons for this well-known trend have long been debated, but in the context of Veseth’s argument it strikes me as anomalous. At all quality levels, I think South African brandy must surely offer the best value, at least the best quality:price ratio of at least the spirits. There is now as much whisky drunk per adult as brandy, and three times as much “other spirits” (cane, vodka, gin, etc); something very irrational there!

Back to another strand of Mike Veseth’s exploration of wine value. He suggests that another possible value problem with wine is the way it’s packaged:

“Many consumers hesitate to open a 750 ml wine bottle for only one or two glasses because they are afraid that what’s left will quickly go bad, making the bottle purchase an even worse deal than the per-serving averages suggest. (By comparison, beer comes in single-serving containers and spirits can keep for a long time, so they don’t suffer the same wasted money problem.)”

My own experience with drinkers that aren’t all that fussy is that they don’t think wine “will quickly go bad” and are happy to leave a bottle for a week or even two with no compunction. But still, point taken. Veseth suggests, supported by a research report, that bag-in-box wine is doing well compared to bottled wine at a similar price point. Do Sawis statistics support this?

Indeed they do. There’s a fascinating table of “Particulars of packaged wine sold on the domestic market”, itemising volumes of still wine in different containers, this time with figures going back to 2017. Firstly, to look at the container which is undoubtedly the one best known to Winemag readers – the 750 ml glass bottle. Over 112 million litres of wine were sold in South Africa in these “standard” bottles in 2017, but just 91.5 million in 2023.

In this survey we can pretty much ignore most of the other containers, including other glass-bottle formats, though it’s surprising how much is sold in  bottles of 1.5–2 litres (more than 38 million litres – would you have guessed that, fellow elitist, given that we’re not talking about the magnums offered of some ultra-prestigious wines to allow for superior long-term ageing?). It’s interesting to note that wine in cans is finally entering the statistics in a meaningful though still small way: 11.5 million litres in 2023.

By far the largest volume of wine sold in South Africa is “bag-in box”: nearly three-quarters of it in five-litre bags and most of the rest two-litre packs. Over 207 million litres in total in 2023, compared with just 148 million in 2017, the numbers climbing steadily each year, not even much affected by the disastrous Covid-inspired prohibition years. Nearly 50% of wine in South Africa is sold in this way, up from 37% in 2017. A visit to smart Woolworths will tell you emphatically that this is not just lowest-end stuff: much of their bag-in-box offering is generic, but the range has included brand names like Alto, Beyerskloof, Diemersdal, La Motte, Spier and Kleine Zalze.

It makes sense. A point not made by Mike Veseth is that producers can afford to put better wines in bag-in-box, increasing the potential value for money to be found there. There’s a fairly significant percentage difference in packaging costs between a cheap chenin in a lightweight bottle with cheap label and cork or screwcap and the grandest cabernet, but as a proportion of the total price it’s fairly negligible. But a glass bottle is an expensive way of packaging wine. Putting the stuff in a bag-in-box is vastly cheaper, and we can hope that some of the saving goes to the wine-lover, including in getting better quality wine for the price. And, except when it comes to (not) buying brandy, the consumer is often pretty canny.

It’s also worth noting another, clearly related, trend in the Sawis statistics on packaging for wine: wine in tetra packs (mostly one-litre) has gone up from 3.6% in 2017 to 5.9% in 2023. That’s an even bigger percentage increase than for bag-in-box.

It might not yet be relevant to the readers of Winemag, but I’d bet something substantial that next year’s Sawis statistical report will record further movement in this direction away from glass.

• Tim James is one of South Africa’s leading wine commentators, contributing to various local and international wine publications. His book Wines of South Africa – Tradition and Revolution appeared in 2013.