Taxing Trends: Tracing the TTB's stats on alcohol levels

A common refrain about wines in recent decades is that alcohol levels have been rising. Whether that statement is meant as lament or praise may vary. But either way, it's not often accompanied by hard data, so it's hard to judge the extent or accuracy of the statement.
   However, data from the federal Alcohol and Tobacco Tax and Trade Bureau (TTB) offers evidence for the debate. Wineries must pay a higher excise tax for wines above 14% than those below, a distinction that has been in place since at least 1951. The current rate, in effect since 1991, is $1.57 per gallon above 14% and $1.07 per gallon below. This translates to $0.31 and $0.21, respectively, per 750-ml bottle.
   Because of this tax differentiation, the TTB tracks the amount of wine in the higher category separately (this includes fortified wine as well as unfortified still wines above 14%). Their annual statistics, graphed above, show that in the mid- to late 1990s, this higher alcohol category consistently accounted for less than six percent of bottled still wines.
   However, a sharp escalation occurred in the last decade, as the amount of higher alcohol wine produced in the United States increased 50%, on a relative basis, to nine percent of bottled still wines.
   According to TTB regulations, the alcohol that appears on the label and what is actually in the bottle may differ by 1.5% in the lower bracket and one percent in the higher. Mari Kirrane, Wine Technical Trade Advisor at the TTB, says that there are no plans to tighten this tolerance in the foreseeable future. She also states that there are about 14 investigators nationally providing enforcement and inspection.
   Comparatively, the United Kingdom taxes wine quite heavily, with the costs borne by the consumer, not the winery. The excise duty hits the low-priced wines comparatively harder, thanks to a £1.69 per bottle excise tax for wines under 15% alcohol by volume (£2.25 for wines above 15%). The Budget 2010 calls for this tax to rise at two percent above the rate of inflation for the next four years. Wines are also subject to a 17.5 percent value added tax (VAT).
   France, by contrast, levies a duty on producers of €3.35 per one hundred liters. VAT is 19.6 percent.
   In the mid-1990s, the federal authorities regulating the wine industry famously disallowed the 1993 Mouton Rothschild label, redacting the art with a beige strip. They deemed the line drawing of a naked young woman too offensive for American sensibilities. Yet every year, they tolerate a number of factual inaccuracies on the label that may be more offensive to some.
   Chief among those is the alcohol level of wine on the label, which is allowed some wiggle room from what's actually in the bottle. Just how much the numbers actually deviate remains an open question.

   Some answers may come next year when the TTB releases the results of their first-ever compliance study of wines in the marketplace, which is currently underway.
   But we didn't want to wait. So here at W&S, we tested the alcohol levels of seven cases of wine, split equally between foreign and domestic. On the East Coast, we used the lab at City Winery in New York City; on the West Coast, Crushpad, a custom winery in Napa, ran our wines through their equipment. Our results were surprising.
   City Winery relies on an electric ebulliometer to determine alcohol levels. Developed about 100 years ago, the device tests the boiling point of a wine sample compared to that of distilled water, a process that controls for changes in barometric pressure. The boiling point is then converted on a slide rule to the level of alcohol by volume. David Lecomte, winemaker at City Winery, says the device isn't perfect (it's accurate within +/- 0.15% vol.), and will not deliver an accurate reading in wines with more than five grams of residual sugar, so we limited our samples to dry wines. Crushpad uses the Metrohm 855, a robotic titrator.
   Our analysis revealed that about half the wines tested were accurate within 0.3 of the alcohol stated on the label—a surprise given the commonly held view that "everybody lies" about alcohol levels. It's time to put such generalizations to bed.
   However, about one in ten of the wines strayed so far from its stated alcohol that it placed in the wrong tax bracket or violated the legal tolerance limit (1.5% deviation in the lower bracket—never crossing 14%—and one percent if over 14%). So given the high percentage in our analysis, the TTB may be leaving some revenue on the proverbial table.
   Moreover, the consumer who seeks accuracy may be disappointed—and get a buzz more quickly, whether wanted or not. Look at it this way: The TTB defines a serving of alcohol as 1.5 ounces of 80-proof alcohol; the US Department of Health equates this with 12 ounces of beer or 5 ounces of wine. The math works out if the wine is 12% alcohol—but if that glass of wine is 14.2%, it contains almost 20% more alcohol than that shot of scotch.
   While we didn't have enough data to discern many patterns, one did stand out. You might reasonably assume that wines from California would be the most likely to deviate from stated levels for at least two reasons: the generally hot climate and the fact that producers have to contend with only one set of national regulations, while imported wines may have to contend with more stringent standards in their home countries. In fact, California wines appeared to stay within the legal boundaries quite well.
   Lecomte, who oversees the approval of 250 labels a year at City Winery, says that label approval is costly and time-consuming and that many small wineries don't have in-house lab equipment. He adds that the TTB doesn't require a new label approval if the vintage and alcohol levels are the only changes on the label, and the wine remains in the same tax bracket.
   Our analyses here remain merely anecdotal, but the deviations we found make us all the more curious to see the results of the TTB study next year.
—Tyler Colman