Does belonging to an appellation make a difference? New evidence from Ontario Viticultural Areas

This article has been accepted for publication and undergone full peer review but has not been through the copyediting, typesetting, pagination and proofreading process, which may lead to differences between this version and the Version of Record. Please cite this article as: Gokcekus O. (2021), Does belonging to an appellation make a difference? New evidence from Ontario Viticultural Areas. Wine Economics and Policy, Just Accepted


Introduction
Until recently, Canadian wine was almost an oxymoron. Other than its infamous ice wine, even wellinformed wine drinkers were not aware of Canadian wines. The reason being that Canada resides out of conventional wine growing zone between 30 and 50 degrees latitude north and south of the equator; Canada was dismissed as a wine country. Yet, Canadians have been making wine for centuries. (Phillips, 2017). Moreover, recently their wine industry is growing at an accelerating rate. Since 1999, Canadian wine industry has been heavily regulated. (Carew and Florkowski, 2012.) For instance, Vintners Quality Alliance, or VQA Ontario Wine Appellation Authority is a regulatory agency responsible for "maintaining the integrity of local wine appellations and enforcing winemaking and labelling standards in Ontario." 1 As is shown in Figure 1, VQA Ontario divides Ontario into three primary Viticultural Areas or appellations of origin: Niagara Peninsula, Lake Erie North Shore, and Prince Edward County. Within the Niagara Peninsula appellation, ten sub-3 appellations are identified; four of them on the plains close to Lake Ontario (Four Mile Creek, Niagara Lake Shore, Niagara River, and St. David's Bench) and three on the bench lands of the Niagara Escarpment (Beamsville Bench, Short Hills Bench, Twenty Mile Bench). The other three subappellations are Creek Shores, Lincoln Lake Shore, and Vinemount Ridge. Thus, there are effectively four layers of appellations within the Ontario wine region. Similarly, as is argued in Gokcekus and Finnegan (2017, p. 345-346) Lake Erie -North Shore Prince Edwards County matters as a fundamental reality or solely economically due to the perceived reputation of a particular area. Nevertheless, the prices at which winemakers can sell their wines vary depending on the wines' geographic origins." (Matthews, 2016;Landon andSmith, 1997, Lecocq andVisser, 2006;Patterson and Buechsenstein, 2018).
In this study, we ask the following questions: Does it make a price difference whether a wine is coming from a particular Ontario appellation? In particular, is there a regional reputation premium attached to a particular appellation or sub-appellation? Moreover, does the premium for vintage, variety, and vinification change whether the regional differences are taken into account or not?

Data
For 4,213 table wines from Ontario wine region, between 2015-2018, we have information regarding their retail price, vintage, size, grape variety(ies), appellation, as well as vinificationwhether they are from a name vineyard or estate bottled. Table 1 provides summary statistics for these wines.

Calculations and Findings
To set a benchmark, first, as an ad-hoc regression model, the real price of each bottle of Ontario wine is set as a function of vintage, variety, and vinification 2 without taking appellations differences into account.
where, ln is natural logarithm operator; Vintage is the age of the wine; Varietyj is a dummy variable, where j = chardonnay, pinot noir, cabernet sauvignon, cabernet franc, riesling, sauvignon blanc, merlot, or pinot grigio; Vinification is a dummy variable, where k = single variety, estate bottled, named vineyard, or named vineyard and estate bottled; finally is a well behaving error term.
According to the robust regression results, as is summarized in column (2) of Table 2, there is a statistically significant relationship between real price and vintage, variety, and vinification: (1) Vintage matters; (2) there is a premium for red wines-pinot noir, cabernet sauvignon, cabernet 2 The wine must be made from at least 85% of the grape variety named to be considered a single variety. Estate Bottled are variations such as "Estate Grown" or "grown, produced and bottled by" are permitted if the wine qualifies for the estate bottled designation, close variations are not permitted for non-VQA wines. Vineyard (any named vineyard indicating origin is not permitted for non-qualifying VQA wines or non-VQA wines, two or more vineyards may not be named but general references to vineyards or multiple but unnamed vineyards are permitted).
6 franc, and merlot as well as a white wine, chardonnay; (3) making wines by using a single variety and estate or/and named vineyard bottling increases the price. Note: Significance levels (two-tailed) 1% (***), 5% (**), and 10% (*) Second, the real price of each bottle of Ontario wine is set as a function of vintage, variety, vinification as well as fifteen appellation/sub-appellations: Table 2 presents the robust regression results for this specification. These findings indicate that for Ontario wine prices 'terroir' matter. For example, compared to an Ontario wine with no particular appellation designation, a wine from Four Mile Creek acquires an additional 35% premium of which 19% is due to having its own sub-appellation, 7% for being under Niagaraon-the-Lake, and 9% for a Niagara Peninsula appellation. The size of the premium differences from one appellation to another is striking: It varies between 5% (Lake Erie North Shore) and 39% (St.

Column (3) of
David's Bench). These differences highlight the importance of appellation designations even in a relatively new and also small wine region of Ontario.
Another finding is about the size of the over-estimation regarding the premium attached to vintage, variety, and vinification, if we ignore appellation/sub-appellation differences. According to the regression results in Table 2-the difference between column (2) and column (3)-the estimated coefficients for vintage, variety and vinification variables were over-estimated by between 1% points and 18% points. For instance, premium attached to grape varieties pinot noir, chardonnay, and cabernet franc are 7.2%, 4.2%, and 4.1% points respectively were over-estimated unless appellation designations were taking into account. Similarly, regarding vinification, overestimations are 17.6%, 8.7%, 7.4%, and 6.8% points for name vineyard and estate, named vineyard, estate bottled, and single variety, respectively. These differences are not only statistically significant; indeed, they are economically significant too.

Discussion
To summarize, these findings 1) highlight the economically significant effect of terroir or a regional reputation even in a relatively new wine region; 2) indicate that ignoring the importance of terroir clearly could result in an overestimation of the premiums attached to different vintages, varieties, and vinification; and, moreover, 3) show that these premiums are not uniformly overestimated; there are variations among vintage, varieties as well vinification. Consequently, a wine maker should be paying attention to things that they can choose or control, but meanwhile they should keep in mind the location of their winery in order to set realistic expectations for the return on their investments and efforts.