Monday, August 29, 2016

Two centuries of Bordeaux vintages — Tastet & Lawton

Vintage charts can be an interesting guide to the general features of each vintage, particularly for those vineyard regions where there is a long series of vintage assessments. In these cases, we can study the long-term patterns of vintage variation, without needing to concern ourselves with differences between individual producers.

I have already blogged about Fifty years of Bordeaux vintages, based on the Wine Cellar Insider Bordeaux Vintage Ratings from 1959 to 2014. In this post I expand the time-frame to 200 years for the same wine-producing region.

These long-term data on the Bordeaux red-wine vintages comes from Tastet & Lawton, the oldest surviving courtier firm in Bordeaux.

Tastet & Lawton (Courtiers Assermentés) was established in 1740 by Abraham Lawton, an Irish immigrant. In the Bordeaux wine trade, the courtiers (brokers) act as intermediaries between the chateaux (wine producers) and the négociants (wine merchants) (see En primeur, negociants, courtiers, the Bordeaux wine system explained). There are c. 100 significant chateaux and c. 300 négociants, along with c. 90 courtiers.

There are two versions of the Tastet & Lawton vintage chart available online, one in French and one in English. I have put my own English-language version online because there are a number of inconsistencies in the originals, which I have tried to correct.

The Tastet & Lawton Bordeaux Vintage Reports, 1795 to 2005

Although vintage assessments are available from the 1795 vintage up to and including 2005, the recording of harvest quality did not start until 1900. The quality is scored on an increasing scale:
Mauvaise (Poor), Médiocre (Mediocre), Moyenne (Average), Assez bonne (Rather good), Bonne (Good), Très bonne (Very good), Exceptionnelle (Exceptional)
The harvest volume is not recorded using any consistent wording through time, with at least 20 different expressions being used. General comments on the vintage are provided in all cases.

Year Harvest
Harvest volume Harvest
1795  24-sept.  Not very abundant  Very good year. 
1796  30-sept.  Not very abundant  Mediocre year, thin wines. 
1797  02-oct.  Not very abundant  Mediocre year, thin wines. 
1798  13-sept.  Passably abundant  Marvelous vintage, still spoken of 20 years later; full, robust silky wines. 
1799  05-oct.  Not very abundant  Poor year. 
1800  23-sept.  Not very abundant  Poor year. 
1801  14-sept.  Not very abundant  Poor year. 
1802  23-sept.  Average harvest  Marvelous year, almost rivals 1798. 
1803  25-sept.  Average harvest  Less than 1802, nonetheless good. 
1804  15-sept.  Average harvest  Mediocre year. 
1805  23-sept.  Abundant harvest  Mediocre year. 
1806  20-sept.  Ordinary harvest  Poor year. 
1807  11-sept.  Small harvest  Good year. 
1808  13-sept.  Ordinary harvest  Good year, ordinary wines. 
1809  06-oct.  Small harvest  Very poor year. 
1810  19-sept.  Ordinary harvest  Passable year. 
1811  14-sept.  Rather abundant  The most remarkable wines; known as “The Year of the Comet”. 
1812  21-sept.  Very ordinary harvest  Weak, insignificant year. 
1813  04-oct.  Ordinary harvest  Mediocre year, lacking character. 
1814  29-sept.  Ordinary harvest  Very good. 
1815  25-sept.  Very little wine  Marvelous, on a par with the 1798 and 1811. 
1816  27-oct.  1/4 volume of an abundant year  Detestable year. 
1817  03-oct.  1/5 normal harvest  Very ordinary year. 
1818  17-sept.  Half-harvest  Rather good year, although a tad hard and of little reputation. 
1819  20-sept.  Ordinary harvest  Perfect, admirable year. 
1820  30-sept.  Good half-harvest  Year lacking character, without color. 
1821  04-oct.  Smaller harvest than 1819  Insignificant, mediocre year. 
1822  27-aug.  Very ordinary harvest  Rather dry but good, robust wines. 
1823  07-oct.  Very ordinary harvest  Without color, value or reputation the first year; later a very elegant wine with a well-deserved success. 
1824  04-oct.  Very little wine  Disappointing year. 
1825  11-sept.  Ordinary harvest  Very robust wines which became quite admirable, but which took a very long time to develop. 
1826  20-sept.  Rather abundant  Very mediocre. 
1827  20-sept.  Abundant harvest  Without reputation at first, but ending up admirably well. 
1828  15-sept.  Ordinary harvest  Marvelous year for wines with elegance, grace, fragrance, delicious taste. 
1829  26-sept.  Ordinary harvest  Detestable year. 
1830  16-sept.  Very minimal harvest  Mediocre year. 
1831  14-sept.  Very minimal harvest  Excellent year, very mellow<./td>
1832  22-sept.  Very minimal harvest  Ordinary year, without charm. 
1833  21-sept.  Abundant harvest  Ordinary year. 
1834  09-sept.  Very minimal harvest  Excellent year. 
1835  23-sept.  Abundant harvest  Light, elegant wines (average quality). 
1836  24-sept.  Minimal harvest  Very poor year. 
1837  25-sept.  Abundant harvest  Ordinary year. 
1838  29-sept.  Very minimal harvest  Very mediocre year. 
1839  26-sept.  Average harvest  Ordinary year<./td>
1840  17-sept.  Abundant harvest  Good year. 
1841  18-sept.  Very abundant  Excellent year. 
1842  20-sept.  Minimal harvest  Ordinary year. 
1843  06-oct.  Very minimal harvest  Very poor year. 
1844  15-sept.  Abundant harvest  Ordinary year. 
1845  10-oct.  Very minimal harvest  Very poor year. 
1846  14-sept.  Rather abundant  Good, robust wines. 
1847  27-sept.  Very abundant  Exquisite wines, not very robust. 
1848  20-sept.  Very abundant  Exquisite, robust wines. 
1849  22-sept.  Average harvest  Ordinary year. 
1850  29-sept.  Abundant harvest  Very light, ordinary wines. 
1851  27-sept.  Average harvest  Good, robust wines. 
1852  24-sept.  Average harvest  Very light wines, rather good. 
1853  09-oct.  Very minimal harvest  Very poor year. 
1854  06-oct.  Not very abundant harvest due to powdery mildew  Very mediocre wines, considered good at first and sold at a high price. 
1855  07-oct.  Not very abundant harvest due to powdery mildew  Very mediocre wines, considered good at first and sold at a high price. 
1856  01-oct.  Not very abundant harvest due to powdery mildew  Very ordinary wines showing a taste of powdery mildew. 
1860  26-sept.  Abundant harvest  Very light, very bad wines. 
1861  22-sept.  Very small harvest due to major frost on May 6th  Good wines, elegant and very expensive. 
1862  20-sept.  Rather abundant  Average quality. 
1863  23-sept.  Not very abundant  Adequate quality but lacking maturity. 
1864  17-sept.  Very abundant  Exquisite, extremely mellow, ripe, fragrant and rich (balanced wines). 
1865  06-sept.  Very abundant  Wines good and ripe, but hard; took a long time to come together. 
1866  21-sept.  Average harvest  Very poor. 
1867  18-sept.  Rather abundant  Ordinary. 
1868  07-sept.  Rather abundant  Coarse wines, hard, lacking charm, too expensive. 
1869  15-sept.  Very abundant  Remarkable and complete wines. 
1870  10-sept.  Rather abundant  Wines very good, very ripe, and very full-bodied. 
1871  18-sept.  Rather abundant  Wines light but very elegant. 
1872  22-sept.  Not very abundant  Ordinary year. 
1873  20-sept.  Not very abundant  Ordinary year; terrible frost April 28th. 
1874  14-sept.  Very abundant  Very good wines. 
1875  24-sept.  Very abundant  Wines very good and elegant. 
1876  26-sept.  Not very abundant  Very ordinary year. 
1877  20-sept.  Rather abundant  Light year but charming wines. 
1878  19-sept.  Rather abundant  Very good year. 
1879  09-oct.  Not very abundant  Ordinary year<../td>
1880  21-sept.  Not very abundant  Ordinary year. 
1881  12-sept.  Not very abundant  Wines solid, but without charm. 
1882  28-sept.  Average harvest  Very light wines, rather elegant; touched by mildew. 
1883  27-sept.  Average harvest  Light wines, very ordinary quality. 
1884  25-sept.  2/3 normal harvest  Mediocre wines; mildew. 
1885  29-sept.  1/2 normal harvest  Very ordinary quality, wines touched by mildew. 
1886  25-sept.  2/3 normal harvest  Ordinary quality, wines deeply affected by mildew. 
1887  19-sept.  1/2 normal harvest  Full-bodied, healthy wines thanks to treatment against mildew. 
1888  02-oct.  Abundant harvest  Good year, elegant wines. 
1889  29-sept.  Abundant harvest  Decent year, rather elegant wines. 
1890  29-sept.  Average harvest  Full-bodied, deeply-colored wines; rather good. 
1891  02-oct.  Average harvest  Mediocre and green wines. 
1892  22-sept.  1/2 normal harvest  A blast of hot air August 15th—45°C, wines burnt and lacking color; elegant. 
1893  15-aug.  Exceptionally abundant  Wines thought to be excellent became major disappointments. 
1894  05-oct.  1/2 normal harvest  Wines of little quality, green and meager. 
1895  22-sept.  Average harvest  A year of great heat. Many wines spoiled by acidity. The wines which were spared were remarkable. 
1896  20-sept.  Very abundant  Fine and delicate wines. 
1897  20-sept.  1/2 normal harvest  Very mediocre wines. 
1898  23-sept.  1/2 normal harvest  Wines that were a little hard but came to be appreciated later. 
1899  24-sept.  Abundant harvest  Year of very great quality. 
1900  24-sept.  Very abundant  Exceptional  Great vintage of the century. 
1901  19-sept.  Very abundant  Mediocre  Thin wines which sold for practically nothing. Certain wines turned out to be very pleasant. 
1902  11-sept.  Abundant  Poor  Wines without character. Forget them. 
1903  23-sept.  Abundant  Poor  Insignificant year. 
1904  21-sept.  Abundant  Good  Wines which disappointed with age. 
1905  14-sept.  Abundant  Average  Light but very elegant wines. 
1906  18-sept.  1/2 normal harvest  Good  Wines with exceptionally full body. 
1907  10-sept.  Abundant  Average  Light, elegant, resembling the 1905s. 
1908  23-sept.  Average  Mediocre  Wines a little hard, lacking charm. 
1909  17-sept.  Average  Mediocre  Disappointing wines. Believed to be good at first. 
1910  23-sept.  1/4 normal harvest  Poor  Very mediocre vintage. 
1911  04-oct.  Average  Rather good  A very hot year. 
1912  18-sept.  Abundant  Poor  Wines lacking a lot of structure. 
1913  23-sept.  Abundant  Poor  Flat wines. 
1914  24-sept.  Average  Mediocre  Wines which were believed to be very good turned out to be great disappointments. 
1915  17-sept.  1/2 normal harvest  Poor  A year to forget. 
1916  20-sept.  Average  Good  Solid and full-bodied wines, lacking a bit of charm. 
1917  22-sept.  Average  Mediocre  Light and fragrant wines. 
1918  17-sept.  Average  Average  Wines healthy but harsh. 
1919  20-sept.  Abundant  Average  Light wines without great quality. 
1920  20-sept.  Average  Good  Still several fine bottles. 
1921  17-sept.  Average  Very good  Extremely hot year, difficult to age wines. Still several great bottles. 
1922  11-sept.  Very abundant  Mediocre  Wines light and flat. 
1923  11-sept.  Average  Average  Wines with little color, somewhat burnt. 
1924  27-sept.  Abundant  Very good  Certain wines are still exceptional. 
1925  16-sept.  Abundant  Mediocre  Green wines lacking maturity. 
1926  28-sept.  1/2 normal harvest  Very good  Certain wines are still exceptional. Much coulure. 
1927  27-sept.  Average  Poor  Very little vintage. 
1928  23-sept.  Average  Exceptional  Remarkable wines; some a little hard. 
1929  18-sept.  Average  Exceptional  Great vintage of the century. 
1930  19-sept.  1/2 normal harvest  Poor  A vintage to forget. 
1931  30-sept.  Average  Mediocre  Year close to the preceding vintage. 
1932  20-sept.  1/2 normal harvest  Poor  Comparable to 1930. 
1933  10-oct.  Average  Average  Light and fragrant wines. 
1934  21-sept.  Abondant  Very good  Several great bottles that are still pleasant to drink. 
1935  10-sept.  Abundant  Mediocre  Green wines lacking maturity. 
1936  25-sept.  Average  Mediocre  Green wines lacking maturity. 
1937  26-sept.  Average  Very good  Still many great wines. 
1938  13-sept.  Average  Average  In decline. “Useful” wines without great quality. 
1939  21-sept.  Very abundant  Average  In decline. Light and fragrant wines. 
1940  24-sept.  Average  Rather good  Rather light wines. 
1941  20-sept.  Average  Poor  No successes. 
1942  30-sept.  Average  Average  With the 1943s, the best year of this period. 
1943  14-sept.  Average  Very good  Several great successes. 
1944  09-sept.  Average  Average  Little year. 
1945  21-sept.  1/2 normal harvest  Exceptional  Disastrous frost on May 2nd. Very concentrated wines, many still surprisingly young. 
1946  07-sept.  Average  Average  In decline. Green wines, lacking maturity. 
1947  25-sept.  Average  Very good  Several legendary bottles which approach perfection. Wines full of charm. 
1948  15-sept.  Average  Good  Several exceptional wines. 
1949  22-sept.  Average  Very good  Magnificent wines: can still continue to age. Similar to 1947. 
1950  22-sept.  Abundant  Good  Light, pleasant wines. Several great wines. 
1951  18-sept.  Average  Mediocre  Very average wines. 
1952  02-oct.  Average  Very good  Firm, tannic, long-aging: some great bottles. 
1953  11-sept.  Average  Very good  Great classic year: balance, richness, finesse, long-lasting. 
1954  22-sept.  Average  Mediocre  In decline. Clean, healthy wines but lacking in maturity. 
1955  14-sept.  Average  Very good  Firm, concentrated, long to open up. Several very great bottles. 
1956  07-oct.  1/4 normal harvest  Mediocre  Catastrophic frost in February; little or no production. 
1957  30-sept.  Not abundant  Average  Alarming degree of coulure. With few exceptions, wines in decline. 
1958  06-oct.  1/2 normal harvest  Average  Little vintage. 
1959  20-sept.  1/2 normal harvest  Exceptional  Hot and dry year. Difficult vinifications. Very great wines, still improving with age. 
1960  15-sept.  Average  Average  Rather average vintage, on the decline. 
1961  17-sept.  Very little  Exceptional  Great vintage of the century. 
1962  01-oct.  Abundant  Very good  Pleasant, charming wines to taste; reminiscent of the 1924s. 
1963  01-oct.  Abundant  Mediocre  Very average year. 
1964  21-sept.  Abundant  Very good  Uneven quality; several great wines for long aging. 
1965  30-sept.  Abundant  Mediocre  Disappointing year. 
1966  20-sept.  Average  Very good  Classic vintage of great elegance. The decade’s best, after 1961. 
1967  25-sept.  Average  Rather good  Average vintage in the Médoc. 
1968  22-sept.  Average  Mediocre  On the decline compared to the previous vintage. 
1969  23-sept.  Little  Average  Average year. 
1970  24-sept.  Very abundant  Very good  Tannic and rather hard wines. The best will age into the next century. 
1971  27-sept.  Little  Very good  Wines that are elegant and a little tender, the finest are at their best. 
1972  09-oct.  Average  Average  Often average wines. 
1973  20-sept.  Very abundant  Average  In decline; very few successes. 
1974  24-sept.  Abundant  Average  Disappointing year. 
1975  22-sept.  Average  Very good  Tannic wines of uneven quality. Some great successes. 
1976  13-sept.  Abundant  Good  Quickly evolving. 
1977  04-oct.  Little  Average  Frost on March 31st and April 9th. Hard wines without a future. 
1978  05-oct.  Average  Very good  Classic wines, elegant, at their best. 
1979  01-oct.  Very abundant  Good  The best wines are harmonious and balanced. 
1980  10-oct.  Average  Rather good  Supple and fruity. Most are on the decline. 
1981  28-sept.  Average  Good  Wines at their maturity. 
1982  13-sept.  Very abundant  Exceptional  Magnificent wines, long-lasting, through to the middle of the next century. 
1983  26-sept.  Abundant  Very good  Fully expressive wines. 
1984  01-oct.  Average  Average  Year of often disappointing wines. 
1985  25-sept.  Abundant  Very good  Already remarkable wines. 
1986  26-sept.  Very abundant  Very good  Very great year for long-aging wines in the Médoc. 
1987  01-oct.  Average  Rather good  Supple wines, fruity and elegant; satisfying to drink while young. 
1988  28-sept.  Very abundant  Very good  Classic wines which need to age. 
1989  03-sept.  Abundant  Exceptional  Very great year for long-aging wines. 
1990  10-sept.  Very abundant  Exceptional  Great, hot year. Wines with great aging potential. 
1991  27-sept.  Very little  Rather good  Frost on April 21st. Red wines at their plateau. 
1992  24-sept.  Very abundant  Average  Red wines at their plateau. 
1993  20-sept.  Very abundant  Rather good  Rather rapidly evolving red wines. Very good. 
1994  13-sept.  Abundant  Good  Classic red wines. 
1995  13-sept.  Abundant  Very good  Homogeneous year. Great red wines, supple and harmonious. 
1996  18-sept.  Abundant  Very good  Great year for long-aging wines in the Médoc. 
1997  07-sept.  Abundant  Good  Fruity, supple, charming wines. 
1998  23-sept.  Abundant  Very good  Very successful wines, endowed with good potential. 
1999  30-aug.  Abundant  Good  Very harmonious vintage which can age well. 
2000  18-sept.  Abundant  Exceptional  Greatest vintage of the century, similar to the 1900! 
2001  25-sept.  Abundant  Very good  Better structured than 1999. Elegant vintage. 
2002  24-sept.  Average  Good  Cabernet very successful. Powerful and structured. 
2003  02-sept.  Average  Very good  Heat wave in summer. Much potential to be a great vintage. 
2004  16-sept.  Abundant  Very good  Close to 2002, more fruity, more dense. 
2005  18-sept.  Average  Exceptional  Historically great vintage. Generally successful. 

Analysis of the data

The harvest quality has also been given a rating score (scale: 0-20) by Tastet & Lawton, based on the following equivalence with the words above:
   19-20  Exceptional
   17-18  Very good
   15-16  Good
   13-14  Rather good
   10-12  Average
     3-9  Mediocre
     0-2  Poor

These scores are shown as a time series in the first graph. The blue line connects the data points, whereas the pink line is a running average (it shows the average value across groups of 9 consecutive years, thus smoothing out the long-term trends).

Bordeaux vintage scores 1900-2005

Clearly, through time there has been a general increase in score, along with a decrease in between-year variation. That is, vintage quality has both improved and become more consistent. This matches the trend discussed in the post on Fifty years of Bordeaux vintages, and extends it back a further 50 years. Being a Bordeaux vigneron has been a more reliable profession than it was before then.

The worst years for Bordeaux red wines were during the periods 1910-1915 and 1930-1935, while the best years were 1920-1925, 1945-1950, and the time since 1980. Indeed, there has been no vintage score below 10 since 1970.

On a technical matter, we can compare the Tastet & Lawton vintage-scoring system with that of the Wine Cellar Insider (produced by Jeff Leve). This is done in the second graph, where each point represents a vintage scored by both systems.

Vontage quality scores from Tastet & Lawton versus Wine Cellar Insider

If the two scoring systems are identical then the points should lie on or near the pink line. However, the Tastet & Lawton score is almost always greater than that of the Wine Cellar Insider (so that the points lie below and to the right of the line). This means that a curved relationship fits better than a linear one, as shown by the blue line (the coefficient of variation = 0.87 for this line, which indicates a good fit between the line and the points). This line indicates that Tastet & Lawton have given higher scores for the poorer vintages compared to the Wine Cellar Inside, but they have been in closer agreement for the better vintages. Perhaps this is related to the fact that the Wine Cellar Insider scores were give retrospectively whereas the Tastet & Lawton scores were produced at the time of the vintage

The Tastet & Lawton data also include the official start date for each vintage. Regarding harvest dates, V. Daux et al. (2012. An open-database of grape harvest dates for climate research: data description and quality assessment. Climate of the Past 8: 1403-1418) note:
Prior to the French revolution, in most areas, wine owners were not free to harvest at their convenience. They had to wait for a public order to harvest (“ban des vendanges”) ... After the revolution, vine-growers were theoretically free to begin the harvest when they wanted to. In practice, most parishes maintained a compulsory minimum date for grape harvest in order to preserve law and order. After 1791, the date was set by the mayor, on vine growers’ advice. When several varieties were grown in the same area, the official date corresponded to the earliest and/or to the most important variety. The law of 9 July 1889 entitled the municipalities to keep or give up the “ban des vendanges” practice, which led to its disappearance in almost all France ... In France, since 1979, the harvest date has been determined by the Prefect’s service on a proposal of the National Institute of Origin and Quality, after the notice of the agency defined for each vineyard, taking into account the planting and the status of vineyards.

Bordeaux vintage start dates 1795-2005

The harvest-date information is plotted in the third graph, which shows a count of how many vintages started on each date. These data cover the full 210 years (1795-2005), unlike the vintage scores.

Mostly, vintage has started in the last 2 weeks of September. However, two outlying vintages are shown, one a month earlier (1893) and one a month later (1816) than usual.

There are also intriguing patterns within the general harvest period. For example, 20 September is greatly over-represented compared to either 19 September or 21 September, and no harvest is listed as starting on either 8 September or 8 October. Perhaps these patterns relate to not officially starting work on commemorative dates. For example, 21 September 1792 was during the French Revolution, and on that date the National Convention declared France a republic and abolished the absolute monarchy. Similarly, on 8 October 1453 an agreement was signed whereby the bastille of Bordeaux would be handed over to the French, along with all prisoners in English hands, thus marking the end of English rule of Gascony (where Bordeaux in located).

Bordeaux vintage start dates 1795-2005

The starting-date data can also be plotted as a time series, as shown in the final graph. The harvest-start dates have been numbered 1-75, so that 1 = the earliest starting date. Once again, the blue line connects the data points, whereas the pink line is a running average (it shows the average value across 9 years).

There are two patterns worth noting here. First, the data show long-term cycles (ie. peaks and troughs), with peaks every 35 years during the 1800s and peaks every 45 years during the 1900s. This presumably relates to long-term patterns in the weather. Second, the recent harvests (since 1990) have been among the earliest ever. This coincides with the generally higher quality of the vintages during that time. This is a consequence of the generally warmer weather (see the post on Fifty years of Bordeaux vintages), as the grapes ripen faster and therefore need to be picked earlier.

Monday, August 22, 2016

United States wine production: 1880 versus 2015

The first successful commercial vineyard and winery in the USA was established in 1798 in Lexington, Kentucky, by John James Dufour. There had been attempts as early as 1619 in Virginia, but none succeeded. After Dufour, Nicholas Longworth built the Ohio wine industry, and George Husmann founded the Missouri wine industry; and in the 1850s, Ohio was the largest wine-producing state, based on the American catawba grape. Vineyards had been established at Spanish missions in California in the late 1700s, but a commercial industry was not established until the mid 1800s, by people such as Charles Kohler and Andrea Sbarboro. You can read about this history in the book by Thomas Pinney, The Makers of American Wine: A Record of Two Hundred Years (University of California Press, 2012).

From these beginnings, wine production has clearly increased in the USA. There was a serious set-back during Prohibition (1920–1933), but there has also been a dramatic increase since 1970. It therefore could be of interest to make some sort of quantitative comparison of "then" versus now.

The data that I have used are for the years 1880 and 2015, and come respectively from these two sources:
In both publications, annual wine production data are provided in gallons for each state of the USA for which there was significant wine volume at the time.

These data are shown in the graph (click to enlarge). In the scatterplot, each point represents a single state, located according to its volume of wine production during 1880 (horizontally) and 2015 (vertically). Note the logarithmic scale for both axes. Also, note that the states labeled in red actually had less wine production than is shown by their position on the graph — they have been placed directly on the graph axes for convenience.

US wine production in 1880 and 2015

The pink line on the graph shows equivalence. That is, any state on (or near) the line had the same volume of wine production in 2015 as it did in 1880; any state above the line had a greater production in 2015 than in 1880; and any state below the line had greater production in 1880 than in 2015.

Clearly, most states increased their wine production from 1880 to 2015. This includes five states that now produce considerable amounts of wine but which were not listed as wine-producing states in 1880 — these are labeled in red at the left of the graph.

There are also five states that have increased their production by more than two orders of magnitude: Oregon (800 times), Florida (165), Pennsylvania (135), Minnesota (130) and Massachusetts (110); along with Wisconsin (90 times).

California has always been the biggest wine-producing state in the USA, and it now produces 50 times as much wine as it did in 1880. Washington, New York and Pennsylvania all now (2015) produce more wine than California did back in 1880, with Oregon producing about the same amount.

Interestingly, there are two states (labeled in blue on the graph) that have hardly changed their wine production at all — Iowa and New Mexico. This does not necessarily mean that their production has never changed through time. In fact, it probably decreased during Prohibition, and has now returned to former levels but no higher.

Finally, there are eight states whose wine production is now considerably less than it was 135 years ago. Indeed, two of these states (labeled in red at the bottom of the graph) are not now recognized as producing much wine at all, even though they were ranked 14th and 15th back in 1880. Other states are making a comeback, after having their wine volume drop to a low level. Missouri, in particular, was the second-biggest producer of wine in 1880, and it is now slowly making a comeback, with the establishment of new vineyards.

This post was inspired by an earlier one from the Hogshead blog: The top ten wine producing states in 1880 as compared to 2012.

Monday, August 15, 2016

Quantifying value-for-money wines - part 4

This is the fourth of a four-part set of blog posts looking at how we might identify value-for-money wines. The topic is not as simple as we might like.
Quantifying value-for-money wines - part 1
   — issues with quantifying value for money
Quantifying value-for-money wines - part 2
   — empirically comparing wines within a specified wine group
Quantifying value-for-money wines - part 3
   — formulae for assessing individual wines against a baseline wine
Quantifying value-for-money wines - part 4
   — empirically comparing wines across wine groups

Empirically comparing wines across wine groups

In a previous post (Part 2) I considered empirical methods for trying to quantify value for money as related to wine, which is usually expressed as a quality to price ratio (QPR). The QPR method that I presented involved fitting the exponential (log-linear) model to the price and quality data for a set of comparable wines.

In this post I discuss a more sophisticated version of this approach. It is still based on the exponential model (and thus assumes a non-linear relationship between wine price and quality; see the post The relationship of wine quality to price), but it incorporates many more of the factors that might influence wine price. It estimates the QPR value for an "average" wine, which then allows the QPR value to be calculated for any other wines, of any type.

Hedonic Pricing

Hedonic Pricing is a method to estimate the "expected" price of the goods within a specified market, based on the principle that the goods are affected by certain factors that can raise or lower the "base" price. The factors can be intrinsic to the goods, or part of the external environment, or solely in the mind of the purchaser.

For wine, the obvious factors that could potentially affect the price include: grape type, vineyard location, winery reputation, vintage year, wine quality rating, cellaring potential of the wine, and current market demand. However, we can include as many factors as we like.

Based on this economic model, we can estimate the "expected" market price for wines by using the mathematical technique of multiple regression (called a hedonic regression). We collect a large dataset of wines, along with a record of all of their required characteristics, and we then perform the mathematical analysis. This yields a single equation that summarizes the model.

In practice, we use the model by entering into the pre-calculated equation the data for any specified wine, and this will produce the expected average price for that wine. If the wine costs less than this expected average then it is good value for money (a high QPR).


This is quite a popular approach to estimating expected prices, not just for wine but for many things, notably house prices. There is a list of many of the applications of Hedonic Pricing to wine in the post on The relationship of wine quality to price. In particular, the References listed below have explicitly applied hedonic regressions in order to identify value-for-money wines.

Unfortunately, I cannot draw a simple graph to illustrate Hedonic Pricing, the way I could in the previous Parts of this series of blog posts — the model forms a complex multi-dimensional pattern (one dimension for each factor included in the analysis). (The best way to illustrate the hedonic pricing model as applied to finding value-for-money wines would be to plot the wine value as predicted by the model against the actual wine price.) However, the principle is similar to that used in Part 2 of these blog posts. In particular, the model used is usually log-linear / exponential.

Importantly, the method is more comprehensive than that used in Part 2, because more influences on price are included, not just wine quality. Moreover, we do not need separate data analyses for different grape types, but instead they are all included in one giant analysis.

Also, the method is potentially easy to use in practice, because we have a single equation that we can use to estimate whether any given wine is good value for money. This is similar to the methods discussed in Part 3.

However, because everything is "built in", the model cannot be adjusted by the user for a new dataset. In particular, if vintage year is included as a factor in the model, then the equation needs to be recalculated after each new vintage.


Jon R. Miller, Robert W. Stone, Eric T. Stuen (2015) When is a wine a bargain? A comparison of popular and regression-based approaches. Journal of Wine Research 26:153-168.

Eddie Oczkowski (1994) A hedonic price function for Australian premium table wine. Australian Journal of Agricultural Economics 28:93-110.

Edward Oczkowski (2001) Hedonic wine price functions and measurement error. Economic Record 77:374-382.

Edward Oczkowski (2010) Hedonic wine price predictions and nonnormal errors. Agribusiness 26:519-535.

David A. Priilaid, Paul Van Rensburg (2006) Non-linearity in the hedonic pricing of South African red wines. International Journal of Wine Marketing 18:166-182.

David A. Priilaid, Paul Van Rensburg (2012) Nonlinear hedonic pricing: a confirmatory study of South African wines. International Journal of Wine Research 4:1-13.

Paul Van Rensburg, David A. Priilaid (2004) An econometric model for identifying value in South African red wine. International Journal of Wine Marketing 16:53-75.

Australian Wine Price Calculator

The first application of Hedonic Pricing to wine was by Eddie Oczkowski in 1994 (see the References). In 2002, he provided a web page for applying his pre-calculated Hedonic Pricing model for Australian wine, which is available at the Australian Wine Price Calculator. This appears to be the only available online calculator for wine that uses Hedonic Pricing.

This model includes five factors, as shown in the picture. It is currently based on a sample of 8,774 Australian premium wines (average price AU$34). The data cover the vintages 2004-2014 (as well as non-vintage wines), with cellaring potential until 2060. There are 82 grape varieties and blends included (with Shiraz as the most commonly occurring one in the dataset), along with 80 named grape-growing regions (McLaren Vale is the most common).

Where to from here?

The empirical QPR approach (see Part 2) should work best for identifying those wines with a good quality : price ratio, because it provides the best fit to the data for each wine category. However, this is rather impractical for assessing individual wines.

The wwpQPR approach (see Part 3) is much more practical for on-the-spot decisions when purchasing wine, because it uses a simple formula to evaluate any given wine. You do need to access the Wellesley Wine Press web page to do the calculations.

The Hedonic Pricing model (as discussed above) tries to combine the best aspects of these two approaches. It provides a good fit to the data across wine categories, as well as providing a simple formula to evaluate any given wine. However, there is currently only one web page available (for Australian wine only), which needs to be kept up to date after each vintage.

Monday, August 8, 2016

Fifty years of Bordeaux vintages

Vintage charts are intended to show us how the harvest quality has varied from vintage to vintage. They are often disparaged, because they simplify the complexities of each harvest (where there can be considerable spatial variation) down into a single number. They also make little sense if they are applied to a very large area, which they too often are in practice.

Nevertheless, they can be an interesting and informative guide to the general features of each vintage within a restricted area, even if they tell us nothing about each wine producer within that area, and how each producer dealt with the vagaries of harvest time.

This particularly applies to those vineyard regions where there is a long series of vintage assessments. In these cases, we can study the long-term patterns of vintage variation, without needing to concern ourselves with differences between individual producers, within or between vintages.

There are several regions where the written vintage records go back at least a century, and I will be covering these over the next few weeks. For example, there are several series for both Bordeaux and Piemonte, as well as a very long one for the Rhine valley.

Here, I will start with a somewhat shorter series, covering only a half-century, but which I will analyze in some detail.

The Wine Cellar Insider Bordeaux Vintage Ratings, 1959 to Today

Jeff Leve, at The Wine Cellar Insider, has provided vintage rankings for the red wines of the Bordeaux region of France, for every year from the famous 1959 vintage up to and including 2014 (ie. 56 vintages). This assessment is based on the standard 100-point quality scale (ie. ranging from 50–100). The data are shown as a time series in the first graph.

Bordeaux vintage quality 1959-2014

This graph shows several interesting features.

First, there is no score between 50 and 60, and also no score between 70 and 80. This means that three distinct types of vintages can be recognized: 1 disastrous vintage (score 50), 11 mediocre vintages (score 60-70), and 44 good vintages (score 80–100). This is a pretty impressive record for the vignerons.

Second, there is no score of 94, and so we can conveniently score "great" vintages as being ≥ 95 points (this is the pink line on the graph). This gives us 9 such vintages, out of the 44 "good" ones, which is also pretty good for the wine-makers. Oddly, there are no scores of either 98 or 99, but there are two scores of 100 points.

Third, there have been no mediocre vintages since 1992, and only two mediocre vintages after 1980. Before that, there were 10 mediocre vintages and 12 good ones. Indeed, before 1993 there were 5 great years, 17 good years, 11 mediocre years and 1 very poor year (in 34 years), while since then there have been 4 great years and 18 good years (in 22 years). This is quite a dramatic change in fortune during the latter part of the series.

This change has often been attributed to improved vineyard management since the 1990s. In particular, nowadays there is usually much more rigorous selection of grapes at harvest, so that poor grapes are simply not included in the best wine. In this sense, it has been claimed that there are no longer "good" years and "poor" years for vintages, but instead there are years with a larger or smaller amount of good wine.

There are also what are called "green harvests", which refers to the taking of unripe (ie. green) bunches of grapes off the vines a few weeks before harvest. This results in uniformly riper grapes, which ensures softer tannins in the ensuing wine, as well as reducing the yield, and thus concentrating the flavor in the remaining grapes. There has also been a recent trend towards late picking of ultra-ripe grapes, which makes the wine more fruity (and more alcoholic). This all makes today’s wines drinkable earlier than those of 20 years ago.

Relationship between Bordeaux vintage quality and summer temperature

However, Orley Ashenfelter (see the references listed below) has suggested, instead, that the recent improvements in harvest quality may actually be due to increased summer temperatures. That is, higher temperatures during the vines' growing seasons have mitigated the need for fancy vineyard management. The relationship between the daily-average summer temperature and the vintage quality score is shown in the second graph, for the period 1959–2009 (the temperature data come from Ashenfelter and Pablo Almaraz). Each point represents one Bordeaux vintage, mapped by its average temperature (horizontally) and its quality score (vertically).

These data show that, mathematically, 37% of the variation in the vintage quality scores during the past half-century is related solely to variation in the summer temperature. Clearly, summer warmth has a large influence on improving harvest quality.

Moreover, the distribution of the summer temperatures in Bordeaux over time from 1952–2009 is shown in the third graph. The pink points indicate the years from 1989 onwards (as they also do in the previous graph), showing that there has been a significant increase in daily-average summer temperature since that time. Indeed, there has been a steady increase in temperature since the low of 1977. European vignerons are in no doubt about the existence of global warming!

Bordeaux summer temperature 1952-2009

Ashenfelter backed up his original assessment of the relationship between vintage quality and summer temperature with some forecasts (inaccurately called "predictions"). These appeared in the New York Times in 1990: Wine equation puts some noses out of joint.

The first forecast was for the 1989 vintage:
Perhaps the most dramatic Ashenfelter prediction is for the 1989 vintage. By Professor Ashenfelter's calculations, the hottest growing season in memory, combined with a very dry harvest, all but guarantee that the 1989 Bordeaux will be stunningly good.
Jeff Leve's assessment in 2015 (15 years later) was a score of 95 for the 1989 vintage. So, this forecast turned out to be spot on.

The second forecast was for the 1986 vintage:
According to the Ashenfelter system, below-average growing season temperatures and above-average harvest rainfall doom the 1986 Bordeaux to mediocrity. When the dust settles, he predicts, it will be judged the worst vintage of the 1980s, and no better than the unmemorable 1974s or 1969s.
Jeff Leve's 2015 assessment was: 1969 vintage = 60 points, 1974 = 65 points, and 1986 = 88 points. Furthermore, there were five worse vintages than 1986 during the 1980s, and four better ones. Indeed, the worst vintage of the 1980s was 1980, with 70 points. So, this forecast was pretty far off.

Literature references

Pablo Almaraz (2015) Bordeaux wine quality and climate fluctuations during the last century: changing temperatures and changing industry. Climate Research 64:187-199.

Orley Ashenfelter, David Ashmore, Robert Lalonde (1995) Bordeaux wine vintage quality and the weather. Chance 8(4):7-14.

Orley Ashenfelter (2008) Predicting the quality and prices of Bordeaux wine. Economic Journal 118:F174–F184.

Orley Ashenfelter (2010) Predicting the quality and prices of Bordeaux wine. Journal of Wine Economics 5:40-52. [a reprint of the previous paper]

Monday, August 1, 2016

European wine imports

In a recent blog post (Global wine exports) I chastised the Australian wine industry for principally exporting large amounts of cheap wine, rather than exporting high-quality wine, which they do actually make as well. The latter pays better in the long run, as shown by the leading export countries, such as France and New Zealand. In this post, I look at this same phenomenon at a smaller scale: the European Union.

The data cover the years 2012–2014, and come from:
EU-27: Wine Annual Report and Statistics 2015. United States Department of Agriculture, Foreign Agricultural Service. Global Agricultural Information Network Report No. IT1512.

The graphs show the average cost per litre of wine imported from various countries. The first graph refers to imports into the European Union as a whole.

Wine imports into the European Union 2012-2014

This shows 10 of the top 11 sources of imported wine, excluding Switzerland. Both Israel and Switzerland export very little wine to the European Union, but they make considerably more money per litre out of it than does anyone else (for Switzerland it is $55–85 per litre!).

The big sources of wine for the European Union are, in decreasing order of volume, Chile, Australia, South Africa and the United States. As shown in the graph, these countries make considerably less money per litre out of their exports than do some of the other countries, notably New Zealand, Georgia and Argentina. In other words, they are purveyors principally of cheap wine.

We can look at somewhat finer detail by considering individual countries within the European Union. Two of the biggest destinations for the imported wine are Germany and the United Kingdom. Indeed, these are the two biggest wine importers by volume in the world (see the post Global wine imports).

Wine imports into Germany 2012-2014

The graph for Germany shows that the countries making most money per litre for their wine exports are from the European Union itself (France, Austria and Italy). Of the outsiders, the USA makes considerably more money per litre than does Australia, which makes no more money than does Chile.

The New Zealanders do not appear in the Germany graph because most of their wine goes to the United Kingdom. Indeed, as shown in the next graph, they also make much more money per litre from this wine than does anyone except the French. Australia and Italy are the two biggest suppliers by volume to the UK, and yet they are principally supplying cheap wine, unlike France and New Zealand. [The 2015 data come from GAIN report United Kingdom Wine Market Report 2016.]

Wine imports into the UK 2012-2015

These figures do not shine a healthy light on the Australian wine industry, as I noted previously. There is money to be made by exporting wine, and yet the Australians are not making much of it, in spite of being globally the fifth biggest exporter (with the UK, USA and China as the biggest markets). Exporting cheap wine, rather than better-paying wines, is not a route to financial advancement.

As Max Allen (Australia at the crossroads?) has noted:
The volume of annual wine exports is roughly the same, 670 million litres, as it was a decade ago (after reaching a high of 768 million in 2007) but the value of those exports has dropped by a staggering A$1 billion, fostering a persistent image overseas of Australia as a commodity producer of cheap plonk.
Part of this issue, of course, is the matter of bulk wine, which is bottled in the destination country rather than at the source — this reduces the added value in the country of origin. As Jancis Robinson has pointed out (How wine travels nowadays – in bulk):
In 2008 fewer than three in every 10 bottles of Australian wine on British shelves contained wine that had been shipped from Australia in bulk rather than in bottle. Four years later that figure was eight in every 10, and the total amount of wine shipped out of Australia in bulk overtook the volume exported in bottle.

The Australian wine industry is frequently reported to be in a state of crisis (see, for example Anatomy of wine profit and (mainly) loss: South Africa versus Australia). This is principally due to an over-supply of grapes of mediocre quality (principally from the irrigated vineyards of the Murray-Darling basin), along with a very centralized production and distribution system (focused on two supermarket chains) that is interested in quantity rather than quality. Indeed, in 2013 John Angove noted that Australia is well overpopulated with wineries, as it has one winery for every 8,800 population, compared to the USA figure of one winery for every 55,000 population.

Viv Thomson noted earlier this year:
It was the mid-90s when the Winemakers Federation of Australia put out a forecast for the next 30 years. By 2010 we were well ahead on plantings, but we weren’t ahead on infrastructure, we weren’t ahead on marketing and that brought about the current glut and disaster. Then the high Australian dollar really put the crunch on things ... it got out of hand, the investors got involved, grape prices were sky high ... it was just not sustainable.
Australia has an abundance of high-quality wines, but they currently seem uninterested in letting the world see many of them. They might solve a few of the reported problems if they tried to earn more export dollars per litre than they currently do.