US states do not import wine from the same places

The USA is not a united wine market, if only because of the state-based Tree-Tier system of alcohol distribution, which gives the distributors in many states great power to determine which wines end up where (see Prohibition still exists in the USA). It would therefore be interesting to look at the data regarding which wines are preferentially imported into which states of the USA.

There do not seem to be complete data freely available regarding his question. However, a recent report from Wine Intelligence, on Opportunities in the US Market for Portuguese Wines (October 2019), does provide data for some of the most populous states: California, Texas, Florida, New York, Illinois, New Jersey, and Massachusetts. The first four of these states have the largest populations, accounting for 33% of the nation’s people between them, and the seven states together make up 42%. So, these data are worth looking at, even though they are incomplete.

The data come from the Vinitrac survey, and refer to the percentage of respondents who stated that they had consumed wine from each of 15 named regions in the previous 6 months. The data are shown in this 3-D bar chart, with the 15 wine-producing regions along the bottom, the 7 import states along the right, and the respondent percentage vertically.

The patterns are very similar across the seven states, but there are notable differences. The rank-order of the wine regions does differ between the US states, although Californian wine is most preferred in all of the states, with Italy ranking second, and Slovenian wine ranking last (of these 7 regions only).

The most obviously different state is California, where the people clearly have a greater preference for Californian wine (78% of respondents) than do the other people (c. 66% of respondents). Perhaps this does not surprise you? This extra consumption is counter-balanced by less consumption of wine from the other US states, along with Italy and France.

French wine is preferred to non-Californian US wine in California, Florida, New Jersey and Massachusetts, but it is the other way around in Texas, New York and Illinois. Spanish wine ranks next, everywhere, followed by Australian wine.

Chilean wine is more popular than Argentinian wine in all of the states except Massachusetts. German wine does slightly better than New Zealand wine, except in California and Massachusetts. Portuguese wine does better in New Jersey and Massachusetts than it does elsewhere (twice as well as in Texas and Illinois).

The wines of South Africa are not as popular in Texas as they are elsewhere. Canadian wines are surprisingly popular in New York, while Greek wines do best in New York and Illinois.

It is interesting to note the presence of Slovenia in the survey. It has recently been noted that Slovenian wine is fast gaining recognition in the USA, although it clearly has a long way to go to challenge any of the other wine-producing regions in popularity.

Mind you, this whole situation is currently under threat from Mr Trump, who wants to remove Italy, France, Spain, Germany, Portugal, Greece and Slovenia from these lists (see Alder Yarrow's compendium: The disastrous tariff edition). This seems very much like: Cutting off your nose to spite your face.

What do wineries, corporations and Benedictine abbeys have in common?

This title sounds like click-bait, but it is not, as you will discover below — the question is real, as is its answer.

A decade ago, the Journal of Management History published an interesting paper by Katja Rost, Emil Inauen and Margit Osterloh, entitled “The corporate governance of Benedictine abbeys: What can stock corporations learn from monasteries?” (2010 16: 90-115). This unlikely topic was explained thus:

Objective: to analyse the governance structure of monasteries, to gain new insights, and apply them to solve agency problems of modern corporations ... The paper uses a dataset of all Benedictine abbeys that ever existed in Bavaria, Baden-Württemberg, and German-speaking Switzerland, to determine their lifespan and the reasons for closures. The governance mechanisms are analyzed in detail ... [Results:] The monasteries that are examined show an average lifetime of almost 500 years, and only a quarter of them dissolved as a result of agency problems. This paper argues that this success is due to an appropriate governance structure that relies strongly on internal control mechanisms, [based on] the tendency to promote groupthink, the danger of dictatorship, and the life-long commitment ... Benedictine monasteries and stock corporations differ fundamentally regarding their goals.

You can read the paper if you want more details, but the three points that the authors highlight do seem to be fundamental differences between corporations and monasteries. My point in this blog post is to note that the same differences can exist among wineries, some of which survive and some of which end up being “dissolved” in various ways (frequently by a corporate predator).

The essential point, I feel, is the one about life-long commitment. This is a feature of many religious groups, of course, who therefore have had longer-term effects than have most other groups (eg. Centuries later, Jesuit missions in South America are still strengthening communities). Wineries cannot claim the same social value as can educators, but a number of their creators have tried to generate similar longevity.

I have discussed before that Keeping the family wine business is often hard. I noted that, throughout much of the world, vineyards and their associated wineries are often family-run concerns, either small or large, but that most family wineries outside of Europe are still in the hand of first-generation owners. I used some data from Australia to explore this point in more detail.

Sadly, things have gotten worse in Australia, with the recent news about financial problems at McWilliam’s, which has been family owned for 6 generations (143-year-old McWilliam’s Wines falls into administration). McWilliam’s was one of the founders of Australia’s First Families of Wine, a group of 11 of the oldest “family owned multi-generational” wineries. The problems seem not to have been short-term (see McWilliams Wines in receivership).

Succession problems are not a feature only of wineries, of course. In a discussion of Porto’s Bakery, in Los Angeles (How long lines keep Porto’s Bakery affordable — and growing), we are told that: “Only 30% of family-run companies hand off a business from the first generation to the second. Only 12% make it to third generation, and less than 5% make it to the fourth generation.” Indeed, not many winemaking families still exist in America that have been serially and continually involved in the wine industry for more than three generations (see 3 years of hurt on horizon, but despite challenges wine remains relevant).

Family businesses are reported to account for 64% of U.S. gross domestic product, generate 62% of the country’s employment, and account for 78% of all new job creation (see America’s economic engine). Moreover, family firms apparently out-perform non-family firms, economically. On the other hand, family-firm CEOs seem to earn on average nearly 10% less than their non-family counterparts (see Family Business Alliance).

As Nick Butler recently noted:

We’re all quite familiar with stories of the rise of family empires, with one hard-working, founding generation handing over a burgeoning business to a second, equally hungry and focused one ... Sadly, we’re also quite used to the disintegration of those powerhouses when an entitled third or fourth generation gets their hands on the steering wheel. There is a genuine art to succession and longevity, in any business.

One basic issue, I suppose, is often that children see how hard their parents work, and few of them then choose to work just as hard, in the same business. Agriculture, in particular, has undergone a personnel drain since the beginning of the Industrial Revolution, with more and more people moving off the land and into a city. Indeed, the European Commission’s EU Agricultural Outlook report recently commented on: “the increasing abandonment of small vineyards due to ageing farm owners and/or difficulties to compete on the market.”

Family business succession can, of course, be planned. In the recent reports about the death of Beaujolais wine-producer Georges Duboeuf, it was noted that he retired only last year (at age 85). This certainly represents a life-long commitment, but it is not a way to ensure a family succession. The idea is to generate life-long commitment in the next generation, not keep them at bay for decades. A similar idea applies to non-family employees, of course (see the example discussed in The unusual success of this family owned Rioja winery).

Wealth-planning companies suggest the following (Family business transitions: rising to the challenge):

1. Develop a plan and process to prepare the next generation of leadership, as this can improve the odds of a successful transition
2. Learn specific steps you can take to overcome two of the most common roadblocks:

a. The next generation is not interested or prepared to lead the business
b. Leadership has not planned or prepared for the leadership transition.

Monasteries do not have to concern themselves about direct succession, of course, but only about an on-going source of novices, each of whom will make a life-long commitment. They still exist in the modern world, although there are many fewer than there once were. Monarchies have also usually been quite good at organizing successions, in a very strict manner — although the unusual longevity of the current British (and Australian) monarch seems to be creating succession problems. There are many fewer monarchs these days, as well.

On the other hand, there are many more wineries now than ever before — but they are not nearly so good at succession, as recently pointed out by Betsy Andrews (How vintners in Sonoma County are tackling succession planning). So, maybe there is a better lesson to be learned from monasteries, rather than from corporations.

Quality scores changed the wine industry, and created confusion

Towards the end of last year there was some more of the ongoing discussion about the pros and cons of wine-quality scores, especially when used as a marketing tool (involving Simon Solis-Cohen, Tom Wark, and Roger Morris).

It seems to me that there was something missing from the discussion, which I wish to highlight here. The fundamental problem with quality scores is that they are an opinion but they get treated as a piece of mathematics. Until this issue is resolved (which it may never be), confusion will continue (along with the discussion), which cannot be good for the wine industry.

If wine evaluation stayed as an opinion (as Tom Wark correctly puts it: A wine rating is an adjective, not a calculation) then, indeed, things would be alright — evaluations would function as a marketing tool, just like any other adjectival opinion. But they don’t stay that way — they implicitly and explicitly have mathematical operations performed on them, such as averaging, which never happen to real adjectives. There is a contradiction here, between what we should do (in theory) and what is happening (in practice).

Wine evaluation

At heart, the process of wine evaluation involves three characteristics, which may be only tenuously connected: (i) the physical wines, which can vary in many chemical ways, some of which are generally considered to be desirable; (ii) the physiology of the tasters, who may vary greatly in their ability to distinguish smells and tastes; and (iii) the psychology of the tasters, who may have very different wine preferences. The sticking point is the last one. We expect that experts can truly detect the wine differences (points i and ii), but can they tell us whether we will like them (point iii)? ^{Note 1}

Wine commentators use the 20-point and 100-point quality scales in an apparent attempt to be seen to be objective, rather than merely opinionated. They tell you their opinion on the chemistry of the wine based on their expert physiology, and then express that as a number. However, we can never justify the difference between a score of 89 and 91, let alone deduce from that number that we will like the wine enough to pay for the marketing difference between those two scores.

In that sense, the benefit to the wine business is very parochial — it makes the marketing easier but does not necessarily benefit the consumer. The score has a one-way effect. This is often the way with any marketing technique, of course, but we should not necessarily condone it (unless we are a marketer).

So, the relation between scores and marketing is a one-edged sword, which is why people disagree about scores being used as a marketing tool. Simon Solis-Cohen dislikes it (Wine scores are the worst marketing technique — so stop it!), while Tom Wark (Why reviews and wine scores ARE good winery marketing) and Roger Morris (How scores changed the wine industry — for the better) can see some benefits.

One thing is for sure: if a wine gets a high score from an accepted expert then its price will immediately go up a long way (see Bob Henry's example in my post How many 100-point scores do critics really give?). This is very effective marketing.

Wine scores

My point here is that if quality scores are adjectives then they should not be treated as numbers, because the only purpose of numbers is their mathematical properties, not their linguistic ones. I have discussed the rather bizarre mathematical properties of wine-quality scores in some of my previous posts (see the Wine scores link in the Labels For Posts list at the right of this page).

For my purpose here, the pertinent issue is the idea that the the numbers express more than merely rank order. We expect that a score of 90 is better than a score of 89 — their rank order should mean something. But we do not know how much better a 90 wine is than an 89, nor do we know what criteria were used to decide on this difference. So, the quantitative difference has no explicit meaning.

This becomes a problem when we try to compare the scores to other characteristics that are also expressed as numbers. The obvious one is price, which is always a very precise number (although it may be flexible from time to time, at the whim of the sellers). We anticipate that there will be some sort of reasonably strong relationship between score and price — as price goes up the score should also go up, and vice versa. I have illustrated this general phenomenon a number of times, many of them listed in my post on The relationship of price to wine-quality scores.

However, can we really do very much with this expectation? Can we evaluate whether a wine’s price is a rip-off (the price is way too high for the assigned quality score)? Can we evaluate whether a wine is a bargain (the price is pretty low for the score)? We can certainly try, but this does assume that the scores have more than merely rank-order meaning. In other words, it depends on their mathematical properties.

This sort of analysis is applied in many ways in the wine industry. For example, wine investors use this when proffering advice (see Fine wine investment: super anomalies). In this process, the investment advisor is looking for wines that are anomalous, in the sense that their price is very low for their assigned score, particularly wines with almost identical scores yet very different pricing. Anomalous wines are seen as a good investment (whether they are also a good drink is beside the point).

In this sort of procedure, there is a mathematical formula (or algorithm) that connects all of the numbers, whether they be scores, prices, aging factors, or whatever. The advisor is then a slave to this algorithm, which brooks no argument. Any formula combines numbers to produce a final number, and that number is THE answer. We will either invest or not. But what if one of the numbers (the score) has little in the way of mathematical usefulness? Surely the output must be nonsense. Where is our adjectival opinion now? It has disappeared in a flurry of calculating.

The problem with mathematics is that you can do almost anything with it. Maybe we should give points for peoples’ opinions about the usefulness of wine scores? If you want to see how ridiculous this sort of thing can get, turning words into numbers, then a tolerably well-known example is included below. ^{Note 2}

Scores may be good marketing, but they are bad mathematics. It is truly said that there cannot be too much data; but there can be too many opinions about those data. And there can be too many opinions that are not based on any data. Wine evaluation is subjective, not objective. So, why do we pretend otherwise? Marketing, of course!

---

Note 1.

Denton Marks has recently had this to say about wine quality scores (If this wine got 96 out of 100 points, what is wrong with me? A critique of wine ratings as psychophysical scaling. AAWE Working Paper No. 239, 2019):

Dispassionate expert wine evaluation to educate consumers might seem aimed at increasing market efficiency, since consumer ignorance likely inhibits wine market growth ... Considerable research has explored how ratings correlate with transaction prices, testing whether they help “explain” willingness to pay (WTP); that could suggest that they do indeed educate consumers.

While mixed results say that wine ratings are not necessarily reliable guides to wine quality and WTP, a more fundamental structural difficulty is that, as a form of hedonic quality index, they involve questionable interpersonal comparisons — say, between experts or between an expert and oneself. For example, different tasters may differ over a taste’s appeal (e.g., the existence of “supertasters”). Saying that experts can tell consumers what they will like so they can reliably determine relative enjoyment and willingness to pay is flawed logic.

Well-established critiques of hedonic scaling illuminate the difficulties and raise fundamental questions about the reliability of ratings — in effect, adopting someone else’s preferences as one’s own — and the interpretation of any price-rating correlation.

---

Note 2.

There is an old joke that periodically does the rounds of the internet. The earliest version that I know of dates from 2004 (What does it mean to give MORE than 100%?). It goes like this:

Ever wonder about those people who say they are giving more than 100%? We have all been to those meetings where someone wants you to give more than 100%. How about achieving 103%?

If:
A B C D E F G H I J K L M N O P Q R S T U V W X Y Z
is represented as:
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26

Then:

H-A-R-D-W-O-R-K
8+1+18+4+23+15+18+11 = 98%

K-N-O-W-L-E-D-G-E
11+14+15+23+12+5+4+7+5 = 96%

A-T-T-I-T-U-D-E
1+20+20+9+20+21+4+5 = 100%

B-U-L-L-x-x-x-x
2+21+12+12+19+8+9+20 = 103%

x-x-x-K-I-S-S-I-N-G
1+19+19+11+9+19+19+9+14+7 = 127%

So, one can then conclude, with mathematical certainty, that while Hard Work and Knowledge will get you close, and Attitude will get you there, it is other things that will put you over the top.

Which countries export most of their wine production?

Exports are of interest to most wine-producing countries, even if only as a national revenue stream. When a country is in a good location for wine production, then that production will usually exceed local consumption. The excess can then be exported to places where wine is harder to produce.

An interesting question, then, is how much of each country’s wine production is exported. For example, it has been noted by Vinex that: despite being sixth largest producer in the European Union, Romania exports less wine than does Sweden. Clearly, exports are not a big thing for the Romanian wine industry.

The data graphed below come from the OIV database, which has collected worldwide data since 1995 (ie. the past 25 years). I have looked only at countries with (almost) complete data for both “Wine Production” and “Wine Exports” up to 2017; and I have excluded those countries where exports exceed production (ie. imports + re-exports are the dominant form of export). This results in 41 countries, 22 of which appear in the following graphs.

Of the 41 countries, one third have changed in one consistent direction through time, either increasing or decreasing their export percentage. Three other countries changed between the first and second halves of the time period.

We can start this survey by looking at the three biggest wine producers in the world (see Global wine exports). Both Spain and Italy have increased their export percentage, while France has done so to a lesser extent. Only Spain regularly exports >50% of its production these days, indicating that the locals in these places are consuming quite a lot of wine.

Next we can note three other European wine producers that have shown some increases in export percentage during the time period. Only Germany has been consistently increasing; and all three seem to have reached a plateau. Germany is, of course, also one of the world’s biggest importers of wine, especially from Italy and Spain (see Global bulk wine routes visualized), which makes its move into exporting an interesting trend (much of it going to the Netherlands, the UK and the USA). Both Germany and Portugal come close to exporting 50% of their wine production.

As noted above, Romania exports very little of its wine production (as shown in the next graph), usually only 3-4% these days. Some of the other eastern European wine producers, on the other hand, have greatly varying export percentages. Bulgaria has often exported >50% of its wine production. On the other hand, Uzbekistan had a big export boom between 2005 and 2010, and Czechia had one between 2008 and 2015. The latter has since returned to c.10% exports.

In contrast to Czechia, its former compatriot Slovakia has steadily increased its wine export percentage this century (shown in the next graph), so that it is now consistently >50%. Further south-east, Georgia has experienced a revival of its wine industry since 2007, so that it might now exceed 50% exports. Lebanon showed an increase until 2010, although it has since held steady at 20-25% exports.

The big losers in terms of wine exports have been the former big exporters to France, all located in northern Africa: Algeria, Morocco, and Tunisia). Their bulk wine was used to beef-up some of the lighter French wines; and in recent decades this role has been replaced by wine from first Italy and now Spain. So, these African countries now export <3% of their wine production.

The so-called New World wine countries of the southern hemisphere have all grown their wine industries in recent decades (as shown next), and all except South Africa now export >50% of their wine production. Indeed, for Chile and New Zealand the exports sometimes reach 90% of production, which is by far the greatest in the dataset. I have looked at exports from Australia and New Zealand in more detail in previous blog posts.

Finally, we can look at those two countries with relatively large wine productions but small export percentages, and with little recent increase in that percentage: Argentina and the USA. The locals are, indeed, rather thirsty. I have also previously looked in more detail at United States wine imports and exports.

So, the answer to the question posed in the title appears to be: Australia, Chile, Georgia, New Zealand, Slovakia, and Spain, although several other wine-producing countries come close.