Alteration of Class and Type: Vodka

DEPARTMENT OF THE TREASURY

Bureau of Alcohol, Tobacco, and Firearms

27 CFR Part 5

[T.D. ATF-360; Re: Notice Nos. 782, 780, 91F009P]

RIN: 1512-AB22

Alteration of Class and Type: Vodka

AGENCY: Bureau of Alcohol, Tobacco, and Firearms (ATF), Department of

the Treasury.

ACTION: Final rule, Treasury decision.

-----------------------------------------------------------------------

SUMMARY: ATF is amending the distilled spirits regulations, 27 CFR

5.23(a)(3), to authorize the use of a trace amount (defined as up to

300 milligrams per liter or 300 ppm) of citric acid in the production

of vodka, without changing its designation as vodka. This level is

intended to ensure that distillers may continue to use citric acid as a

smoothing agent to correct objectionable tastes which might result from

such things as the water used in reducing the proof, the charcoal used

in distillation, or the glass in which packaged. This level is also

intended to protect the integrity of the standard of identity for

vodka, a product, which by definition, may not have any distinctive

character, aroma, taste, or color. Pursuant to this document, T.D. ATF-

306 [55 FR 49994, December 4, 1990] is rescinded.

DATES: This document is effective on January 30, 1995.

FOR FURTHER INFORMATION CONTACT: David W. Brokaw, Wine and Beer Branch,

(202) 927-8230.

SUPPLEMENTARY INFORMATION:

Background

The standard of identity for vodka was promulgated in 1949 in TD

5707, 1949-2 C.B. 252. The hearing record indicates that a standard of

identity for vodka was proposed because many different types and

qualities of vodka were being offered for sale in the United States.

With the proliferation of vodka products, the Alcohol Tax Unit of the

Internal Revenue Service (IRS), a predecessor to ATF, was concerned

that, consistent with the mandate of section 105(e) of the Federal

Alcohol Administration Act (FAA Act), 27 U.S.C. 205(e), consumers were

adequately informed regarding the identity and quality of such

products. The standard for vodka provided that it was to be defined as

neutral spirits distilled from any material at or above 190 proof,

reduced to not more than 110 proof and not less than 80 proof and,

after such reduction in proof, so treated as to be without distinctive

character, aroma, or taste. Although no explicit definition of the term

``distinctive'' could be found in the hearing record, the testimony

indicates that vodka is to be as tasteless and odorless as possible.

In 1956, Revenue Ruling 56-98, 1956-1 C.B. 811, concluded that

citric acid and sugar were not considered to be flavoring ingredients

which would materially affect the taste of vodka or change its basic

character so long as the sugar did not exceed two-tenths of 1 percent

and citric acid was only added in a ``trace amount.'' The ruling did

not quantify ``trace amount.''

Revenue Ruling 56-98 was issued after testing of a vodka sample

prepared by the IRS' Alcohol and Tobacco Laboratory pursuant to a

formula submission which contained two-tenths of 1 percent of sugar and

13 ppm of citric acid. The Alcohol and Tobacco Laboratory stated that

the addition of such amounts of sugar and citric acid were advantageous

from the consumer-standpoint, in that such addition would make a

smoother and more neutral product. The Alcohol and Tobacco Laboratory

further stated that the use of two-tenths of 1 percent of sugar and 13

ppm of citric acid would in no way render vodka flavored but, rather,

would correct objectionable tastes which might be obtained from the

water used in reducing the proof, the charcoal used in distillation, or

the glass in which packaged. The Alcohol and Tobacco Laboratory

concluded that the use of sugar and citric acid in the amounts stated

was in conformance with the standard of identity for vodka in that such

amounts were not considered to be flavoring materials and would not

materially affect the taste of the product, or change its basic

character.

The Alcohol and Tobacco Laboratory therefore stated that the use of

these ingredients in the amounts prescribed would conform with a method

of production of vodka outlined in T.D. 5707, supra, which stated that

vodka could be produced ``[b]y purifying or refining the distillate by

any other method which the Deputy Commissioner finds will result in a

product equally without distinctive character, aroma, or taste, and

which has been approved by him.'' Simply stated, the addition of up to

two-tenths of 1 percent of sugar and ``trace amounts'' of citric acid

was allowed because those ingredients act as a smoothing agent which

results in rendering vodka without distinctive character, aroma, taste,

or color. ``Trace amounts'' was not quantified in view of the fact that

the Alcohol and Tobacco Laboratory recognized that the appropriate

amount of citric acid would vary on a case by case basis depending on

the extent of the treatment used to neutralize the product.

The issue of citric acid took on a new dimension in 1980, when 26

U.S.C. 5010 was enacted. The section 5010 credit created an incentive

for industry to add additional amounts of flavors to distilled spirits

products for the purpose of reducing the effective rate of tax. Under

current law, distilled spirits products are taxed at the rate of $13.50

a proof gallon, with a credit under section 5010 for the flavor

content. The credit for alcohol derived from flavors is equal to the

distilled spirits tax, but only 2 1/2 percent of the alcohol content of

a distilled spirits product qualifies for the credit. The net effect of

this credit is to reduce the tax rate applicable to such product.

Vodka producers apparently began to use citric acid as a method to

receive the credit. The ATF Laboratory has stated that 1,000 ppm

represents, as a practical matter, the amount of citric acid on which

the maximum section 5010 credit may be claimed. Moreover, the use of

1,000 ppm of citric acid reduces the effective rate of tax on a proof

gallon of neutral spirits by 34 cents. Such a reduction can result in a

tax benefit of hundreds of thousands, if not millions of dollars a year

for the vodka industry.

Notice No. 403

With the increasing use of citric acid, the ATF Laboratory tested

various vodka samples produced in accordance with Revenue Ruling 56-98

and found that vodka which contained sugar and citric acid within

levels authorized by the ruling had a measurable solids content due to

the presence of sugar, and displayed a change in the titratable acidity

due to the presence of citric acid. The ATF Laboratory stated that such

changes in the physical and chemical characteristics of the samples

resulted in a vodka having impermissible distinctive character even if

an organoleptic examination indicated no distinctive character, aroma,

taste, or color. Based on the conclusions of the ATF Laboratory with

respect to sugar and citric acid, ATF subsequently issued an Advance

Notice of Proposed Rulemaking (ANPRM) on January 11, 1982, Notice No.

403 (47 FR 1148), requesting comments on whether ATF should revoke

Revenue Ruling 56-98 and prohibit the treatment of vodka with sugar and

citric acid, or whether ATF should establish a separate class and type

of vodka containing specified quantities of sugar and citric acid.

Additionally, the ANPRM noted that a conflict seemed to exist between

the provisions of Revenue Ruling 56-98 authorizing treatment of vodka

with sugar and citric acid, and 27 CFR 5.23(a)(3) which prohibits any

addition of harmless coloring, flavoring, and blending materials to

neutral spirits. The ANPRM, therefore, requested comments on clarifying

the standard of identity for vodka to preclude the addition of any

ingredients to vodka, such as sugar and citric acid, that would

materially affect its chemical and physical characteristics and

comments on establishing another class of vodkas authorizing the

addition of sugar and citric acid within specified amounts.

Comments on Notice No. 403

ATF received 16 comments pursuant to the ANPRM. Eight respondents

opposed the addition of citric acid and sugar to vodka, and favored

revoking Revenue Ruling 56-98 for the following reasons: It is

inconsistent with 27 CFR 5.23(a)(3), the public perceives vodka as a

pure material spirit, there is no purpose to the addition of sugar and

citric acid, and allowing sugar and citric acid will allow other

materials to be added to vodka. Seven respondents favored permitting

the use of sugar and citric acid in the production of vodka by

retaining Revenue Ruling 56-98. These comments indicated mainly that

the standard of identity should be based on sensory factors such as

taste, aroma, and color, and not on physical and chemical

characteristics because consumers judge vodka by taste. These comments

further stated that consumers have not been harmed or deceived by the

addition of sugar and citric acid and would not benefit by a

prohibition on such ingredients. In addition, eight of the sixteen

respondents objected to the proposed establishment of a new class and

type of vodka containing sugar and citric acid for various reasons.

Notice No. 583

Based on the comments on Notice No. 403, ATF issued an NPRM on

February 19, 1986, Notice No. 583 (51 FR 6009), proposing to revoke

Revenue Ruling 56-98, incorporate its provisions into 27 CFR 5.23, and

define a ``trace amount'' as 150 ppm. The NPRM noted that trace amounts

varied widely with different formulas. The notice explained that 150

ppm was being proposed because it is sufficient to neutralize residual

alkalinity derived from charcoal treatment of some vodkas, or from the

use of certain glass in manufacturing bottles. The NPRM sought comments

on the appropriate level of citric acid, both above and below 150 ppm.

Comments on Notice No. 583

ATF received 10 new comments pursuant to the NPRM. Five commenters

opposed the addition of citric acid and sugar to vodka even though ATF

had previously decided to continue to allow the use of sugar and citric

acid in vodka based on comments received pursuant to the ANPRM. The

remaining commenters favored allowing the use of sugar and citric acid

so as not to interfere with eligibility for the section 5010 credit.

The comment received from the LeVecke Corporation included the results

of organoleptic tests which showed that the test panelists could

consistently detect the presence of citric acid at 400 ppm.

Accordingly, the company recommended that citric acid be allowed up to

400 ppm.

T.D. ATF-306

On December 4, 1990, ATF issued a Final Rule, T.D. (T.D. ATF-306,

55 FR 49994), that vodka with more than 150 ppm of citric acid has a

distinctive character, in violation of the standard of identity, and

must be labeled as flavored vodka. T.D. ATF-306, amended 27 CFR

5.23(a)(3) to authorize the use of up to 2 grams per liter (2,000 parts

per million) of sugar, and a trace amount (defined as 150 milligrams

per liter or 150 parts per million) of citric acid in the production of

vodka. T.D. ATF-306 was effective January 3, 1991, with a formula and

label cancellation date of March 4, 1991, for products not made within

the limitations of the Treasury decision.

The TD incorporated the comment received by the LeVecke Corporation

which showed that the test panelists could consistently detect the

presence of citric acid at 400 ppm. The TD states that ATF agrees that

the presence of citric acid is detectable at 400 ppm but that ``ATF's

Laboratory conducted similar tests and found that 150 ppm is

approximately the threshold at which the presence of citric acid in

vodka can be detected.'' The TD, therefore, concluded that any amount

of citric acid higher than 150 ppm was distinctive in violation of 27

CFR 5.22(a)(1) based on the position that citric acid, as with any

other ingredient, should not be present at a level at which it is

detectable to a consumer, irrespective of whether the consumer can

specifically identify the ingredient as citric acid.

Petition for Reconsideration

On March 4, 1991, in response to a petition from Heublein, Inc.,

(Heublein) for the reconsideration of T.D. ATF-306, ATF issued T.D.

ATF-311 (56 FR 8922). T.D. ATF-311 deferred the compliance date with

respect to the citric acid limitation set forth in 27 CFR

5.23(a)(3)(ii) by T.D. ATF-306 to December 4, 1991. Heublein's petition

was based on a representation that new scientific information and data

not previously available had come to their attention concerning maximum

levels for the use of citric acid in vodka. The compliance date set

forth in T.D. ATF-306 regarding a maximum level for the use of sugar in

vodka remained unchanged.

Notice No. 716

On April 29, 1991, ATF issued Notice No. 716 (56 FR 19623) to

gather additional information with respect to the 150 ppm limitation

set forth in T.D. ATF-306. Notice No. 716 was issued in conjunction

with Heublein's request for reconsideration which was granted in T.D.

ATF-311. Notice No. 716 proposed to retain the reasoning and conclusion

of T.D. ATF-306 as set forth above. The purpose of the Notice was to

solicit comments from the public and industry concerning the

appropriateness of sensory threshold citric acid levels higher and

lower than the proposed maximum level of 150 ppm.

Comments on Notice No. 716

In response to Notice No. 716, ATF received nine comments. All of

the comments were opposed to setting a maximum limitation as low as 150

ppm for the addition of citric acid to vodka. However, the only

commenter submitting substantiating test data was Heublein. In January

1991 and July 1991, Heublein submitted results of sensory (taste and

smell) testing which it claimed demonstrate that the majority of

consumers could not detect a difference between vodka containing 150

ppm and vodka containing 480 ppm. The purpose of these tests was not to

determine the level at which citric acid can be detected by the

consumer, but rather whether a statistically significant difference

could be detected between two levels of citric acid in vodka. For

example, if a panel were asked to determine whether there was a

difference between 1,000 ppm and 1,110 ppm of citric acid, the

statistical analysis of the results of such testing would only show

whether the two samples can be reliably distinguished but would not

generate any information concerning the threshold level at which citric

acid could be detected.

Heublein relied on these tests to support the conclusion that 480

ppm is not detectable. That is, if vodka at 480 ppm tastes, smells, and

looks no different than 150 ppm, then it is no more distinctive.

Heublein also tested vodka containing 528 ppm of citric acid and stated

that several of its tests showed significant and perceptible difference

between vodka containing 150 ppm of citric acid and vodka with 528 ppm

citric acid. Heublein further stated that changing the formulation of

the product from 528 ppm to 150 ppm would ``significantly change the

sensory character'' of their product. The ATF Laboratory interpreted

Heublein's tests as concluding that, while the majority of consumers

could not detect a change in the level of citric acid from 150 ppm to

480 ppm, such a change could be detected from 150 ppm to 528 ppm. As

discussed later, ATF does not believe that these tests can be relied

upon in setting a citric acid limitation.

Independent Testing

During the comment period, ATF secured an outside testing firm,

Odor Science and Engineering (OS&E), to conduct independent testing on

sensory threshold levels for citric acid addition to vodka. The outside

testing firm was secured in order to provide data for comparison to the

results secured by ATF and industry members.

OS&E was requested to conduct a test which would produce results

with respect to the taste threshold of citric acid in vodka. The

initial tests were conducted on April 25 and 29, 1991. Various

concentrations of ethanol solution and citric acid were presented to a

taste panel comprised of ten experienced sensory panelists familiarized

with the taste of the ethanol solutions. The initial ethanol solutions

were supplied to OS&E by ATF and consisted of 40 percent ethanol in

water (80 proof). The flasks were stoppered to prevent any opportunity

of interference because of odor perception. The initial levels of

citric acid in the various ethanol solutions were: 50 ppm, 100 ppm, 150

ppm, 200 ppm, and 1,000 ppm.

The presentations consisted of two flasks containing only ethanol

solution and one flask containing ethanol solution mixed with one of

the levels of citric acid described above. One panelist at a time

entered the sensory laboratory and tasted each of the three flasks.

Each of the panelists tasted each of the three flasks and indicated to

the panel moderator which of the three samples was different, thus

requiring the panelist to guess if he or she expressed uncertainty.

This procedure is known as the forced-choice ascending concentration

method and was repeated for each of the citric acid concentrations

described above. The forced-choice ascending method is stacked against

guessing which sample contains the citric acid because two out of three

choices do not contain citric acid.

As results of the initial trials were inconclusive, repeat trials

were conducted with eleven panelists using citric acid concentrations

of 100 ppm, 250 ppm, and 500 ppm. However, as with the initial tests,

the ethanol solution had a pronounced anesthetic effect which precluded

accurate determinations of taste thresholds. As a result, the ethanol

solution was diluted to 20 percent ethanol in water (40 proof) as

directed by American Society for Testing and Materials he American

Society for Testing and Materials (ASTM) methodology and three

concentrations of citric acid in the diluted solution were evaluated:

25 ppm, 50 ppm, and 500 ppm. The anesthetic effect discussed above was

presumably eliminated because all of the panelists detected taste

differences at lower concentrations of citric acid.

An additional test was run on May 28, 1991, using twelve series (2

by each panelist) in order to develop more representative data using a

20 percent ethanol in water (40 proof) solution. The concentrations of

citric acid in the diluted solution were 25 ppm, 50 ppm, 100 ppm, 200

ppm, and 400 ppm. The threshold of citric acid in vodka could then be

determined using the forced-choice ascending concentration method. OS&E

concluded that a majority of people could detect the presence of citric

acid in vodka at a level of approximately 180 ppm.

However, such test results collected by OS&E were not calculated in

accordance with the methodology prescribed by the ASTM. ASTM Procedure

E-679 is entitled ``Standard Practice for Determination of Odor and

Taste Thresholds By a Forced-Choice Ascending Concentration Series

Method of Limits.'' This method requires use of a geometric progression

of concentrations, i.e., each concentration of citric acid is a factor

or multiple of the previous concentration. In this test, the factor was

two. In order to calculate the mean of such a progression, one must

take the geometric mean. This is calculated by taking the nth root of

the product of the numbers. This differs from the familiar arithmetic

mean which is found by dividing the sum of the numbers by n. In both

cases, n is the number of values under consideration.

ATF Analysis of Independent Test Data

The OS&E tests were supported by the ATF Laboratory in lieu of the

tests relied on in setting the 150 ppm limit in T.D. ATF-306 because

those earlier tests did not fully conform to the methodology

established for sensory threshold testing established by the ASTM.

However, although the OS&E tests were conducted in accordance with the

ASTM methodology, OS&E did not calculate results from the data it

generated in the ASTM prescribed method. The ATF Laboratory therefore

used the data generated by OS&E and recalculated the results according

to the approved ASTM method. The detection threshold of citric acid in

vodka was determined using the best-estimate criterion, or the group

geometric mean of all the panelists' thresholds. Each panelist's

threshold is the geometric mean of the last missed concentration (the

last concentration of citric acid not detected) and the next higher

concentration. This kind of analysis is considered most reliable when

most of the panelists have had at least two hits in a row at the high

concentration end. For those panelists who have a miss at the highest

concentration, it is assumed by the ASTM method that they would have

had a hit at the next higher concentration had there been one more

sample in the series (800 ppm citric acid for this test). For those

panelists who have no misses, it is assumed by the ASTM method that,

had the testing begun one sample lower in the series (12.5 ppm citric

acid for this test), they would have missed it.

For normally distributed data, half of the population lies on each

side of the center of the data curve. Using the OS&E data, the ATF

laboratory determined that two-thirds of the population lies within the

range of citric acid concentrations between 202 ppm and 700 ppm and

one-sixth lies outside this region in each tail of the distribution. A

normal distribution did not exist because the data were skewed toward

lower concentrations of citric acid. Because of this, the geometric

mean (which is calculated from the average of the log of the values)

would not yield as accurate a measure of where the majority of people

would detect the presence of citric acid in vodka as the median, or

middle point, of the distribution ordered from lowest point to highest

point.

The ATF Laboratory also utilized the OS&E data to determine that

the actual range of concentrations of citric acid that would include

two-thirds of the population results in a geometric mean of 376 ppm and

a range of 202 ppm to 700 ppm at 40 percent alcohol or 80 proof. The

results of the OS&E testing also show that a majority of the panelists

were able to detect citric acid at the median level of 282 ppm in 80

proof vodka. That is, the OS&E data revealed that 282 ppm was the

initial point of detection for the majority of panelists. The median

level of 282 ppm is lower than the group geometric mean of 376 ppm

because the data are skewed away from a normal distribution toward

lower concentrations of citric acid.

The sample size of the OS&E study was sufficient to be

statistically valid based on the ASTM method in place at the time of

the testing. As is always the case, a larger sample size would have

certainly given a distribution more closely representative of the total

population. ATF has allowed for the uncertainty inherent in any

determination where the entire population is not sampled. The standard

of identity for vodka would prescribe that the lower limit of

detectability be used to ensure that a majority of the population could

not detect the presence of citric acid.

Disparity in Test Results

Heublein was provided with the opportunity to comment on the data

secured on behalf of ATF by OS&E. Heublein acknowledged the

discrepancies in the OS&E report in their comments to ATF when they

stated that ``we were unable to reproduce some * * * values in the

report.'' Nevertheless, Heublein used the incorrect OS&E calculations

and represented to ATF that the level at which the majority of people

could detect citric acid in vodka is 734 ppm. This level is much higher

than the 282 ppm level generated by ATF from the OS&E data because the

value of 734 ppm was erroneously based on the calculation of the

arithmetic mean.

The citric acid level of 734 ppm proposed by Heublein would be the

concentration at which greater than two-thirds of the population would

detect the presence of citric acid in vodka. In addition, the tests

submitted by Heublein were designed to determine whether there is a

statistically significant difference between two solutions of different

concentrations of citric acid in vodka, not the threshold level of

citric acid. Accordingly, ATF does not believe that the tests submitted

by Heublein are germane to the issue of setting a level of citric acid

that will not be in conflict with the current standard of identity for

vodka which defines the product as being without distinctive character,

aroma, taste, or color. Indeed, ATF believes that the use of a level of

citric acid which greater than two-thirds of the population could

detect is inconsistent with the above-stated standard of identity for

vodka. ATF believes instead that it is in the interest of preserving

the standard of identity for vodka that any statistical treatment be

applied in the opposite direction, i.e., where most of the population

would not detect the presence of citric acid.

Deferrals of Compliance Date

A significant disparity existed between the methodologies of the

studies done by ATF and Heublein, and the conclusions reached by

Heublein and OS&E with respect to the study conducted by OS&E. Based on

these disparities, ATF concluded that more time would be needed to

properly evaluate the sensory tests and results derived from Heublein

and the outside firm hired by ATF. The compliance date of December 4,

1991, with respect to citric acid, set forth in T.D. ATF-311 was

subsequently deferred to September 3, 1992, by T.D. ATF-319 (56 FR

63398, December 3, 1991) in order to allow for time to resolve the

disparity in the test results. As ATF had not yet completed its review

of all data submitted relative to the citric acid limitation as of

September 3, 1992, ATF issued T.D. ATF-333 (57 FR 40323, September 3,

1992), which deferred the compliance date with respect to the citric

acid limitation set forth in 27 CFR. 5.23(a)(3)(ii) until September 3,

1993. On August 27, 1993, ATF issued T.D. ATF-348 (58 FR 45251)

deferring the compliance date set forth in T.D. ATF-333 until August

28, 1995, with respect to the citric acid limitation set forth in 27

CFR 5.23(a)(3)(ii) by T.D. ATF-306. The two year compliance date

deferral was necessary in order to allow ATF to analyze the comments

received during the comment period relative to the issuance of Notice

No. 780 and to allow for additional unforeseen contingencies.

Notice No. 780

Upon consideration of both the OS&E and Heublein tests, ATF

published Notice No. 780 (58 FR 46141) on September 1, 1993, proposing

to amend the regulations authorizing the use of a trace amount (defined

as up to 300 milligrams per liter or 300 ppm) of citric acid in the

production of vodka, without changing its designation as vodka. The

NPRM was issued in order to provide the public with an opportunity to

comment on the OS&E study. Notice No. 780 contains the detailed

description and analysis of the OS&E study and the critique of the

Heublein tests submitted in response to Notice No. 716 set forth above.

The NPRM proposed a level of 300 ppm as a rounding off of the 282 ppm

level derived by OS&E in order to facilitate regulatory administration

and consumer understanding. Because citric acid is not an essential

component of vodka, the NPRM proposed amending 27 CFR. 5.23 which

regulates additions of substances to distilled spirits, rather than 27

CFR 5.22(a)(1) which is the standard of identity of vodka. Under this

proposal, vodka made with a greater concentration of citric acid would

be designated ``flavored vodka'' or labeled with a fanciful name under

27 CFR part 5.

The comment period for Notice No. 780 ended on October 18, 1993. At

that time, Notice No. 782 [58 FR 53682] was published in the Federal

Register extending the comment period for Notice No. 780 until January

3, 1994. Notice No. 782 was issued in response to a request from the

Distilled Spirits Council of the United States (DISCUS) for an

extension of the comment period in order to provide sufficient time for

all interested parties to respond to the issues addressed in the NPRM.

Comments on Notice No. 780

In response to Notice No. 780, 8 comments were received. Seven of

the commenters felt that the maximum authorized level for the addition

of citric acid to vodka should be 1,000 ppm without changing the

product's designation as vodka. The remaining commenter, Mr. David A.

Owen, did not object to the proposed 300 ppm level but felt that the

level of all such ingredients should be indicated on the label.

Three of the commenters, the President's Forum of the Beverage

Alcohol Industry, the David Sherman Corporation, and Barton

Incorporated, while opposed to setting a maximum limitation as low as

300 ppm, did not submit any substantiating data to support their

recommended level of 1,000 ppm.

Jim Beam Brands Co., another of the seven commenters supporting

1,000 ppm, felt that consumer response to vodka produced with citric

acid is the most important issue. Jim Beam pointed out that the actual

presence of citric acid has not raised any health or safety issues, nor

has it resulted in any consumer deception since being introduced in

vodka production in 1956. Rather, Jim Beam argued that consumers would

be confused by requiring vodka containing greater than 300 ppm to be

labeled as a ``flavored vodka'' when such products traditionally have

been marketed as ``vodka.'' Jim Beam notes that in decreasing the

allowable amount of citric acid, ATF must depend on valid scientific

studies which demonstrate that consumers can reliably detect a

distinctive difference between vodkas at or above that threshold and

vodkas below that threshold. Jim Beam also includes in its comment a

critique of the OS&E study which Beam believes was neither reliable nor

valid. As such, Jim Beam believes that ATF is fulfilling its statutory

mandate in this matter without further action.

The comment submitted by United Distillers also criticizes the OS&E

study. In addition, United Distillers argues that there are detectable

differences even amongst vodkas that are untreated. These differences

are said to result from ingredients, manufacturing processes, and the

subjective tastes of the consumer. Therefore, based on the historical

acceptance of vodkas with up to 1,000 ppm of citric acid, international

practices, and product standardization, United Distillers supports a

level of 1,000 ppm.

Heublein argues in its comment that the OS&E study was unreliable.

Heublein also states that ATF should not determine a minimum threshold

level since they believe that it has no relevance to the

distinctiveness requirements of the vodka standard of identity.

Similarly, Heublein states that the test results it submitted in

response to Notice No. 716 demonstrate the insufficiency of the minimum

threshold standard and the appropriateness of distinctiveness standard.

Heublein argues that ATF misinterpreted the test results it submitted

in connection with Notice No. 716. Heublein maintains that those tests

do in fact demonstrate that vodkas containing 480 ppm and 528 ppm are

not distinguishable from vodka containing 150 ppm. Heublein further

argues that ``the effort being expended to justify an additive limit

below 1,000 ppm is in clear conflict with the instruction of the

President to ATF and other agencies,'' as stated in the President's

Executive Order on Regulatory Planning and Review dated September 30,

1993 (58 FR 51735 (Oct. 4, 1993)). The President stated, in part, that

Federal agencies should promulgate only such regulations as are

required by law, are necessary to interpret the law, or are made

necessary by compelling public need, such as the material failure of

private markets to protect or improve the health and safety of the

public. The President further stated that, in deciding whether and how

to regulate, agencies should assess all costs and benefits of available

regulatory alternatives, including the alternative of not regulating.

Heublein therefore asserts that since there is no compelling public

health or safety reason for the change proposed, ATF should allow up to

1,000 ppm. Heublein supports its conclusion by arguing that no reliable

evidence exists in support of 300 ppm, and that it has provided

evidence in support of 1,000 ppm,

Finally, DISCUS submitted substantive comments which were supported

by Heublein in its comment. Initially, DISCUS reiterates the argument

that a maximum level of citric acid can be based only on distinctive

differences between vodkas with levels of citric acid at or above a

specific level and vodkas without citric acid. DISCUS also argues that

ATF's longstanding approval of formulas containing up to 1,000 ppm and

the confusion that would result from the adoption of 300 ppm militate

toward allowing 1,000 ppm.

In support of its conclusions, DISCUS issued a critique of the OS&E

study. It criticized OS&E for, in general, using too few test subjects,

failing to verify the qualifications of the testers that it did use,

apparently using the study authors as test subjects, collecting too few

test responses, and failing to use proper testing procedures. DISCUS

states that these criticisms demonstrate the unreliability of the OS&E

study.

DISCUS also submitted the results of an independent test conducted

by the Tragon Corporation (a sensory testing and marketing research

company) which purports to demonstrate that vodka containing up to

1,000 ppm has no distinctive character, aroma, taste, or color. This

test was not ``threshold testing'' in which the minimum amount of

citric acid which could be detected by a test subject is determined

but, rather, ``difference testing'' in which a test subject is asked

whether a difference can be detected between vodka containing 0 ppm,

500 ppm, and 1,000 ppm of citric acid. This test also inquired as to

whether test subjects could ascribe any characteristics of citric acid,

such as tartness, to any of the test samples and concluded that such

characteristics could not be identified. These tests were characterized

as ``reliable scientific evidence'' in Heublein's and DISCUS' comments

and both commenters now urge ATF to set a limit of 1,000 ppm.

The substantive comments on Notice No. 780 related in large part to

the relative merits of the OS&E and DISCUS studies. With respect to the

test data submitted by DISCUS, ATF has several criticisms which serve

as the basis to discount its conclusions. Initially, ATF does not

believe that DISCUS' study conforms with ASTM procedures. For example,

the DISCUS study utilizes commercially available vodkas which contain

80 proof alcohol while the ASTM procedures require that distilled

spirits used in testing be diluted to 25 proof or 45 proof. In

addition, ATF believes that the inclusion of a flavored vodka and a

vodka with two tenths of 1 percent sugar added in the testing samples

cast doubt on the validity of the study's conclusions. Most

importantly, ATF believes that the DISCUS study improperly involved

``difference testing'' while the proper focus should be ``threshold

testing'' as those terms are described above. With respect to the

criticisms of DISCUS and all other commenters of the OS&E study, ATF

believes that the study conformed with ASTM procedures. For example,

although the size of a testing panel and the number of test responses

collected could always be larger, both were within acceptable ASTM

procedures. Moreover, ATF ensured that the data collected by OS&E was

calculated in accordance with ASTM procedures. As such, ATF believes

that the criticisms of the OS&E study by seven of the commenters on

Notice No. 780 do not diminish the validity of the study.

Based on all of the above, ATF believes that its reliance on the

OS&E study in reaching a conclusion in this matter is justified. In

that regard, no commenter during the course of this rulemaking

submitted a study which both involved threshold testing and conformed

with ASTM procedures. As the OS&E study did meet these criteria, and

its results are scientifically valid and reliable, it is reasonable to

utilize the data from that study rather than data submitted by

commenters that are in favor of 1,000 ppm.

The other principal contentions of the comments on Notice No. 780

are that the acceptable level of citric acid should be based on

distinctiveness rather than on detectability and that, in essence,

there is no reason not to allow 1,000 ppm because that level represents

the status quo regarding citric acid. ATF believes that detection is

the appropriate standard because the standard of identity for vodka

states that vodka should be without distinctive character, aroma,

taste, or color. If an ingredient is present in an amount at which it

can be distinguished from other ingredients in the product, then its

presence contravenes rather than supports the standard of identity.

Indeed, the ingredient in question would itself become distinctive in

violation of the standard of identity. With respect to the argument in

favor of maintaining the status quo, ATF believes that its statutory

mandate to protect consumers is best served by adhering as closely as

possible to the standard of identity. Moreover, ATF does not believe

that consumers would be misled or confused by the setting of a citric

acid limitation. Indeed, consumers receive no information as to the

level of citric acid in vodka products. Thus, the setting of a citric

acid limitation, which will require that vodkas above 300 ppm be

labeled as ``flavored'' while vodkas below that level retain the

designation ``vodka,'' will prohibit deception in that consumers will

be able to relate such terms to the level of citric acid contained in

vodkas. Consumers will, therefore, be better informed as to the

identity and quality of the products in question. ATF believes that the

preservation of the standard of identity is best served by ensuring a

change in labeling for vodkas over 300 ppm. As such, ATF believes that

its regulation of citric acid is in accordance with the Executive Order

on Regulatory Planning, and Review because it is a necessary

interpretation of the law. Stated another way, ATF believes that

regulating citric acid is necessary to maintain both the original

intent and integrity of the standard of identity for vodka.

Discussion

Any discussion of a citric acid limitation must begin with the

premise that the standard of identity for vodka contemplates a neutral

product. The addition of any ingredient to such neutral spirits must

not only be clearly justified, but must not contravene the standard of

identity for vodka by contributing distinctiveness to the product. No

comments since the inception of this rulemaking have offered such a

justification for the use of more than 300 ppm of citric acid.

In that regard, ATF has historically maintained that the use of

citric acid should be for the purpose of correcting deficiencies in

vodka; i.e., acting as a smoothing agent to correct objectionable

tastes which might be obtained from the water used in reducing the

proof, the charcoal used in distillation, or the glass in which

packaged. ATF has viewed vodka as a product which, in essence, was

neutral in character, aroma, taste, and color for forty-five years;

although this policy has been deferred during the pendency of the

rulemaking. While sugar and citric acid are recognized as acting as

smoothing agents in vodka, it was very clear that they were authorized

only for the purpose of correcting objectionable tastes which might

result from such things as the water used in reducing the proof, or

from the glass in which packaged. The ATF Laboratory has further stated

that it believes that such use of citric acid should not be detectable

to a majority of consumers. ATF therefore takes the position that the

addition of citric acid to a level equal to or greater than that of

detection would contribute a distinctive character to the product in

violation of the standard of identity.

ATF is fulfilling its statutory mandate in 27 U.S.C. 205(e) by

ensuring adherence to the standard of identity for vodka. That is, ATF

is setting a citric acid limitation in order to ensure that consumers

are adequately informed as to identity and quality of vodka; a product

which is understood to be without distinctive character, aroma, taste,

or color. The addition of sugar, citric acid, or any other ingredient

for that matter, has and should be allowed only to facilitate rendering

vodka without distinctive character, aroma, taste, or color. Therefore,

there is a strong presumption against the use of any ingredients, in

any amount, particularly where they are likely to contribute to

distinctiveness rather than neutrality. Moreover, the point at which

the amount of such an ingredient begins to contravene the standard of

identity, as in the point at which citric acid can be detected by a

majority of consumers, is the maximum allowable level of that

ingredient in vodka.

ATF therefore believes that the rulemaking record in this case

supports a level of 300 ppm of citric acid in vodka without changing

the product designation. The current standard of identity mandates

consideration of not only differences attributable to taste, but also

differences attributable to character, aroma, or color. To that end,

the ATF Laboratory has conducted tests based upon chemical and physical

characteristics and designed the OS&E tests based upon organoleptic

factors, all of which lend support to a level of 300 ppm.

This position is clearly supported not only by an examination of

the history of 27 CFR 5.22(a)(1), but also by 27 CFR 5.23(a)(1), which

provides that no material whatsoever may be added to neutral spirits

(which encompasses vodka). ATF believes that this section clearly

reflects the original intent of the regulations in that vodka was

presumed to be a product which was not altered in any way by the

addition of any material, and would, therefore, be as tasteless and

odorless as possible. Thus, sugar and citric acid were recognized as

acting as smoothing agents in vodka, which were authorized only for the

purpose of correcting objectionable tastes which might result from such

things as the water used in reducing the proof, the charcoal used in

distillation, or the glass in which packaged.

In light of this purpose, ATF believes that this interpretation

correctly applies the standard of identity for vodka since the level of

detection to a majority of consumers would demonstrate that the amount

of citric acid used reached a point with regard to the product that was

no longer viewed only as corrective. That is, the addition of citric

acid to a level equal to or greater than that of detection would

contribute a distinctive character to the product in violation of the

standard of identity.

It is apparent that the reason for the increased use of citric acid

in vodka has been directly tied to the tax credit savings available by

virtue of the 26 U.S.C. 5010 wine/flavor credit enacted in 1980. While

ATF recognizes that such a rationale is not, in and of itself, a basis

for denial of citric acid in vodka production, it certainly can be used

to determine whether the amounts that are currently used reflect the

levels which were initially contemplated by Revenue Ruling 56-98. That

is, the inquiry becomes whether the level of citric acid proposed by

vodka producers whose comments support the use of greater than 300 ppm

is necessary to correct objectionable tastes, etc., in vodka or whether

such levels are merely a tax saving mechanism used to reduce the

effective tax rate on vodka products. Commenters supporting the use of

greater than 300 ppm have offered no rationale for the use of citric

acid in amounts up to 1,000 ppm since the initiation of this rulemaking

effort. To the contrary, it is apparent that the use of up to 1,000 ppm

of citric acid is directly related to the section 5010 flavors credit.

Absent such a rationale, ATF is unable to conclude that the addition of

that quantity of citric acid serves a corrective purpose in the

production of vodka. Even assuming it has such a purpose, the

detectability of 1,000 ppm would preclude the authorization of such a

level. As such, 27 CFR 5.23(a)(1) precludes, as it would the addition

of any other material whatsoever, the addition of greater than 300 ppm

of citric acid.

ATF recognizes in setting a limit of 300 ppm that formulas

containing up to 1,000 ppm have been approved for several years.

Rulemaking with respect to this issue was initiated in January 1982.

ATF has therefore allowed the use of up to 1,000 ppm during the

pendency of the rulemaking. Vodka producers began submitting formulas,

for the most part, after the enactment of section 5010 and ATF did not

believe that it was equitable to deny such formulas until an ultimate

decision was reached in this matter. Moreover, ATF has only now

obtained reliable scientific data in support of 300 ppm, allowed

interested parties the opportunity to comment on this data, evaluated

the comments on such data, and reached a final conclusion in this

matter.

ATF notes that two options exist with respect to vodka products

containing more than 300 ppm of citric acid that are currently in the

marketplace. First, such products may continue to be produced in their

current form and relabeled as ``flavored vodka'' or labeled with a

fanciful name, followed by a truthful and adequate statement of

composition under 27 CFR part 5. In addition, the level of citric acid

can be reduced to 300 ppm and thereby retain the designation ``vodka.''

ATF does not believe that such a reduction is unreasonable in light of

the fact that the use of citric acid in vodka must serve a purpose

relative to maintaining the product's standard of identity and no

commenter advocating the use of greater than 300 ppm has, since the

inception of this rulemaking, offered such a purpose, or indeed any

purpose, for the use of such a level of citric acid.

ATF believes that the 300 ppm figure demonstrates sensitivity to

the amounts of citric acid used in the marketplace in that it is closer

to the level of citric used by some producers than the 282 ppm level

that the ATF Laboratory derived from the OS&E data. The 300 ppm level

also recognizes the uncertainty inherent in any determination where the

entire population is not sampled. Moreover, a level of 300 ppm is

acceptable to the ATF Laboratory because that level of citric acid

represents a de minimis difference from its conclusions with respect to

the OS&E tests and because it believes that the integrity of the

standard of identity for vodka continues to be maintained at that

level. That is, the 300 ppm level is within a reasonable range of the

initial point of detection for the majority of people. As such, a level

of 300 ppm is a reasonable interpretation in light of both the OS&E

data and the practical considerations surrounding the marketplace.

ATF believes that a level of 1,000 ppm as suggested by several

commenters throughout the rulemaking can only be supported if the

standard of identity for vodka is changed. Thus, ATF believes that a

limit of 300 ppm is reasonable for all of the above-stated reasons.

Accordingly, ATF is amending 27 CFR 5.23(a)(3)(ii) to authorize the

use of a trace amount (defined as up to 300 milligrams per liter or 300

ppm) of citric acid in the production of vodka, without changing its

designation as vodka.

Formula and Label Approvals

Vodka products bottled or imported for sale or distribution in the

United States on or after December 29, 1995 must comply with the

standards set forth in this final rule. It is the responsibility of

domestic bottlers to provide, upon request, sufficient evidence to

establish that the vodka product was bottled in the United States prior

to December 29, 1995. It is the responsibility of importers to provide,

upon request, sufficient evidence to establish that the vodka product

was bottled in a foreign country prior to December 29, 1995.

Existing formulas and corresponding labels that are not in

compliance with the standards set forth in this final rule are

effectively canceled December 29, 1995, and where necessary, new

formulas and affected labels should be re-submitted to the Product

Compliance Branch, Bureau of Alcohol, Tobacco and Firearms, Room 5408,

650 Massachusetts Avenue., NW., Washington, DC 20226.

Regulatory Flexibility Act

It is hereby certified that this regulation will not have a

significant economic impact on a substantial number of small entities.

No new recordkeeping or reporting requirements are proposed.

Accordingly, a regulatory flexibility analysis is not required.

Executive Order 12866

It has been determined that this regulation is not a significant

regulatory action as defined in E.O. 12866 because (1) it will not have

an annual effect on the economy of $100 million or more or adversely

affect in a material way the economy, productivity, competition, jobs,

the environment, public health or safety, or state, local, or tribal

communities; (2) create a serious inconsistency or otherwise interfere

with an action taken or planned by another agency; (3) materially alter

the budgetary impact of entitlement, grants, user fees, or loan

programs, or the rights and obligations of recipients thereof; or (4)

raise novel legal or policy issues arising out of legal mandates, the

President's priorities, or the principles set forth in Executive Order

12866.

Paperwork Reduction Act

The provisions of the Paperwork Reduction Act of 1980, Public Law

96-511, 44 U.S.C. Chapter 35, and its implementing regulations, 5 CFR

Part 1320, do not apply to this final rule because no requirement to

collect information is imposed.

Drafting Information

The principal author of this document is David W. Brokaw, Wine and

Beer Branch, Bureau of Alcohol, Tobacco, and Firearms.

List of Subjects in 27 CFR Part 5

Advertising, Consumer protection, Customs duties and inspection,

Imports, Labeling, Liquors, Packaging and Containers

Authority and Issuance

27 CFR Part 5--Labeling and Advertising of Distilled Spirits, is

amended as follows:

PART 5--LABELING AND ADVERTISING OF DISTILLED SPIRITS

Paragraph 1. The authority citation for 27 CFR, Part 5 continues to

read as follows:

Authority: 26 U.S.C. 5301, 7805, 27 U.S.C. 205.

Par. 2. Section 5.23(a)(3)(ii) and (iii) are revised to read as

follows:

Sec. 5.23 Alteration of class and type.

(a) Additions. * * *

* * * * *

(3) * * * (ii) any material, other than caramel, infusion of oak

chips, and sugar, in the case of Cognac brandy; or (iii) any material

whatsoever in the case of neutral spirits or straight whiskey, except

that vodka may be treated with sugar in an amount not to exceed 2 grams

per liter, and, on and after December 29, 1995, with citric acid in an

amount not to exceed 300 milligrams per liter.

* * * * *

Par. 3. Section 5.23(c) is revised to read as follows:

Sec. 5.23 Alteration of class and type.

* * * * *

(c) Exceptions. (1) This section shall not be construed as in any

manner modifying the standards of identity for cordials and liqueurs,

flavored brandy, flavored gin, flavored rum, flavored vodka, and

flavored whisky or as authorizing any product which is defined in

Sec. 5.22(j), Class 10, as an imitation to be otherwise designated.

(2) Vodka products bottled in the United States or a foreign

country prior to December 29, 1995 are exempt from the citric acid

limitation set forth in Sec. 5.23(a)(3)(iii). It is the responsibility

of the bottler or importer to provide, upon request, sufficient

evidence to establish that the vodka was bottled prior to such date.

Signed: November 11, 1994.

Daniel R. Black,

Acting Director.

Approved: November 23, 1994.

John P. Simpson,

Deputy Assistant Secretary (Regulatory, Tariff, and Trade Enforcement).

[FR Doc. 94-32010 Filed 12-28-94; 8:45 am]

BILLING CODE 4810-31-P