Category Archives: Economics

The Proximity Principle of Wine: How Neighboring Wineries Affect the Market Price of Wines in a Region

Posted on April 1, 2013 by Becca| 6 Comments

 

—————————————————————————————————-

Welcome to The Academic Wino! If you are new here, please read the “About Me” page to find out more about myself and the blog. If you would like to receive free updates on articles like this by email, then sign up here or you can subscribe to the RSS feed. Also, check us out on Twitter, Facebook, Google+, and or Pinterest. Thanks for visiting!

—————————————————————————————————-

“One bad wine in the valley is bad for every winery in the valley. One good wine in the valley is good for everyone.” –Robert Mondavi

Leading in to the article presented today (Yang et al. 2012) was the above quote by Robert Mondavi in regards to the Napa Valley in the 1960s. It’s pretty self-explanatory in that he and many others believe that one bad winery in an area will decrease the attractiveness of all other wineries in the region, and one good winery in the area will increase the attractiveness of all other wineries in that region. Past research has found that the location of a particular winery will have a significant influence on the market prices of those wines and subsequently those wines in the immediate area. Consumers look for recognizable locations on a label when purchasing wines, as well as the price, expert opinions of the wine, and quality.

There are many reasons Yang et al. (2012) cited as to why it is beneficial for wineries to be in close proximity to each other. A winery that is geographically isolated may suffer from the fact that they may not be in an appellation or wine region that consumers recognize. Also, being geographically isolated means that visitor attraction may be decreased, since consumers tend to prefer visiting several wineries in a short span of time instead of visiting only one. Other benefits include potentially lower production costs, as labor is often shared between wineries in close proximity to one another.

Yang et al. (2012) also cited a few reasons why spatial clustering of wineries could negatively affect the price and reputation of the wineries in that area. First, pests or disease pathogens could easily spread from one winery or vineyard to another when in close proximity, thus damaging or destroying the crop from multiple wineries in the area. A winery in isolation is less likely to be infected by this type of pest or pathogen outbreak. Also, a winery focused on producing cheap mass-produced wines could take advantage of the close proximity principle and move into an area with higher quality wines, thus artificially increasing their own prices and increasing their initial chances of being purchased by a consumer, simply by having a recognizable region name on the label. Land prices also tend to be significantly higher in an area with higher quality wines, thus making it more difficult for new wineries to become established.

Taking both the positive and negative effects together ultimately determines how a particular wine region will develop over time. Specifically, these factors influence how wineries will establish themselves spatially, as well as the quality and price of the wines produced in that area. Though it has been generally shown that location affects price of wine, according to Yang et al. (2012), the specific spatial effects of the economics of wine production has been largely unexplored.

The paper presented today aimed to use knowledge obtained from spatial hedonic literature as well as knowledge obtained from hedonic modeling of consumer goods literature and created a spatial econometric model to determine how geographic proximity affects economic relationships among

By Matt Pourney [Public domain], via Wikimedia Commons

By Matt Pourney [Public domain], via Wikimedia Commons

wineries as well as the general reputation of a wine region. Using GIS (geographic information system) data and spatial analyses of the California and Washington wine regions, the authors aimed to get a better understanding of how economic factors as well as other relationships are influenced by geographical location.

Methods

The data used in this analysis was for wineries that produced red wines in Washington and California. For each wine, data on price, scores, case numbers produced, years of aging, vintage, and the production region were collected. Data for each individual winery was pooled to obtain one value per winery.

Wine regions for California that were used in the analysis were: Napa, Sonoma, Mendocino, Bay Area, South Coast, Carneros, Sierra Foothills, and “other” California. Wine regions for Washington that were used in the analysis were: Columbia Valley, Yakima Valley, Walla Walla, Puget Sound, and “other” Washington. A total of 79 wineries were analyzed for Washington and 876 wineries for California. There were more wineries in each area, however, exact location information was not available at the time and were thus not included in analysis.

All information included winery names and addresses were collected from the Wine Spectator online site. Addresses were converted to latitude and longitude GIS coordinates. It was assumed that wine production took place at the address provided by the Wine Spectator database. Since winery information came from the Wine Spectator, the authors acknowledged that the quality limits for wineries used in the analysis were slightly higher than what they may be in reality, as there were no boxed wine or jug wine producing wineries included in the analysis.

Two mathematical models were created for this spatial analysis. The first model was a hedonic price model with a spatial lag parameter which included the following variables: wine score, number of cases produced by the winery, the number of year the wine was aged before release, and the wine region. The second model examined the spatial correlation in wine prices independent from any other variable.

Results

• Modeling results showed that close neighbors have a significant and positive effect on the price of a winery’s own wine.
o Higher-performing wineries have a significant positive impact not only on their own wine prices, but also the prices of their close neighbor’s wine prices. This was true for both Washington and California wineries.
• Washington:
o Wine scores had a significant and positive effect on wine price. (The higher the score, the higher the price).
o The number of cases produced had a significant and negative effect on wine price. (The more cases produced the lower the price).
o The number of years the wine aged before release had a significant and positive effect on wine price. (The longer the aging, the higher the price).
o Specific wine region did not affect wine prices.
• California:
o Wine scores had a significant and positive effect on wine price. (The higher the score, the higher the price).
o The number of cases produced had a significant and negative effect on wine price. (The more cases produced the lower the price).
o Specific wine region significantly affected wine prices. Specifically, the model showed a clear benefit to being in the Napa or Sonoma wine regions (i.e. higher prices in these areas).

Conclusions

According to the authors, this is the first study to use GIS to evaluate the effects of geographic proximity of wineries on wine prices. Their results suggest that there is, in fact, a clustering effect of wineries on the market price of wines in that area, and that the interactions between the wineries in an area are very important in regards to price. It was clear from the results that expert ratings and scores positively affected price (i.e. high score = higher price), and that being located in a more recognized region (i.e. Napa or Sonoma) increased the prices of all wines in that area. It was interesting to note that there was no region effect on prices in Washington state, the reasoning of which was not explored in this paper but may be due to the fact that the region is younger and in general smaller than in California, so brand awareness occurs more on a state-wide level instead of a region-wide level (my theory anyway).

While the results of this study are fascinating, it brings up many more questions that were not addressed and which could make interesting follow-up studies. For

By Agne27 (Own work) [CC-BY-SA-3.0 (http://creativecommons.org/licenses/by-sa/3.0) or GFDL (http://www.gnu.org/copyleft/fdl.html)], via Wikimedia Commons

By Agne27 (Own work) [CC-BY-SA-3.0 (http://creativecommons.org/licenses/by-sa/3.0) or GFDL (http://www.gnu.org/copyleft/fdl.html)], via Wikimedia Commons

example: what are the specific factors or reasons that result in this neighborhood effect? Is it the stories behind the region (i.e. history, romance, etc), or is it the specific terroir and style of the region that has people willing to spend more money?

Also, what happens in the long term when a winery producing lower quality wines tries to take advantage of the good neighbor effect and moves into a region with a lot of higher quality wineries? The authors of this study suggested that there is the incentive for these types of wineries to move into these regions, but may ultimately undermine the area by lower the overall reputation of the area over time. Understanding how these spatial effects equilibrate themselves over the long term would be very fascinating to explore.

Finally, the authors noted that they did not take costs into consideration when developing this model. Would it be economically beneficial for a winery to move into an already established wine region such as Napa or Sonoma where land prices and other costs are markedly higher than in lesser known regions? Do the benefits received from the good neighbor effect outweigh the negative cost effects? Or would it be more economically beneficial to set up camp in a lesser known region that costs less though may not receive as much traffic or as much purchases from consumers since it is a lesser known region? It would be very interesting to see what would happen to the model if these cost factors were taken into consideration.

Overall, the results of this study provide evidence for the close proximity theory for wineries, and that it is beneficial to be close to wineries producing high quality wines. The information found in this study could be helpful for those that are considering investing in a winery in Washington or California, however, future research would need to be done, particularly in regards to costs, in order to be able to make a fully-informed decision on whether or not to set up a winery in a particular region or not.

I’d love to hear your thoughts on this topic! Please feel free to comment and share!

Source: Yang, N., McCluskey, J.J., and Brady, M.P. 2012. The Value of Good Neighbors: A Spatial Analysis of the California and Washington State Wine Industries. Land Economics 88(4): 674-684.

Posted in Economics

Tagged , , , , , , ,

Wine Retail Price Dispersions: A Closer Look into Wine Economics in the United States

Posted on October 4, 2012 by Becca| 4 Comments

Have you ever gone to a store to purchase wine one day and it’s a certain price, then go to buy the same wine in a different store later and you notice it is a few dollars more expensive?  Before you go thinking that you’re just crazy, you should know that the retail wine marketin the United States is known for its large price dispersions.  For example, retail wine price data has shown that a 2007 Yellow Tail Merlot in a wine shop in upstate New York is $4.99, while the same exact wine is $9.99 in Jersey City, New Jersey.  Going to the high end of the spectrum, the same data set found that a 2005 Chateau Latour ranges from $695 in a shop in Petaluma, California to $2,000 in a shop outside Chicago, Illinois.

Photo by Tobyotter: http://farm5.staticflickr.com/4047/4433318139_69f48ef1e6.jpg

Why is it that wine prices are so variable from shop to shop in the United States? 

There could be a variety of factors why price dispersion of wine is so large in the US, including differences in production, differences in consumer demand, or different market regulations set forth by the federal government or the state government.  In regards to regulation, it is primarily up to the individual states how wine is sold within their borders; however, the federal government does require a federal wine tax be paid on all wines.  There is quite a large variation in how each state regulates their wine sales, with some states selling wine only at state-owned wholesale shops, some allowing the sale of wine only at particular types of private businesses or at certain times of the day/week, and some even imparting restrictions on how and where to ship wines to others.

How can price dispersion persist?

If the search for information on a product is ultimately costly for the consumer, price dispersions are able to persist in otherwise stable markets.  In other words, if it takes too much time, effort, and sometimes money to find correct pricing information for a product, the consumer will just forgo this search and assume the sticker price is correct and will willingly pay that amount.  With the advent of the internet, the ability for consumers to search for pricing information has improved dramatically, allowing consumers to more easily find the information they are looking for in much less time than they would have spent without this technological resource.  Even though this is theoretically how it would work, some studies have found that consumers will still give up before they are fully informed on a product, even with the internet at their fingertips.

In some cases, price dispersion may be in jeopardy if the consumer learns through experience.  Since inexpensive wines are purchased more frequently than expensive wines, one can imagine that the same cheap wines are being purchased over and over again by the same individuals.  Therefore, over time, these consumers would learn more about the pricing information of a particular wine by experience or by word of mouth by another consumer, thus forcing a much narrower range of price in the market.  If the consumers know the wine is

Photo by 401(k) 2012: http://farm7.staticflickr.com/6092/6355360253_30e095425d.jpg

only $4.99 at one store, they will be very unlikely to ever purchase the same wine at a different shop for $9.99.  The prices then over time “work themselves out” and form a tighter, less variable price dispersion in the market.

The paper presented today, while brief and more a starting point for further research, aimed to examine whether state effects or county effects explain these price dispersions of wine in the United States, and if the degree of dispersion is the same for both inexpensive and expensive wines.

Methods

The researchers used a large online wine pricing data base encompassing the years 2006 through 2008 in order to collect price information of wines.

To determine how local market characteristics affect wine prices, a complicated mathematically model was employed, which took into account the following factors: wine type, vintage, county, state, year of price posting, number of retail wine establishments in a given county, the county population during the year 2000, race, and age.

Results

  • 186 wine brands were analyzed.
    • A total of 106,000 prices were found for these wines.
  • For nearly all wines (except for one), there were over 200 difference prices found throughout the United States.
    • For many wines (exact number not specified), there were over 1,000 prices found throughout the US.
  • Most of the wines analyzed were from the US, and 2/3 of them were red varietals or blends.
  • The overall price dispersion of the wines analyzed was 23.4%.
    • Price dispersion was higher for reds than for whites.
    • Price dispersion was higher for French wines than for American wines or all other imports.
    • Price dispersion was higher for expensive wines than inexpensive wines.
  • Based on the mathematical model, results showed that prices vary with local market conditions.
    • Even so, local market conditions only explained between 7% (red wines) and 13% (white wines) of the price dispersions.
  • For both red and white wines, the relationship between price dispersion and average price was significantly stronger for vintage wines than nonvintage wines.

Authors Conclusions

The authors of this short study concluded by saying that yes, a wide price dispersion for wine exists in the United States, with the average price dispersion falling around 23.4%.  Based on the mathematical model results, the authors

Photo by markwainwright: http://farm3.staticflickr.com/2063/3528237255_72d3e10f1a.jpg

attributed this dispersion partially to different local market conditions, though differences in this dispersion may also be due in part to price levels, and after controlling for variations due to consumer, market, and state regulations.

Since it was found that more expensive wines carry with them a larger price dispersion than inexpensive wines, the authors conclude that this could be indicative of a “learn by experience” or “word of mouth” phenomenon.  Since more people purchase the inexpensive wines, and since the same individual is likely purchasing the same wines more frequently, they are more informed and aware of what the appropriate price should be.  On the other end of the spectrum, the authors suggest it is also possible that those seeking to purchase expensive wines incur a larger opportunity cost of time, and do not have the time to search out the best price for their wine of inquiry.

Final thoughts

Overall, I thought this study was interesting; however, it left me wanting to know a lot more.  I feel as though this study just gets at the tip of the iceberg that is price dispersion of wine in the United States.  To me, it reads more like a pilot/preliminary study, which is fantastic, but I’m now left with the sense of “oh man, that’s it?”.

I would like to see this study taken a bit further and would like to see some other results teased out.  The authors included many different factors in their mathematical model equation, however, they don’t really go into details about how many of them influence price dispersion.

What about you all?  I would love to hear your thoughts on this study and what sorts of things you would have liked to see done differently (if anything).  What sorts of questions do you think need to be asked next?  Please feel free to leave your comments and participate in the discussion!

Source: Jaeger, D.A., and Storchmann, K. 2011. Wine retail price dispersion in the United States: Searching for expensive wines? American Economic Review: Papers & Proceedings 101(3):136-141.

Posted in Economics

Tagged , , , , ,

The Economic Effect of Climate Change on Viticulture

Posted on April 30, 2012 by Becca| 4 Comments

Regardless of what you believe is the cause, global warming and climate change is occurring.  Depending upon where one looks around the globe, climate change affects specific areas of the world differently.  Specifically, in regards to wine and viticulture in California, studies have shown that global warming could have negative effects on the quality of wine (Pinot Noir, specifically) in the region, which would likely be reflected by lower prices.

Is it possible that some areas of the world will see positive benefits of global warming?

http://www.winerelease.com/Past_Newsletters/
2009/MoselSteepVines1.jpg

The paper reviewed today, though now a couple of years old at this point, aimed to examine the economic impact of global warming on viticulture in the Mosel Valley of Germany, which lies between 49.61o and 50.34o latitude.  Within the Mosel Valley, production of grapes depends upon specific site characteristics such and steep slopes on rocky/infertile soil and specific weather conditions to allow for winter survival and successful ripening.  As a result, wine quality (as well as prices) depend upon weather and can therefore vary widely from year to year.  Due to these specific limitations and characteristics in the Mosel Valley, it is expected that temperature-induced changes due to climate change will have a direct impact on the economics on this part of Germany.

In order to study the economic impact of global warming on viticulture in the Mosel Valley of Germany, the authors used the “Ricardian” approach that has been verified and validated by other studies focusing on the effects of climate change on agriculture.  To be more specific, the authors created their three models based on different price data, including retail, wholesale, and auction prices.

Model and Data

The data this model focused on revenue and its components.  Per hectare revenue between 1997 and 2008 in each of the 5 viticultural areas of the Mosel Valley were examined (Upper Mosel, Middle Mosel, Lower Mosel, Saar, and Ruwer Valley).  Revenue is basically calculated by the product of price and crop yield, though there are some other complexities that result in deviations from this simple formula, such as how wines in the Mosel Valley are labeled and marketed. 

Basically, German wines are classified and labeled according to the natural sugar content of the grape must (unfermented) based on the Oechsle cale (oOe):  the sweeter the must; the higher the alcohol; the stronger the aroma; and finally the higher the quality.  The quality of wines increase in the following order: Quality Wine (no oOe requirement), Kabinett (70oOe), Spätlese (76oOe), Auslese (83oOe), Beerenauslese (110oOe), Eiswein (110oOe), and Trockenbeerenauslese (150oOe).  Wine prices are thereby determined by the vineyard where the grapes were grown and by the quality level.

Data for revenue per hectare are not readily available; however, they can easily be calculated by multiplying crop yield data by the average prices for each region and each wine quality level.  Wine prices by region and by quality are not readily available; however, they can easily be calculated by drawing on various wine price data using three different sources (retail, wholesale, and auction).

Wine production data by region and by quality between 1997 and 2008 was provided by the Statistical Office of the State of Rheinland-Pfalz and its agricultural commission (Landwirtschaftskammer).

A disadvantage to using retail and wholesale price data is that they refer to posted prices, not transaction prices (though sometimes they are).  Conversely, an advantage to using these data is that they cover a wide range of wine producers in the Mosel Valley.  Auction prices, while they do represent actual transaction prices, only a very small percentage of Mosel Valley wines are represented and sold, so auction prices may not be representative of the Mosel Valley region in general.

Retail price data from 1994 to 2008 came from the Gault Millau Wine Guide for Germany.  This guide contained detailed information about wine age, geographic origin, and quality classification, as well as the data to allow for the calculation of wine prices and quality levels per region.  Wholesale price data from 1993 to 2001 came from the Mainz Wine Trade Fair (Mainzer Weinbörse).  Auction price data from 1981 to 2008 came from the wine associations VDP Grosser Ring and Bernkasteler Ring.

Auction wines, though in the past represented a great variety of wines in the Mosel Valley region, primarily serve now as a showcase for a few very high quality wines.  For example, only 0.13% of wines auctioned are Quality Wines (lowest quality), while 74.1% of all wines produced in the Mosel Valley are at the Quality Wine level.  Also, 12% of the wines sold at auction are Eiswein, Beerenauslese, or Trockenbeerenauslese quality levels, whereas these quality levels only represent 0.2% of the total production of the Mosel Valley. 

According to the authors, responses of prices to temperature during the growing season are very sensitive to these higher quality wines, making it likely that these data will suffer from selection bias.  Also, the auction price data are likely to overstate the average effect of temperature on price.  In years of good weather, yield reduction is practices in vineyards production higher quality wines, therefore prices of these quality wines are already partially a result of weather.  Crop yields more fully reflect weather variation in the Upper Mosel region, where quality of wine is lower and yield reduction is seldom practiced.

Results

  •       Wine quality and price are highly dependent upon weather, as seen in other studies.

o   In more northern latitudes, warmer and drier weather are expected to yield higher quality fruit.

o   Warmer weather had a significantly positive effect on prices.

o   Higher quality wines benefitted from a warmer growing season than lower quality wines.

o   The effect of temperature increase on price was greatest in the regions of Saar and Ruwer.

  •        Auction prices were significantly more sensitive to temperature changes than retail or wholesale prices.
  •       There was a greater production of higher quality wines in warmer years.

o   Increases in temperature resulted in an increase in wine prices within each quality level.

o   Increases in temperature resulted in higher number of higher quality wines than lower quality wines.

  •       Revenue per hectare significantly increased with increasing temperatures.

o   The extent of this effect depends on which price structure one is considering:

§  Auction price data suggested increases in revenue of 63% per degrees Celsius increase.

·         Since auction data focuses mostly on high quality wines, this result is most likely an overestimate of the revenue increase due to increasing temperatures.

§  Wholesale price data suggested increases in revenue of 27% per degrees Celsius increase.

§  Retail price data suggested increases in revenue of 37% per degrees Celsius increase.

Conclusions

All three models employed in this study suggest that the vineyards of the Mosel Valley in Germany will increase in value as a result of increasing temperatures caused by climate change.  Auction prices will likely overestimate this increase, whereas retail and wholesale prices more accurately represent the potential effect of global warming on changing prices of wine. 

According to the results of the models, the authors predict that a 3oC increase in temperature would more than double the value of the vineyards in the Mosel Valley region.  A more moderate increase of 1oC is predicted to result in an increase in revenue of about 30%.

One thing that’s not certain is whether we will continue to see this trend as the future progresses, or if we are in a more transitional period with much more change to come.  The only thing we can do is to continue running these models and adjust parameters accordingly depending upon any new changes observed.

The authors described a few more limitations of the models used in this study.  The first limitation is that the model does not take into account any general equilibrium effects that may occur with the restructuring of land prices.  Specifically, if there were to be any dramatic changes in prices of vineyard land itself due to climate change, there could be consequences for the final results of the price analysis.  Another limitation presented by the authors is that the results presented represent only a small fraction of the overall appraisal of the role of climate change on vineyard and general agricultural values.  Finally, it’s possible that too high an increase in temperature would be detrimental on price, if the grapes were subsequently damaged by excessive heat.

I’d love to hear your thoughts on this topic.  Please feel free to comment below (no html tags, please).

Source: Ashenfelter, O., and Storchmann, K. 2010. Measuring the Economic Effect of Global Warming on Viticulture Using Auction, Retail, and Wholesale Prices. Review of Industrial Organization 37: 51-64.

DOI: 10.1007/s11151-010-9256-6
I am not a health professional, nor do I pretend to be. Please consult your doctor before altering your alcohol consumption habits. Do not consume alcohol if you are under the age of 21. Do not drink and drive. Enjoy responsibly!

Posted in Economics

Tagged , , , , ,