[Federal Register Volume 70, Number 201 (Wednesday, October 19, 2005)]
[Proposed Rules]
[Pages 60751-60769]
From the Federal Register Online via the Government Publishing Office [www.gpo.gov]
[FR Doc No: 05-20874]


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

DEPARTMENT OF HEALTH AND HUMAN SERVICES

Food and Drug Administration

21 CFR Part 133

[Docket No. 2000P-0586 (formerly Docket No. 00P-0586)]


Cheeses and Related Cheese Products; Proposal to Permit the Use 
of Ultrafiltered Milk

AGENCY: Food and Drug Administration, HHS.

ACTION: Proposed rule.

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

SUMMARY: The Food and Drug Administration (FDA) is proposing to amend 
its regulations to provide for the use of fluid ultrafiltered milk (UF) 
in the manufacture of standardized cheeses and related cheese products. 
This action responds principally to two citizen petitions: One 
submitted by the American Dairy Products Institute (ADPI) and another 
submitted jointly by the National Cheese Institute (NCI), the Grocery 
Manufacturers of America, Inc. (GMA), and the National Food Processors 
Association (NFPA). FDA tentatively concludes that this action will 
promote honesty and fair dealing in the interest of consumers and, to 
the extent practicable, will achieve consistency with existing 
international standards of identity for cheeses and related cheese 
products.

DATES: Submit comments by January 17, 2006.

ADDRESSES: You may submit comments, identified by Docket No. 2000P-
0586, by any of the following methods:
Electronic Submissions
Submit electronic comments in the following ways:
     Federal eRulemaking Portal: http://www.regulations.gov. 
Follow the instructions for submitting comments.
     Agency Web site: http://www.fda.gov/dockets/ecomments.

[[Page 60752]]

 Follow the instructions for submitting comments on the agency Web 
site.
Written Submissions
Submit written submissions in the following ways:
     FAX: 301-827-6870.
     Mail/Hand delivery/Courier [For paper, disk, or CD-ROM 
submissions]: Division of Dockets Management (HFA-305), Food and Drug 
Administration, 5630 Fishers Lane, rm. 1061, Rockville, MD 20852.
    To ensure more timely processing of comments, FDA is no longer 
accepting comments submitted to the agency by e-mail. FDA encourages 
you to continue to submit electronic comments by using the Federal 
eRulemaking Portal or the agency Web site, as described in the 
Electronic Submissions portion of this paragraph.
    Instructions: All submissions received must include the agency name 
and Docket Nos. or Regulatory Information Number (RIN) for this 
rulemaking. All comments received may be posted without change to 
http://www.fda.gov/ohrms/dockets/default.htm, including any personal 
information provided. For detailed instructions on submitting comments 
and additional information on the rulemaking process, see the 
``Comments'' heading of the SUPPLEMENTARY INFORMATION section of this 
document.
    Docket: For access to the docket to read background documents or 
comments received, go to http://www.fda.gov/ohrms/dockets/default.htm 
and insert the docket number(s), found in brackets in the heading of 
this document, into the ``Search'' box and follow the prompts and/or go 
to the Division of Dockets Management, 5630 Fishers Lane, rm. 1061, 
Rockville, MD 20852.

FOR FURTHER INFORMATION CONTACT: Ritu Nalubola, Center for Food Safety 
and Applied Nutrition (HFS-820), Food and Drug Administration, 5100 
Paint Branch Pkwy., College Park, MD 20740, 301-436-2371.

SUPPLEMENTARY INFORMATION:

Table of Contents

I. Background
    A. Petitions and Grounds
    1. The 1999 ADPI Petition
    2. The 2000 NCI/GMA/NFPA Joint Petition
    B. The Government Accountability Office (GAO) Report
    C. Comments to Petitions
    D. Forms of Milk Permitted as Basic Dairy Ingredients
    E. Temporary Marketing Permit (TMP)
II. The Proposal
    A. Legal Authority/Statutory Directive
    B. Options Considered
    C. Proposed Amendments
III. Executive Order 12866: Cost Benefit Analysis
    A. Need for Regulation
    B. Background and Current Industry Practices
    C. Regulatory Options
    D. Summary of Costs and Benefits
IV. Small Entity Analysis
V. Unfunded Mandates
VI. Small Business Regulatory Enforcement Fairness Act of 1996 (SBREFA) 
Major Rule
VII. Federalism
VIII. Environmental Impact
IX. Paperwork Reduction Act of 1995
X. Comments
XI. References

I. Background

    The standards of identity for cheeses and related cheese products 
are specified in part 133 (21 CFR 133). The general provisions within 
part 133, in part, define ``milk'' and ``nonfat milk'' that may be used 
in the manufacture of cheeses and related cheese products. The 
definitions for ``milk'' and ``nonfat milk'' in Sec.  133.3(a) and (b), 
respectively, list different forms of milk and nonfat milk, including 
concentrated, reconstituted, and dried forms, that may be used in the 
making of cheeses and related cheese products. However, fluid or dried 
filtered forms of milk obtained through mechanical filtration of milk 
or nonfat milk are not included within these definitions. Therefore, 
while current regulations permit the use of concentrated, 
reconstituted, and dried forms of milk and nonfat milk as basic dairy 
ingredients, they do not provide for the use of fluid or dried filtered 
milk or fluid or dried filtered nonfat milk as basic dairy ingredients 
in standardized cheeses and related cheese products.
    Mechanical filtration technologies available for milk processing 
include microfiltration, ultrafiltration, nanofiltration, and reverse 
osmosis (Refs. 1 and 2). In all of these filtration methods, milk is 
passed over a series of semipermeable membranes with varying pore 
sizes. The portion of milk that passes through the membranes is 
referred to as the ``permeate,'' and the portion that does not pass 
through the membranes is referred to as the ``retentate.'' While the 
application of hydraulic pressure is the driving force for these 
membrane separation processes, the nature of the membrane itself (as 
well as the orientation of the components) controls which components of 
milk are separated into the permeate and which components are retained 
in the retentate during these filtration processes (Refs. 1 and 2). In 
a reverse osmosis (RO) filtration, the membrane pore size is such that 
all components other than water in the milk are retained. 
Nanofiltration uses membranes with pores that are larger than RO 
membranes, but smaller than those used in ultrafiltration. In milk 
processing, nanofiltration can be used to remove water as well as some 
soluble salts, yet retain all other components of milk (Refs. 1 and 2). 
Ultrafiltration retains macromolecules and particles larger than about 
0.001-0.02 micrometers, while microfiltration is designed to retain 
particles between about 0.10 micrometers to 5 micrometers (Ref. 1). 
While there is some overlap in membrane pore sizes and operating 
pressures used in ultrafiltration and microfiltration (Refs. 1 and 3), 
in dairy processing, ultrafiltration is typically used to retain all 
protein components of milk, including casein and whey proteins, while 
some of the lactose, minerals, and water soluble vitamins present in 
milk are lost along with water. Microfiltration, on the other hand, is 
primarily used for fat separation, bacterial removal, and casein 
concentration, with a resulting loss of whey proteins, lactose, 
minerals, and water soluble vitamins along with water (Refs. 1, 2, and 
3).

A. Petitions and Grounds

    FDA received two petitions requesting amendments to existing 
regulations to permit the use of filtered milk in the manufacture of 
standardized cheeses and related cheese products.
1. The 1999 ADPI Petition
    The ADPI filed a citizen petition (CP) on December 2, 1999 (Docket 
No. 1999P-5198 (formerly Docket No. 99P-5198); hereafter referred to as 
the ADPI petition) requesting that the FDA amend the definition of 
``milk,'' as provided in Sec.  133.3(a), to include fluid UF milk, 
thereby permitting the use of fluid UF milk in the manufacture of 
standardized cheeses and related cheese products specified in part 133. 
ADPI requested that Sec.  133.3(a) be amended to add that ``milk may be 
subjected to an ultrafiltration process that results in a fluid UF milk 
for use in the manufacture of cheese.'' In its petition, ADPI stated 
that the requested amendment would improve efficiencies in cheese 
manufacturing and result in benefits to consumers without alteration of 
cheese composition, characteristics, or flavor. FDA reviewed the ADPI 
petition and determined that it did not present reasonable grounds in 
accordance with 21 CFR 10.30 to support the requested amendment and, 
therefore, FDA closed

[[Page 60753]]

this petition. However, because the issues raised in the ADPI petition 
are clearly covered under a second citizen petition (Docket No. 2000P-
0586 (formerly Docket No. 00P-0586)/CP2, discussed in section I.A.2 of 
this document), FDA converted the ADPI petition into a comment to this 
second petition. ADPI was informed of FDA's action in a letter dated 
February 26, 2003.
2. The 2000 NCI/GMA/NFPA Joint Petition
    On June 13, 2000, FDA received a joint petition (Docket No. 2000P-
0586 (formerly Docket No. 00P-0586)/CP2; hereafter referred to as the 
NCI petition) from the NCI, the GMA, and the NFPA requesting an 
amendment of Sec.  133.3 to include ``filtered milk'' in the definition 
of ``milk'' and ``filtered skim milk'' in the definition of ``nonfat 
milk'' for use in standardized cheeses and related cheese products. The 
NCI petition also requested that a new subsection be added within Sec.  
133.3 to define ``filtered milk'' as:
    * * * the liquid milk product produced by a physical separation 
technique in which raw or pasteurized milk is passed over one or 
more semipermeable membranes to partially remove the water phase and 
its constituents, including water, lactose, whey proteins, and 
minerals. Either before or after filtration, fat may be separated to 
produce filtered skim milk. After filtration, water may be partially 
removed by means of evaporation to produce more concentrated forms 
of filtered milk.''
Based on this definition, FDA believes that the petitioners requested 
the agency to permit not only ultrafiltration (which typically does not 
result in a loss of whey proteins), but also other filtration 
techniques such as microfiltration and subsequent treatment to further 
concentrate the filtered product, in the manufacture of standardized 
cheeses and related cheese products. The petitioners withdrew a 
previous joint petition (Docket No. 2000P-0586 (formerly Docket No. 
00P-0586)/CP1) that requested amendments to permit both fluid and dried 
forms of filtered milk in the manufacture of standardized cheeses and 
related cheese products.
    In support of their requested amendments, the NCI, GMA, and NFPA 
(hereafter referred to as the petitioners) argued that the amendments 
requested in the NCI petition are consistent with established FDA 
policy. Some cheese standards, in addition to specifying a specific 
procedure for preparing the food, currently provide for the use of 
``any other procedure which produces a finished cheese having the same 
physical and chemical properties'' (see e.g., standard of identity for 
cheddar cheese in Sec.  133.113). The petitioners maintained that these 
``alternate make procedure'' provisions historically have provided the 
legal basis for the use of milk filtration and the resulting filtered 
milk in cheese making, regardless of whether the filtration occurs in 
the same plant as other cheese-making procedures or in a centralized 
filtration facility. The petitioners believe that FDA has previously 
acknowledged that the use of filtered milk to manufacture cheddar 
cheese is covered by the alternate make procedure provision of the 
standard of identity for cheddar cheese. Furthermore, the petitioners 
maintained that the requested amendments are fully consistent with the 
basis and rationale for amendments that FDA previously made to expand 
the scope of the forms of milk recognized as ``milk'' for cheese 
making. The petitioners stated that FDA authorized the use of certain 
forms of milk because these forms of milk may be used in place of fluid 
milk to produce a finished cheese that is equivalent physically and 
chemically to the traditional cheese made using fluid milk.
    In addition, the petitioners stated that mechanical filtration has 
been used in cheese manufacturing in the United States for the past 20 
years, and contended that the extensive use of filtration technologies, 
under the existing ``alternate make procedure'' provisions within some 
standards of identity for cheeses, has produced significant benefits by 
improving product consistency and yields and manufacturing efficiency; 
lowering milk refrigeration, hauling and whey disposal costs; expanding 
milk sourcing options; and enabling cheese makers to respond more 
effectively to regional disruptions in the fluid milk supply. The 
petitioners also stated that because mechanical filtration removes only 
those constituents that are removed by loss of whey in traditional 
cheese making, it functions simply to rearrange the steps in the cheese 
making process to permit the constituents to be removed earlier. The 
petitioners further contended that the long history and widespread use 
of filtration technology under the alternate make procedure provisions 
have clearly established the equivalence of cheese made from filtered 
milk and cheese made from other forms of milk explicitly permitted 
under Sec.  133.3.
    The petitioners also argued that cheese made with filtered milk is 
nutritionally equivalent to traditional cheese because mechanical 
filtration of milk using membranes with pore sizes between 0.0001 and 
0.20 microns removes the water phase constituents (water, soluble 
protein, lactose, minerals, and some water soluble vitamins) that 
otherwise would be removed in the traditional cheese-making process as 
whey. In fact, the petitioners argued, with respect to filtered milk in 
cheese, the retentate may actually contain slightly greater 
concentrations of valuable constituents (e.g., whey proteins) than the 
cheese curd that remains after loss of whey in traditional cheese 
making. The petitioners provided analytical data related to cheddar 
cheese to support their assertion that cheese made with filtered milk 
is not ``nutritionally inferior,'' as that term is defined in 21 CFR 
101.3(e)(4), to cheese made using traditional procedures.
    Finally, the petitioners argued that their proposed amendments are 
consistent with the Codex Alimentarius Commission (Codex) standard for 
cheese. The Codex standard for cheese (Standard A-6-1978, revised in 
January 1999) provides for the use of ``milk and/or products obtained 
from milk.'' The petitioners stated that the Codex standard encompasses 
mechanical filtration technology, provided the finished cheese meets 
applicable requirements for physical and chemical properties, which 
would include nutritional and organoleptic properties.

B. The Government Accountability Office (GAO)\1\ Report
---------------------------------------------------------------------------

    \1\ The GAO changed its name from the ``General Accounting 
Office'' in 2004.
---------------------------------------------------------------------------

    The fiscal year (FY) 2000 FDA appropriations bill from the U.S. 
Senate requested the Comptroller General to conduct a study to 
determine the quantity and end use of UF milk imported into the United 
States and to submit a report describing the results of the study to 
Congress. In March 2001, GAO reported (hereafter referred to as ``the 
GAO report'' (Ref. 4)), in part, that: There are no specific data on UF 
milk imports because UF milk is classified under the broad category of 
``milk protein concentrates'' (MPC) by the U.S. Customs Service. GAO 
reported that imports in the broad category of MPC rose dramatically 
between 1990 and 1999 from about 800 to 45,000 metric tons, the primary 
reasons being the difference between U.S. and international prices of 
milk protein, especially nonfat dry milk (NFDM), and the market growth 
of nutritional supplements and other novel foods using MPC. GAO also 
reported that dry MPC imports are used in several foods other than 
cheeses, such as frozen desserts, bakery products, and sports and other 
nutritional supplement products. Some in the industry note that

[[Page 60754]]

economic disincentives have prevented domestic production of dry MPC. 
GAO noted that there are limited data on domestic production and use of 
fluid UF milk in cheese making but found that 22 dairy plants produce 
fluid UF milk used to make cheese within the plant, while 4 dairy farms 
in New Mexico and Texas produce fluid UF milk for transport to cheese 
plants in the Midwest. GAO also found that FDA and State contract 
inspectors reported no violations related to the use of imported UF 
milk or MPC in standardized cheese in FY 1999, whereas in FY 2000, two 
plants in Vermont were issued warning letters for using imported MPC in 
standardized cheese, and the plants subsequently discontinued this use.

C. Comments to Petitions

    FDA received a total of 58 letters and e-mails, each containing one 
or more comments, to the ADPI (subsequently converted to a comment to 
the NCI petition) and the NCI petitions. A large portion of the letters 
and e-mails received were from individual dairy farmers, organizations 
representing dairy farmers, and consumers. Nearly half of the comments 
opposed both the ADPI and NCI petitions, while the other half opposed 
the NCI petition alone without commenting on the ADPI petition. A few 
comments expressed support for the ADPI petition, but none of the 
comments supported the NCI petition. The primary concern expressed by 
the comments opposing either of the petitions appeared to be the 
potential economic impact of the use of imported milk ingredients, 
particularly dried forms of filtered milk or MPC, on U.S. dairy 
farmers. Some comments also expressed concern about the use of imported 
milk ingredients on the quality and safety of cheese.
    The organizations representing dairy farmers expressed strong 
opposition to both petitions and stated that the use of filtered milk 
would undoubtedly lower the quality of cheese products and greatly 
increase the flood of imports of subsidized MPC and filtered milk with 
the potential to jeopardize the safety of cheese products. They stated 
that the filtration process removes calcium and reduces the lactose 
content of milk and results in cheese that does not have the fullness 
of flavor of traditional cheese. They further maintained that changing 
the definition of milk to allow the use of liquid filtered milk would 
ultimately result in the use of dry filtered MPC and, therefore, they 
reiterated that even if only liquid filtered milk were allowed, while 
disallowing dry MPC, they would still be concerned about product 
quality degradation. In addition, they stated that changing the 
definition of milk could result in increased imports of filtered milk 
from Canada, displacing U.S. milk and causing a surplus. However, these 
comments did not provide any factual data or information that would 
lead FDA to believe that the use of fluid UF milk would impact the 
safety or quality of the product.
    Another comment, from an organization representing milk producers, 
unconditionally endorsed the ADPI petition, but strongly opposed the 
NCI petition, stating that the commenter does not support any change to 
Sec.  133.3(a) that alters which products are currently defined as 
``milk.'' This comment stated that the language in the NCI petition is 
sufficiently vague that it may be subject to interpretation such that 
it subsequently would allow dried forms of UF milk. The comment also 
stated that permitting only liquid forms of UF milk has general 
widespread support among different stakeholders, and argued that it is 
essential to establish a definition of ``liquid'' UF milk to mitigate 
potential misinterpretations regarding the use of dried MPC and provide 
clarity for enforcement. In this regard, the comment suggested that a 
limitation of 45 percent total solids be included in the definition of 
``liquid ultrafiltered milk,'' because a requirement of a maximum of 45 
percent total solids would allow for the use of UF technology while 
preserving the liquid state of the ultrafiltered product and preventing 
subsequent treatment for concentration beyond ultrafiltration.

D. Forms of Milk Permitted as Basic Dairy Ingredients

    The definitions of ``milk'' and ``nonfat milk'' in Sec.  133.3 do 
not provide for the use of filtered milk or filtered nonfat milk as 
basic dairy ingredients in standardized cheeses and related cheese 
products. In 1983, with respect to the use of the forms of milk that 
are permitted as basic ingredients in cheesemaking, FDA amended Sec.  
133.3 to define the class designations ``milk,'' ``nonfat milk,'' and 
``cream'' and provide for alternate forms of milk, nonfat milk, and 
cream, i.e., concentrated, dried, and reconstituted forms to be used in 
standardized cheeses and related cheese products (48 FR 2736, January 
21, 1983). In the proposed rule, FDA advised of its opinion that these 
alternate forms can be used to produce the same cheese as produced from 
fluid cow's milk (43 FR 42127 at 42128, September 19, 1978), which was 
the only form of milk permitted as the basic ingredient for cheese 
manufacture at that time. Filtered forms, however, are not included 
within ``milk'' or ``nonfat milk'' permitted in standardized cheeses 
and related cheese products.
    In the NCI petition, the petitioners argued that the alternate make 
procedure that is provided for in some cheese standards historically 
has provided the legal basis for the use of milk filtration and the 
resulting filtered milk as an ingredient in cheese making. FDA does not 
agree with the petitioners. The alternate make procedure provision 
provides for the use of ``any other procedure which produces a finished 
cheese having the same physical and chemical properties'' as the 
procedure specified in the standard. For example, the procedure for 
making blue cheese described in Sec.  133.106(a)(2) requires 
Penicillium roquefortii spores to be added to the curd. In a final 
rulemaking in 1983, in response to a comment that this requirement 
should be changed to permit the addition of spores to dairy ingredients 
rather than only to the curd, FDA noted that a change is not necessary 
because the procedure described in Sec.  133.106(a)(2) may be modified 
as provided for in Sec.  133.106(a)(1), which states that any other 
procedure may be used which produces a finished cheese having the same 
physical and chemical properties (48 FR 2736 at 2739). Rather than 
restricting the manufacturing procedure to the one specifically 
described in the standard, this provision allows manufacturers to use 
alternate manufacturing procedures, but not alternate ingredients, 
provided the alternate manufacturing procedure does not adversely 
affect the physical and chemical properties of the cheese. However, the 
alternate make procedure provision does not permit the use of dairy or 
other ingredients that are not specifically provided for in the cheese 
standard. Therefore, the alternate make provision of current cheese 
standards allows manufacturers to appropriately process the basic 
ingredient milk during the cheese-making process. For example, the 
ingredient milk may undergo an additional step of ultrafiltration prior 
to being introduced into the cheese vat in a single within-batch and 
within-plant production line for cheese making. In such a process, the 
ingredient that is introduced into the cheese-making process is milk. 
However, fluid UF milk purchased or brought in from another plant, even 
within the same company, that is then introduced into cheese making is 
considered an alternate ingredient because the ultrafiltration process 
is

[[Page 60755]]

used solely for the production of an ingredient that is subsequently 
used in cheese making. Therefore, in this case, the ingredient is fluid 
UF milk, not milk.
    In the NCI petition, the petitioners also stated that FDA has 
previously acknowledged that the use of filtered milk to manufacture 
cheddar cheese is covered by the alternate make procedure provision of 
the cheddar cheese standard, including when filtration occurs in a 
separate centralized facility. FDA clarifies that it has previously not 
objected to the use of fluid UF milk in cheddar cheese under specific 
circumstances. In 1996, FDA granted temporary permission to Bongards 
Creamery in Minnesota to manufacture cheddar cheese using fluid UF milk 
that is produced on a farm in New Mexico. That permission was granted 
on a limited basis in response to a request from the T.C. Jacoby & 
Company, Inc., to run a testing program at Bongards Creamery during a 
pilot period to demonstrate that the finished cheddar cheese made with 
fluid UF milk as an ingredient has the same physical and chemical 
characteristics as traditional cheddar cheese (Ref. 5). In its response 
to T.C. Jacoby & Company, Inc., FDA stated that based on its 
understanding that ``cheddar cheese produced with the retentate that 
results when milk is subjected to processing in a ultrafiltration 
system is nutritionally equivalent to and is physically and chemically 
identical'' to cheddar cheese prepared by the standardized procedure, 
it would not object to the use of fluid UF milk in the manufacture of 
cheddar cheese at Bongards Creamery on the limited basis described by 
T.C. Jacoby & Company, Inc. (Ref. 6).
    Subsequently, FDA stated its interpretation of the cheese standards 
that, as written, they do not allow for the use of UF milk as an 
ingredient (Ref. 7). FDA reaffirms that the use of filtered milk, dried 
or fluid, including fluid UF milk, as an ingredient is not covered 
under the alternate make procedures provided for in certain 
standardized cheeses. However, while FDA has considered the use of UF 
milk in standardized cheeses, it has stated that it would not object to 
the experimental use of fluid UF milk as an ingredient in cheddar and 
mozzarella cheeses (Ref. 7) and that enforcement regarding the use of 
UF milk as an ingredient in Swiss cheese is not a priority (Ref. 8).
    Substances commonly referred to as MPC are also not permitted as 
ingredients in standardized cheeses. While there is no current FDA 
regulation that defines ``MPC'' and this term does not appear to have a 
standard definition within the industry, the term ``MPC'' is generally 
used to refer to dried forms of filtered milk and dried blends and 
coprecipitates of milk proteins (Ref. 9). The existing standards of 
identity in part 133 do not list MPC as a permitted optional ingredient 
in the manufacture of standardized cheeses or related cheese products. 
Ingredients that are not specifically provided for by the standard 
cannot be used in the manufacture of a food named with the standardized 
term. FDA reiterated this statement in 1983 when FDA amended the 
standards for nine natural cheeses to bring them into closer 
conformance with the recommended Codex standards for those cheeses (48 
FR 2736). FDA advised that dairy ingredients that may be used in 
manufacture of standardized cheeses are specifically listed in the 
individual standards, and that milk-derived ingredients other than 
those specifically provided for may not be used in these cheeses (48 FR 
2736 at 2737). In addition, specific to the use of caseinates in 
standardized cheeses, FDA previously addressed comments on the use of 
caseinates in previous rulemakings (48 FR 2736 at 2737 and 58 FR 2431 
at 2439, January 6, 1993), and advised that caseinates are not among 
the dairy ingredients provided for use in the manufacture of 
standardized cheeses in part 133 and, therefore, cannot be used. FDA 
reaffirms that ingredients other than those specifically provided for 
by the individual standards cannot be used in the making of 
standardized cheeses and related cheese products.
    Therefore, under the current regulations, use of filtered milk, 
including fluid UF milk, as an ingredient in a cheese whose applicable 
standard(s) does not provide for its use would constitute a deviation 
from the standard, and such cheese cannot be named by the standardized 
term. However, under the provisions of 21 CFR 130.17, food 
manufacturers may request from FDA a temporary marketing permit (TMP) 
to market a food that is named by the standardized term but that 
deviates from its standard of identity.

E. Temporary Marketing Permit (TMP)

    On August 1, 2002, FDA received an application from Wells' Dairy, 
Inc. (Wells' Dairy), for a TMP for the use of UF milk in the 
manufacture of cottage cheese. In the Federal Register of December 9, 
2004 (69 FR 71418), FDA announced the issuance of a TMP to Wells' Dairy 
to market test cottage cheese that deviates from the standard of 
identity for cottage cheese in that the product is formulated using 
fluid UF skim milk. For the purpose of this TMP, fluid UF skim milk was 
described as ``the product obtained by subjecting skim milk to a 
physical separation process called ultrafiltration using a membrane 
with a pore size of 10,000 Daltons (Da) molecular weight cut-off 
(MWCO), resulting in the partial loss of lactose, minerals, water-
soluble vitamins, and water present in skim milk.'' The TMP also 
specified that the casein-to-whey protein ratio of skim milk is not 
altered during the ultrafiltration process and that the moisture 
content of fluid UF skim milk is about 80 percent. The TMP permitted 
the addition of such fluid UF skim milk to skim milk at a level needed 
to increase the total solids of the cheese milk (or final milk used to 
make cheese) by 5 to 25 percent, and required fluid UF skim milk to be 
declared in the ingredient statement of the finished cottage cheese as 
``ultrafiltered skim milk.'' The purpose of the permit was to allow 
Wells' Dairy to measure consumer acceptance of the product, identify 
mass production problems, and assess commercial feasibility. The permit 
provided for the temporary market testing of 15 million pounds (lb) 
(6.8 million kilograms) of the test product for a period of 15 months.

II. The Proposal

A. Legal Authority/Statutory Directive

    Section 401 of the Federal Food, Drug, and Cosmetic Act (the act) 
(21 U.S.C. 341)) directs the Secretary of Health and Human Services 
(the Secretary), to issue regulations fixing and establishing 
reasonable definitions and standards of identity, quality, or fill of 
container whenever such action will promote honesty and fair dealing in 
the interest of consumers. Section 701(e) of the act (21 U.S.C. 371(e)) 
directs the Secretary to publish a proposal for the amendment or repeal 
of any definition and standard of identity under section 401 of the act 
for any dairy product (e.g., cheese) that is based on a petition of any 
interested persons showing reasonable grounds.

B. Options Considered

    FDA considered several options in response to the two petitions, 
including the following: (1) Denying the two petitions, (2) proposing 
to permit the use of all fluid forms of filtered milk, (3) proposing to 
permit the use of all fluid and dried forms of filtered milk, and (4) 
proposing to permit the use of fluid UF milk. FDA concluded that the 
first option would not be appropriate given that the NCI petition 
includes within its scope allowing the use of UF milk in standardized 
cheeses, which FDA

[[Page 60756]]

tentatively concludes, for reasons discussed under option 4, should be 
permitted.
    The second option, to provide for the use of all fluid forms of 
filtered milk in standardized cheeses, was also determined to be 
inappropriate. Standards of identity regulations establish the name of 
the food, which identifies and describes the food's basic nature (43 FR 
42118 at 42120, September 19, 1978). As FDA discussed in 1950 during 
the establishment of the cheese standards of identity, the starting 
point for all varieties of cheese is milk. In preparing milk for use in 
cheese making, adjustments may be made by adding or removing milk fat 
in the form of cream, fresh skim milk, NFDM solids, or concentrated 
skim milk so that the ratio of milk fat to the nonfat milk solids is at 
a desired level (15 FR 5656 at 5657, August 24, 1950). FDA reiterates 
its longstanding interpretation that a basic nature of cheese is that 
it is a food made using milk as the starting ingredient. Proposing to 
allow the use of all fluid forms of filtered milk in standardized 
cheeses was rejected because some forms of filtration concentrates are 
specific individual components of milk resulting in a retentate that is 
no longer milk. For example, microfiltration can be used to separate 
whey proteins along with lactose, minerals, and water-soluble vitamins 
from milk resulting in the concentration of casein fractions. FDA 
tentatively believes that such products that are merely concentrates of 
certain individual milk components are not milk. The use of individual 
components of milk, such as specific milk proteins, as the basic or 
starting ingredient in cheese is not consistent with the basic nature 
of cheese in that cheese is a food prepared using milk, not specific 
individual components of milk. Moreover, as FDA previously noted, when 
providing flexibility for use of advances in food technology, food 
standards should ensure that the basic nature of the food remains 
essentially the same (60 FR 67492 at 67499, December 29, 1995). FDA 
tentatively concludes that allowing for the use of technologies that 
could potentially result in the use of a specific component of milk as 
the starting ingredient of cheese would seem to violate the intent of 
the cheese standards of identity to preserve the basic nature of 
cheese.
    In the NCI petition, the petitioners also stated that because 
mechanical filtration removes only those constituents that are removed 
by loss of whey in traditional cheese making, it functions simply to 
rearrange the steps in the cheese-making process to permit the 
constituents to be removed earlier. FDA believes that food standards 
should provide for flexibility in manufacturing procedures and 
ingredients, provided that the basic nature and essential 
characteristics of the food are preserved. In determining which 
filtered products are appropriate for use as ingredients in cheeses, 
FDA considered how the use of a type of filtered milk affects the basic 
nature and essential characteristics of cheese. While filtration 
selectively and variably removes different constituents of milk that 
are lost, to varying degrees, during the whey removal process in the 
traditional cheese-making process, we do not agree that this fact can 
form a sufficient basis to support the use of all forms of fluid 
filtered milk as ingredients. Some forms of filtration result in 
retentates that are specific individual components of milk and are no 
longer milk. In addition, research suggests that milk that is 
concentrated to higher levels of protein is not suited for use in all 
types of cheeses, with adverse effects on quality being reported 
particularly in the case of hard and semi-hard cheeses (Refs. 1, 10, 
and 11). Moreover, FDA believes that in determining the appropriateness 
of different forms of filtered milk as ingredients in cheese a primary 
criterion, based on a fundamental principle of food standards, is 
whether the use of the filtered milk ensures the integrity of the 
standardized cheese--its basic nature and essential characteristics. As 
explained in the previous paragraph, FDA tentatively concludes that the 
use of a product of microfiltration as the starting ingredient of 
cheese is not consistent with the basic nature of cheese. Therefore, we 
do not agree that it is appropriate to provide for the use of all types 
of fluid filtered milk nor do we agree that the argument about the 
``rearrangement'' of the steps of cheese making (as described by the 
petitioners) sufficiently supports the appropriateness of the use of 
all forms of fluid filtered milk an ingredient.
    A third option that was also considered inappropriate was to 
provide for all filtered milk, including both fluid and dried forms. 
Under this option, substances such as MPC, dry microfiltered (MF) milk, 
and caseins would be permissible in standardized cheeses or related 
cheese products. FDA's concerns regarding the use of all fluid filtered 
milk, which are stated in the two previous paragraphs, also apply to 
the use of dried filtered milks. Allowing for the use of technologies 
that could potentially result in the use of specific components of 
milk, such as caseins, rather than milk, as the starting ingredient of 
cheese would be inconsistent with the basic nature of cheese.

C. Proposed Amendments

    Based on all the information available, including the information 
presented by the two petitions and the comments received thus far, FDA 
is proposing to amend the definitions of ``milk'' and ``nonfat milk'' 
in Sec.  133.3 to do the following: (1) Provide for ultrafiltration of 
milk and nonfat milk and (2) define UF milk and nonfat milk as raw or 
pasteurized milk or nonfat milk that is passed over one or more 
semipermeable membranes to partially remove water, lactose, minerals, 
and water-soluble vitamins without altering the casein-to-whey protein 
ratio of the milk and resulting in a liquid product. FDA is also 
proposing that the name of such treated milk is ``ultrafiltered milk'' 
or ``ultrafiltered nonfat milk,'' as appropriate. Consequently, when 
this type of milk is used, it would be declared in the ingredient 
statement of the finished food as ``ultrafiltered milk'' or 
``ultrafiltered nonfat milk.''
    First, providing for the use of fluid UF milk is consistent with 
the basic nature of cheese in that the starting ingredient is milk. 
During the process of ultrafiltration, some of the lactose, soluble 
salts, and water-soluble vitamins of milk pass through the membranes 
and are removed, while protein, fat, fat-soluble vitamins, and some of 
the insoluble salts are retained. Therefore, unlike microfiltration, 
ultrafiltration does not result in the separation of specific fractions 
of milk proteins.
    Second, FDA tentatively concludes that fluid UF milk can be used in 
standardized cheeses while maintaining the essential characteristics of 
these cheeses specified in the individual standards of identity in part 
133. Scientific literature suggests that fluid UF milk, especially at 
low concentration factors, can be used in different cheeses (including 
soft, semi-hard, hard, and direct-acidified cheeses and process cheese) 
without adversely affecting the physical, chemical, or organoleptic 
properties of the cheese (Refs. 1, 2, and 11 through 20; Appendix F of 
the NCI petition). This appears to be especially true with soft cheeses 
such as cottage cheese (Refs. 1, 14, and 15) and some direct-acidified 
cheeses (Ref. 12). Specifically with respect to cottage cheese, as 
noted in section I.E of this document, FDA reviewed relevant scientific 
information related to the use of fluid UF milk as an ingredient and 
determined that fluid UF milk may be used in cottage cheese without

[[Page 60757]]

adversely affecting the essential physical or chemical characteristics, 
including nutritional composition and organoleptic properties of 
cottage cheese. FDA issued a TMP to Wells' Dairy to market test cottage 
cheese that deviates from the standard of identity for cottage cheese 
in that the product is formulated using fluid UF skim milk (69 FR 
71418).
    FDA notes, however, that the scientific literature also includes 
some reports of adverse effects from the use of fluid UF milk on the 
texture and development of flavor and aroma of certain cheeses, 
particularly in semi-hard and hard cheeses and with the use of fluid UF 
milk at higher concentration factors (Refs. 1, 11, 17, and 21 through 
24). FDA points out that the use of fluid UF milk must not adversely 
affect the physical or chemical characteristics of the cheese. The 
cheese standards of identity ensure the integrity of the cheese by 
setting limits on its fat, milk solids-not-fat, and moisture content. 
In addition, FDA considers nutritional equivalency and organoleptic 
properties of the cheese among other factors to determine whether the 
essential characteristics of the cheese are maintained. Providing for 
the use of fluid UF milk does not preclude a standardized cheese from 
meeting the existing requirements within the applicable individual 
standard(s) of identity in part 133. Rather, the use of fluid UF milk 
would be optional and any cheese made using fluid UF milk would have to 
meet all the requirements, including the physical and chemical 
characteristics, specified in the applicable individual standard(s) of 
identity.
    Third, FDA anticipates that providing for the use of fluid UF milk 
would enable cheese manufacturers to benefit from advances in milk 
filtration technology and provide them with greater flexibility in 
cheese making, while preserving the basic nature and essential 
characteristics of standardized cheese. Further, using ultrafiltration 
technology may result in better retention of milk proteins and greater 
cheese yields as well as more uniform product quality (Ref. 1). In 
addition, the petitioners claimed that using fluid filtered milk 
(including fluid UF milk) helps manage seasonal imbalances in milk 
supplies and demand for cheese, and reduces the costs associated with 
bulk milk distribution, resulting in cost savings that ultimately could 
be passed on to consumers. Furthermore, declaring fluid UF milk in the 
ingredient statement of the cheese as ``ultrafiltered milk'' or 
``ultrafiltered skim milk,'' as appropriate, would enable consumers to 
identify cheeses made with milk that has undergone ultrafiltration.
    Finally, providing for the use of fluid UF milk would bring the 
standards of identity for cheeses in closer conformity with the 
international standards adopted by Codex and facilitate increased 
harmonization. In response to the ADPI and NCI petitions, FDA 
considered the relevant Codex standards for cheeses and related cheese 
products. Specifically, FDA reviewed the Codex standards for cheese 
(Codex Stan A-6), cheeses in brine (group standard) (Codex Stan 208), 
cottage cheese including creamed cottage cheese (Codex Stan C-16), 
cream cheese (Codex Stan C-31), extra hard grating cheese (Codex Stan 
C-35), unripened cheese including fresh cheese (group standard) (Codex 
Stan 221), named variety process(ed) cheese and spreadable process(ed) 
cheese (Codex Stan A-8(a)), process(ed) cheese and spreadable 
process(ed) cheese (Codex Stan A-8(b)), process(ed) cheese preparations 
(Codex Stan A-8(c)), and whey cheeses (Codex Stan A-7) (Refs. 25-34). 
FDA notes that several Codex standards such as the standard for cheese, 
group standard for cheeses in brine, and group standard for unripened 
cheese including fresh cheese all permit the use of ``milk and/or 
products obtained from milk,'' which encompasses fluid UF milk, as the 
raw material in the manufacture of theses cheeses, provided the 
finished cheese meets the relevant physical and chemical properties. 
Additionally, the Codex standard for whey cheeses provides for the 
addition of ``raw materials of milk origin,'' including fluid UF milk. 
Providing for the optional use of fluid UF milk as a basic dairy 
ingredient in cheeses would be consistent with, although not as 
expansive as, the provisions of some Codex standards.
    In a recent proposed rule (70 FR 29214, May 20, 2005) (the food 
standards proposal), FDA and FSIS proposed a set of general principles 
that define how modern food standards should be structured. The 
agencies also proposed that, if finalized, the agencies will require 
that a CP for establishing, revising, or eliminating a food standard be 
submitted in accordance with these general principles. Conversely, the 
agencies proposed that they may find deficient a petition to establish, 
revise, or eliminate a food standard that does not follow these general 
principles. FDA believes that the action proposed here to provide for 
the use of fluid UF milk as an ingredient in standardized cheeses and 
related cheese products is consistent with the general principles 
proposed in the food standards proposal.
    For the reasons explained previously in this section, FDA 
tentatively concludes that providing for the use of fluid UF milk only, 
rather than for the use of all fluid filtered milk (as requested by the 
NCI petition), would promote honesty and fair dealing in the interest 
of consumers by providing greater flexibility in cheesemaking while 
preserving the basic nature and essential characteristics of the food. 
Therefore, FDA proposes to amend the definitions of ``milk'' and 
``nonfat milk'' within Sec.  133.3 to do the following: (1) Provide for 
ultrafiltration of milk and nonfat milk and (2) define UF milk and 
nonfat milk as raw or pasteurized milk or nonfat milk that is passed 
over one or more semipermeable membranes to partially remove water, 
lactose, minerals, and water-soluble vitamins without altering the 
casein-to-whey protein ratio of the milk and resulting in a liquid 
product. FDA also proposes that the name of such treated milk is 
``ultrafiltered milk'' or ``ultrafiltered nonfat milk,'' as 
appropriate. Consequently, when this type of milk is used, it would be 
declared in the ingredient statement of the finished food as 
``ultrafiltered milk'' or ``ultrafiltered nonfat milk.''
    FDA seeks comment on the appropriateness of the proposed 
amendments, including the provision to permit the use of fluid UF milk 
and fluid UF nonfat milk. The proposed amendments would allow for 
optional ultrafiltration of the starting ingredient, milk or nonfat 
milk, used in cheese manufacturing. Under these proposed amendments, 
whether a manufacturer uses fluid UF milk is optional and entirely up 
to the manufacturer.
    FDA also seeks comment on the appropriateness of the proposed 
definition of ultrafiltration. With respect to the requirement for an 
unaltered casein-to-whey protein ratio during ultrafiltration, FDA 
acknowledges that some loss of small molecular weight whey proteins may 
occur during ultrafiltration of milk with the extent of loss partially 
dependent on the nature of the membrane and the orientation of the 
molecules in milk (which may be influenced by the treatment of milk 
prior to or during ultrafiltration). While casein and most whey 
proteins are retained in the retentate, proteose-peptones with low 
molecular weights may be lost in the permeate. Proteose-peptones have a 
molecular weight between 4,100 and 20,000 Da (Ref. 35). Because there 
is expected to be free cross-flow of these proteins across the 
membranes, the loss of the very low

[[Page 60758]]

molecular weight proteose-peptones may be small and, therefore, as 
noted in published reviews, the casein-to-whey protein ratio of milk 
would not be significantly altered during ultrafiltration (Refs. 36 and 
37). Studies also have demonstrated complete retention of whey proteins 
and a relatively constant casein-to-whey protein ratio in milk that has 
been ultrafiltered to increasing volume concentration (Refs. 13, 38, 
and 39). The information presented by Wells' Dairy, Inc., as part of 
its TMP submission also demonstrates that there is minimal, 
insignificant loss of true protein in the ultrafiltration permeate 
resulting in an ultrafiltered retentate with its casein-to-whey protein 
ratio intact (Docket No. 2004P-0519; 69 FR 71418).
    FDA notes that a comment received in response to the two petitions 
suggested that any definition of ultrafiltration also include a 
requirement that the fluid UF milk must contain a maximum of 45 percent 
total solids (or a minimum moisture content of 55 percent). The comment 
stated that this requirement is necessary to define ``liquid'' UF milk 
and preclude any treatment following ultrafiltration to further 
concentrate UF milk. However, the comment did not provide any 
supporting information or data on the appropriateness of this minimum 
level of moisture. In the proposed definition of UF milk, FDA is not 
proposing a requirement related to minimum moisture content of UF milk; 
however, the proposed definition states that UF milk is a liquid 
product. FDA seeks comment on whether there is a need for an added 
measure to ensure the liquid nature of this ingredient and/or to 
preclude any subsequent treatment following ultrafiltration to further 
concentrate the fluid UF milk. If so, does a minimum moisture content 
requirement sufficiently address this concern and what is an 
appropriate minimum level of moisture?
    FDA also seeks comment on the need for, and appropriateness of, the 
following: (1) Not permitting other forms of mechanical filtration, 
such as microfiltration; and (2) the requirement that the casein-to-
whey protein ratio remain unaltered during ultrafiltration and the 
feasibility of such a requirement for compliance and enforcement 
purposes. If the requirement that the casein-to-whey protein ratio 
remain unaltered is not appropriate, FDA seeks information on what 
constitutes an acceptable variation of this ratio during 
ultrafiltration of milk so that FDA may determine appropriate criteria 
for purposes of enforcement.
    In response to the petitions, FDA received some comments that 
opposed the use of any filtered milk, citing product safety and quality 
concerns; however, these comments did not provide any scientifically 
sound and valid data to support their objections specifically with 
regard to fluid UF milk. At this time, FDA does not have any 
information that raises food safety concerns with the use of fluid UF 
milk in standardized cheeses. FDA specifically requests that any 
comments that address the technical aspects of these proposed 
provisions include sound scientific and factual data or information 
that support the positions presented in the comments. For example, are 
there analytical data or other information that would support a 
determination that standardized cheeses made using fluid UF milk, as 
defined in this proposed rule, are potentially unsafe or are 
nutritionally inferior? Are there scientific data or information that 
demonstrate that the use of fluid UF milk, as defined in this proposed 
rule, adversely affects the physical, chemical, or sensory 
characteristics of a particular standardized cheese or cheese product 
or that would support the determination that the use of fluid UF milk 
is not appropriate in a particular standardized cheese or cheese 
product?

III. Executive Order 12866: Cost Benefit Analysis

    FDA has examined the economic implications of this proposed 
amendment for part 133 as required by Executive Order 12866. Executive 
Order 12866 directs agencies to assess all costs and benefits of 
available regulatory alternatives and, when regulation is necessary, to 
select regulatory approaches that maximize net benefits (including 
potential economic, environmental, public health and safety, and other 
advantages; distributive impacts; and equity). Executive Order 12866 
classifies a rule as significant if it meets any one of a number of 
specified conditions, including: Having an annual effect on the economy 
of $100 million, adversely affecting a sector of the economy in a 
material way, adversely affecting competition, or adversely affecting 
jobs. A regulation is also considered a significant regulatory action 
if it raises novel legal or policy issues. FDA has determined that this 
proposed rule is a significant regulatory action as defined by 
Executive Order 12866.

A. Need for Regulation

    Under current standards of identity for cheese and cheese products, 
the definitions of ``milk'' and ``nonfat milk'' do not encompass 
``filtered milk''. As a result, while these definitions list milk, 
nonfat milk, and the different forms (including concentrated, 
reconstituted, and dried) that can be used in making standardized 
cheeses, they do not explicitly permit the use of filtered milk as an 
ingredient in standardized cheeses. The use of filtered milk in cheese 
making provides greater flexibility and potential cost savings to 
cheese producers while still preserving the basic nature and essential 
characteristics of the food. FDA tentatively concludes that revision of 
the standard is needed to promote honesty and competition in the 
interest of consumers and to allow dairy producers to utilize a safe 
and effective technology.

B. Background and Current Industry Practices

    The sources for this analysis were compiled from food research and 
chemistry journals, milk and cheese industry publications, U.S. 
Department of Agriculture (USDA) data and reports, other government 
agency reports, and expert opinions. Sources cited in this text refer 
to the specific passage or data reported, but all sources found at the 
end of the document were used to formulate the basis of the analysis.
    The standardization of casein and fat content in milk is a common 
practice in cheese production that improves the consistency of the 
final products, reduces the volatility of total milk ingredient costs, 
and increases the amount of cheese produced per vat (Ref. 9). Not all 
cheese producers standardize their milk, but the amount of protein, 
specifically in the form of casein, present in milk for cheese 
production is the single largest factor affecting cheese yield. 
Condensed skim milk and NFDM are widely used to increase the amount of 
casein in cheese milk (Refs. 9 and 40). In 2001, the dairy industry 
purchased 621 million lb of NFDM, 67.5 percent of all domestic sales of 
NFDM. The use of NFDM in hard cheeses made up 43.3 percent of the total 
amount purchased by the dairy industry, and cottage and cream cheeses 
accounted for an additional 6.2 percent (Ref. 41).
    By adding condensed milk or NFDM the cheese producer is adding 
lactose and minerals that must later be removed from the curd at a 
greater rate than the casein that provides the benefits (Ref. 40). 
Ideally, cheese producers would standardize their cheese milk with a 
higher concentration of protein without adding components that later 
have to be removed. The key components of milk products used in cheese 
making are listed in table 1 of this document.

[[Page 60759]]



                                     Table 1.--Composition of Milk Products
----------------------------------------------------------------------------------------------------------------
                                     Nonfat Dry    Fluid UF Milk   Dry UF Milk     Fluid MF     Isolated Casein
     Component\1\        Milk (%)     Milk (%)          (%)            (%)        Milk (%)\2\         (%)
----------------------------------------------------------------------------------------------------------------
Protein                       3.3          36     4.48-11.94      42-80                  7.9   89-94
----------------------------------------------------------------------------------------------------------------
Fat                           3.65          0.8   5.51-14.68      1-2.5                 10.5   1.5\3\
----------------------------------------------------------------------------------------------------------------
Lactose                       4.75         52     4.59-3.68       46-4.1                 4.7   0-0.2\3\
----------------------------------------------------------------------------------------------------------------
\1\ Percentages compiled from the Wisconsin Center for Dairy Research and the Wisconsin Milk Marketing Board
  White Paper (2001), Fassbender (2001), Innovations in Dairy (2001), and GAO (2001).
\2\ As in the case of fluid UF milk, the composition of fluid MF milk can vary but we were unable to find a
  range of values of protein, fat, and lactose content of fluid MF milk in the literature.
\3\ Maximum values.

    Table 1 of this document, reflects the fact that UF milk can be 
concentrated to a greater or lesser extent to meet the needs of 
different manufacturing processes. For some cheeses, the UF milk can be 
highly concentrated then mixed with cream to produce a liquid 
``precheese'' with the same gross composition as the final cheese. It 
has been shown that this precheese can be used in continuous process 
cheese making without the use of vats (Refs. 10 and 42). Some soft 
cheeses, processed cheese, and direct acidified cheese, particularly 
those made from goat's and sheep's milk, have been reported to be 
successfully produced using highly concentrated UF milk (Refs. 12, 13, 
and 43). However, the high concentration of the retentate may affect 
some properties of the milk and require specially designed equipment 
(Ref. 2).
    More widely accepted for the common styles of cheese consumed in 
the United States appears to be the use of lower concentrations of UF 
milk to standardize the protein concentration in cheese milk to produce 
higher final cheese yields (Refs. 4, 10, and 44). Low concentration UF 
milk replaces a percentage of milk, usually between 10 and 20 percent, 
to provide a higher level of casein in the cheese milk without the 
addition of lactose and minerals (Ref. 40). Most of the benefits of 
using UF milk are from standardizing the protein concentrations while 
still allowing conventional cheese-making equipment to be used, or 
easily adapted for use (Ref. 10). Other uses include UF milk 
replacement to eliminate the natural seasonal variation in milk 
quality, improving the consistency of cheese (Ref. 9).
    For the purpose of the economics analysis, and without making any 
declarations about what FDA believes is technically sufficient, we use 
a low concentration of UF milk with approximately 10 percent 
replacement as the appropriate reference for 80 percent of all cheese 
made in the United States. This is based on research that suggests that 
low concentration replacement has been successfully used in Cheddar and 
Mozzarella cheeses (Refs. 1 and 9), whereas continuous process cheese-
making from high concentration UF milk was not (Ref. 9). These two 
cheeses alone made up two-thirds of domestic cheese production in 2002 
with Swiss and other American cheeses, making up an additional 13 
percent (Ref. 45). If this proposed rule is finalized, all standardized 
cheese made in the United States, regardless of the variety and 
including those that implement UF technology, must continue to meet the 
physical and chemical properties specified in the standard.
    Amending the standard of identity of cheese has the potential to 
affect two related sectors of the dairy industry: Dairy processors and 
cheese producers. Milk is produced on dairy farms daily, with the 
volume and composition varying both seasonally and daily. The milk is 
picked up from dairy farms and transported by milk haulers to 
cooperatives or proprietary operations for distribution or further 
processing. Large dairy farms may encompass production, processing, and 
even hard-product manufacturing facilities all at one site, whereas 
other dairy farms may belong to a cooperative or sell their milk to a 
proprietary operation that processes or further distributes the milk at 
its own discretion. Except in the cases of large operations, dairy 
farms do not usually process their own milk. Therefore, while there are 
almost 92,000 dairy operations (an operation is a place with one or 
more milk cows; a farm may include more than one operation) in the 
United States (Ref. 46), the unit of measurement for purchasing UF 
technology is the dairy processor who collects milk from one or more 
dairy operations. In addition to making the capital investment in UF 
technology, dairy processors would benefit from the decreased costs for 
transporting and storing UF milk during shipment to cheese producers.
    Cheese producers, while not the direct purchasers of UF technology, 
would still be affected by the changes in the definition of milk in 
standardized cheese if they choose to replace some of their ingredient 
milk with UF milk. Many of the benefits of using UF milk in cheese 
accrue to the cheese producers directly, including, e.g., higher cheese 
yields and increased production efficiency as well as a greater ability 
to eliminate the natural variation in their milk supplies, and reduced 
storage costs.
    Dairy processors and cheese producers are not mutually exclusive 
categories. A dairy processor is a manufacturer of dairy products made 
using milk as the main dairy ingredient. Therefore, cheese producers 
are all dairy processors, but not all dairy processors produce cheese. 
In 2002 there were 403 cheese plants and 1,153 dairy processors in the 
United States (Ref. 45). Some dairy processors either manufacture 
cheese directly or manufacture dairy products that are sold to cheese 
producers. However, some dairy processors produce no cheese products or 
ingredients whatsoever, and instead, produce a variety of other dairy 
products including fluid milk, butter, ice cream, and whey products. It 
is also worth noting that dairy processors include cooperatives. In 
1997 there were 226 dairy cooperatives that ranged in primary function 
from bargaining-only to hard-product manufacturing and fluid processing 
(Ref. 47).
    We measure benefits as the net decrease in the cost of producing 
cheese. These benefits accrue from all types of protein-
standardization; however, the extent of the benefits will vary 
depending on the milk product used. These benefits lead to cost savings 
that could be passed along to consumers if the market is opened to a 
larger number of dairy producers within the industry and competition 
among cheese producers is enhanced. When only those milk processors 
that are large

[[Page 60760]]

enough to incorporate UF technology in legitimate alternate-make 
procedures (i.e., within plant and within batch) are allowed to use the 
cost-saving technology in standardized cheeses, they will be able to 
sell their goods at the market price, which is based on competition 
among firms with higher production costs. If, however, the market is 
broadened so that all firms, large and small, are able to use the cost-
saving technology, competition among these firms should bid down the 
market price of cheese, passing the savings on to consumers.
    We measure the costs of using filtered milk to make standardized 
cheese as losses to consumers who prefer cheese made under the existing 
milk definitions, domestic and international market adjustments, and 
government purchases required under USDA's Commodity Credit Corp., 
program. Increases in government purchases of dairy products will not 
incur unless the market prices of specific products fall below the 
government floor prices.

C. Regulatory Options

    We analyze several options for amending the standards of identity 
for cheeses and cheese products. Option 1 would amend the definition of 
milk in the standards of identity for cheeses to allow fluid UF milk to 
be used. Option 2 would allow fluid UF milk and dry UF milk. Option 3 
would amend the definition of milk in the standards of identity for 
cheese to allow all filtration methods that resulted in a fluid milk 
product to be used in cheese production. Option 4 would allow all 
filtration methods that resulted in fluid or dry milk products to be 
used. Option 5 would allow all milk or products obtained from milk to 
be used in cheese production, in concert with the Codex general 
standard for cheese.
    We estimate the benefits and costs of the regulatory option 
compared with the benefits and costs of a baseline. The baseline 
reflects the state of the industry before any new regulation is put in 
place. Therefore, in this analysis the baseline is leaving the standard 
of identity for cheese unmodified, i.e., milk, nonfat milk, and the 
concentrated, reconstituted, and dried forms of milk and nonfat milk 
are the only basic ingredients allowed in the production of 
standardized cheese. Due to the ``extensive use of nonfat dry milk 
(NFDM) as an ingredient for cheese manufacture in the United States'' 
(Ref. 9), the baseline assumes NFDM is used as the source of 
supplemental solids in cheese manufacture. For purposes of this 
analysis, we assume that the benefits and costs of the baseline are 
zero.
Option 1: Allow fluid UF milk to be used in the making of standardized 
cheeses
    This option would allow fluid UF milk to be used in the making of 
standardized cheese. For most U.S. cheese production, this option would 
result in replacing a percentage of the milk used in the production of 
cheese with fluid UF milk. This option differs from the baseline by 
substituting fluid UF milk for NFDM as the protein-dense replacement 
milk ingredient.
    Benefits of Option 1: Fluid UF milk retains more moisture from milk 
than NFDM does, so as a percentage of total composition, UF milk has 
less protein than NFDM. However, it also contains less lactose than 
either NFDM or milk. In fact, the more highly concentrated the milk is, 
(the concentrations listed in table 1 of this document, vary from 1.5 
to 4 times the solids concentration of milk), the more protein is 
retained and the less lactose is unnecessarily added. Replacement of 
milk with fluid UF milk during the manufacturing process produces yield 
increases per vat, thus spreading out fixed costs (labor, equipment, 
physical facility) over more total weight of cheese (Ref. 9). According 
to the Technical Director of North American Milk Products, a cheese 
plant that replaces 10 percent of its daily milk inputs with fluid UF 
skim milk would see an increase in cheese yield of 12 percent. This 
increase in yield lowers costs by up to two cents per pound of cheese 
(Ref. 48). In 2002, 8.6 billion pounds of cheese were produced (Ref. 
45). Therefore, the yield increase due to partial replacement of milk 
with fluid UF milk in all U.S. cheese production could save about $172 
million per year ($0.02 per pound x 8.6 billion pounds).
    This estimate may understate the potential cost savings; Fassbender 
(Ref. 49) states that a 10 percent replacement produces a yield 
increase of 25 percent, and an article from Dairy Management, Inc., 
states that a 10-15 percent replacement produces a yield increase as 
high as 18 percent (Ref. 50). In addition, the amount of rennet and 
starter cultures which are added to cheese milk can be reduced due to 
the higher solids content in the cheese milk. In one fluid UF milk 
research study at the Wisconsin Center for Dairy Research, a plant was 
able to reduce the rennet usage by 4 ounces per vat, for a total annual 
savings of over $28,000 (Ref. 49). If we assume this plant is 
representative of all cheese manufacturing plants, then multiplying 
$28,000 by the 403 cheese plants in 2002 (Ref. 45) gives a rough figure 
of $11 million savings in coagulant usage annually. FDA notes that 
these estimates are uncertain and seeks comment on the cost savings 
from rennet and starter cultures.
    Estimating the net social benefits from implementing UF technology 
requires subtracting out the private costs to firms of making the 
necessary capital investments. Milk is increasingly being ultrafiltered 
during the processing stage, usually at manufacturing plants or dairy 
cooperatives, so we assume that no capital investment in equipment by 
the cheese maker is needed to take advantage of UF technology for low 
level fluid UF milk concentration replacement (Ref. 48). Cheese 
producers can simply replace a portion of milk with fluid UF milk 
purchased from a dairy processor without purchasing new equipment.
    An early cost-benefit analysis of fluid UF milk production by 
Slack, et al. (Ref. 51), found that the benefits of UF milk production 
outweighed the costs for dairy farms with over 100 cows. However, this 
threshold has likely changed as the latest Pasteurized Milk Ordinance 
(April 2003 edition) loosened the restriction that allowed only single 
pass UF systems to now allowing for less expensive recirculating UF 
systems. Informal conversations with industry representatives revealed 
that the smallest single pass UF systems being marketed can process 
300,000 lb of milk per day, the equivalent of production from almost 
5,000 cows (300,000 lb is roughly 34,800 gallons, which at 7 to 8 
gallons per cow per day, is 4,350 to 4,971 cows). Recirculating 
systems, on the other hand, are available for flow rates of 800 gallons 
per day, or production from approximately 100 cows (Ref. 52).
    The costs of implementing fluid UF technology differ for four 
categories of dairy processors.
     If a processor already produces fluid UF milk, there is no 
additional cost to allowing the extended definition of milk in 
standardized cheese.
     If a processor collects milk from fewer than 100 cows, UF 
technology may not be economically feasible. If cheese producers switch 
their input purchases away from milk to fluid UF milk, there might be a 
redistribution of income away from these very small dairy processors. 
FDA believes that few, if any, milk processors will fall into this 
category. Even though there are many small dairy farms (72,070 in 2002) 
milk is not necessarily ultrafiltered on-farm. Instead, small dairy 
farms have the option of combining milk with other dairy farms in 
member-owned cooperatives or selling milk to

[[Page 60761]]

proprietary operations that combine milk from several farms for 
processing. The USDA defines a ``small'' dairy cooperative as handling 
less than 50 million lb of milk each year (Ref. 53), which is roughly 
the equivalent of milk from 2,000 cows per day and well above the 100 
cow minimum.
     If a processor collects milk from more than 100 cows but 
less than 4,000 and is not currently producing fluid UF milk, then the 
cost of purchasing recirculating UF equipment ranges from $175,000 to 
$350,000 (Ref. 52).
     If a processor collects milk from 4,000 or more cows and 
is not currently producing fluid UF milk, then the cost of purchasing 
UF equipment ranges from $350,000 for a recirculating system to 
$1,372,500 for a single-pass system (Ref. 52).
    Of the 1,153 dairy processors (which includes dairy cooperatives 
that process milk for members), an unknown portion would purchase UF 
technology in response to this proposed rule if finalized. In 2002, 
cheese production used 64,504 million lb of milk, which is 
approximately 61 percent of the 105,961 million lb used in all 
manufactured dairy products (Ref. 45). Therefore, we estimate that 
61percent of the dairy manufacturing plants process milk for cheese, 
for a total of 703 dairy plants. Given that at least 22 dairy 
manufacturing plants and 4 large dairy farms already produce fluid UF 
milk (Ref. 4), a total of 677 dairy processors may choose to purchase 
UF technology as a result of changing the definition of milk in Sec.  
133.3. Assuming that new purchases of UF equipment would more likely be 
recirculating systems, the total one time capital expenditure would 
range from $118 to $237 million. Given that the UF equipment 
depreciates over 7 to 14 years (Ref. 1), we estimated the annualized 
cost over a 10 year period. With a 3 percent interest rate, the 
annualized cost ranges from $14 to $28 million. With a 7 percent 
interest rate, the annualized cost ranges from $17 to $34 million. The 
annualized cost ranges indicate the capital expenditure ranges based on 
the equipment capacity needs described previously in this document.
    Milk is produced daily, with the volume and composition varying 
both seasonally and daily. Demand for dairy products also varies both 
seasonally and daily, but demand variations are not correlated with 
supply variations (e.g., milk production peaks in the spring, but 
demand for milk and butter peaks in the fall months) (Ref. 53). Cheese 
producers, however, need to provide a consistent quality cheese, 
regardless of the day or season in which the inputs were produced. 
Replacing a given portion of milk with UF milk can eliminate the daily 
variation that occurs in milk composition by standardizing the ratio of 
casein to fat. However, fluid UF milk does not offer any price 
stability from seasonal fluctuations that occur in the supply and 
demand for both milk and cheese, since it cannot be stored past the 
short term in a liquid form. Nonfat dry milk has a shelf-life of 12 to 
18 months (Ref. 50) and may offer more price stability from seasonal 
fluctuations.
    The transportation and storage costs associated with fluid UF milk 
are lower than milk due to the removal of approximately two-thirds of 
the water, lactose, and ash during the filtration process (Ref. 48). 
The 2001 GAO Report cites a shipment of fluid UF milk by Select Milk 
Producers, Inc., in which the cost of transporting fluid UF milk was 73 
percent lower than the cost of transporting milk. In this same year, 
milk hauling charges in the Upper Midwest Marketing Area (which 
includes California and Wisconsin, the top two milk producing states) 
averaged 17.1 cents per hundredweight (cwt) of milk (Ref. 54). A 73 
percent price reduction in this average hauling cost lowers the cost of 
hauling fluid UF milk to an average of 4.62 cents per cwt. As stated in 
the section I of this document, we assume that for approximately 80 
percent of the cheese produced in the United States, fluid UF milk is 
used as a substitute in cheese production, not for milk, but for the 
baseline standardizing ingredient, NFDM. To calculate the 
transportation savings for these cheeses, we take the 64,504 million 
pounds of milk used in cheese production in 2002 (Ref. 45) and multiply 
by 80 percent to capture the amount shipped for American style natural 
cheeses. We then calculate 10 percent of this total to be replaced by 
fluid UF milk and convert it to cwt. This is the amount of milk that is 
subject to a 73 percent reduction in shipping costs, giving a total 
annual cost savings of about $7 million as follows:

  Calculation of Transportation Cost Savings for Fluid UF Milk Used in
                      American Style Natural Cheese
 80% X 64,504       =        51,603 million lb of milk shipped for
 million lb                   American cheese production
 10% X 51,603       =        5,160 million lb of milk filtered before
 million lb                   shipment to cheese factory
 5,160 million lbs/ =        51.6 million cwt of milk filtered before
 100 lb                       shipment
 73% of 17.1 cents/ =        $0.13 savings per cwt of fluid UF milk
 cwt                          shipped
 $0.13 X 51.6       =        $6.7 million
 million cwt
 

    There would be an additional transportation and storage cost 
savings for the varieties of cheese that are well-suited to high 
concentrations of UF milk where replacement values are closer to 100 
percent of the original milk. To get a potential range for what this 
cost savings would be, we calculated the transportation savings 
assuming that the remaining 20 percent of cheese production would use 
only UF milk for an upper bound and assuming only 2 percent of cheese 
production would replace 100 percent of milk in cheese production as a 
lower bound. The annual transportation savings here range from $2 to 
$17 million (See below).

Calculation of Upper Bound of Transportation Cost Savings for 100% Fluid
                           UF Milk Replacement
 20% X 64,504       =        12,901 million lb of milk shipped for all
 million lb                   other cheese production
 100% X 12,901      =        12,901 million lb of milk filtered before
 million lb                   shipment to cheese factory
 12,901 million lb/ =        129 million cwt of milk filtered before
 100 lb                       shipment
 73% of 17.1 cents/ =        $0.13 savings per cwt of UF milk shipped
 cwt
 $0.13 X 129        =        $16.8 million
 million cwt
 


Calculation of Lower Bound of Transportation Cost Savings for 100% Fluid
                           UF Milk Replacement
 2% X 64,504        =        1,290 million lb of milk shipped for other
 million lb                   cheese production
 100% X 1,290       =        1,290 million lb of milk filtered before
 million lb                   shipment to cheese factory
 1,290 million lbs/ =        12.9 million cwt of milk filtered before
 100 lb                       shipment
 73% of 17.1 cents/ =        $0.13 savings per cwt of UF milk shipped
 cwt
 $0.13 X 12.9       =        $1.7 million
 million cwt
 


[[Page 60762]]

    In terms of total transportation cost savings for all cheese 
production, this calculation gives an annual savings between $9 and $24 
million for replacing milk with fluid UF milk in cheese production. 
While this is a cost savings over using milk in cheese production, it 
is not a savings over using NFDM. Reducing the moisture content of milk 
by two-thirds reduced the shipping costs by 73 percent, so it is 
reasonable to assume that NFDM with only 3.2 percent moisture (Ref. 40) 
and an increased shelf-life of 12 to 18 months (Ref. 50) would be 
significantly less expensive to ship and store than UF milk. Compared 
with the baseline then, these savings would be reduced by an amount in 
excess of $7 million due to the actual increase in costs from replacing 
NFDM with fluid UF milk.
    The total annual benefits from using fluid UF milk to make 
standardized cheeses are uncertain, partly because the number of 
additional plants that would use the UF technology is uncertain. The 
cost savings also depend on the size of the plants that decide to 
invest, the amount of milk which cheese producers replace with fluid UF 
milk, and whether fluid UF milk replaces milk or NFDM in the production 
process. If all dairy plants switch to UF technology, the yield and 
coagulant savings would be high, but investment costs would also rise. 
If most plants already use this technology, or decide against 
investing, the yield, coagulant, and transportation savings would be 
low. If NFDM is not extensively used in current cheese production, the 
transportation savings will be greater. Finally, if larger plants 
already have UF technology the total capital investment costs will 
decrease but yield increases will not be as dramatic as only smaller 
systems will potentially invest as a result of changing the definition.
    In addition to the technical benefits in cheese production from 
allowing fluid UF milk to be used in standardized cheese production, 
amending the standards offers another economic benefit. Specifically, 
allowing fluid UF milk to be used as an ingredient in cheese would open 
the benefits of UF technology to a wider range of cheese manufacturers. 
Currently, fluid UF milk can be used in standardized cheese production 
only under ``alternate make'' procedures. Under the alternate make 
procedure provisions, manufacturers of cheese who purchase or produce 
milk in sufficient quantity to use UF technology may substitute the 
ultrafiltration of milk as a step in the cheese-making process as long 
as the final finished cheese has the same physical and chemical 
properties as the cheese produced under the procedure specified by the 
standard of identity. This provision only allows for the use of 
alternate procedures and not for alternate ingredients. Therefore, the 
use of UF technology must be within plant and within batch; fluid UF 
milk purchased from another plant, even within the same company, is 
considered an alternate ingredient. Allowing fluid UF milk as an 
ingredient effectively removes the barriers to shipment of fluid UF 
milk to cheese producers throughout the country and allows for greater 
competition in the market for cheese ingredients.
    As stated previously in this document, approximately 22 dairy 
manufacturing plants and four large dairy farms produce UF milk. It is 
difficult to ascertain how much of the UF milk is being used within 
plants under alternate make procedure provisions, and how much is being 
shipped to outside plants. Few records are kept either by the USDA or 
trade associations regarding intermediate products like fluid UF milk 
(See GAO report). In 1996, the FDA permitted a single New Mexico plant 
to produce cold UF milk for shipment to a cheese-making plant in 
Minnesota for trial purposes only. Subsequently, the New Mexico plant 
is said to have increased shipments of UF milk to 15 plants throughout 
the country (Ref. 49).
    Allowing fluid UF milk to be used in standardized cheese production 
could significantly increase the number of plants using this cost-
saving technology, particularly among smaller operations that cannot 
currently afford to purchase UF technology. These smaller cheese 
producers that cannot afford to filter milk as a step in the production 
process could purchase UF milk from a dairy processor. In 2002, there 
were 403 cheese plants and 1,153 dairy manufacturing plants spread 
across all fifty states (Ref. 45) but only 26 dairy plants and farms 
were producing UF milk. The supply of UF milk is restricted by the 
current definition, potentially increasing its cost as an input to 
cheese production.
    Costs of Option 1: There are no health costs associated with the 
lower production costs of cheese made with fluid UF milk.
    If consumers prefer cheese made under the existing milk definition 
and if they purchase cheese made from fluid UF milk believing it to be 
made from milk under the existing definition, there will be a small 
cost incurred by the consumer. However, even though the total dollar 
amount spent on cheese is large (in 2000, the retail price of 1 lb of 
natural cheddar cheese was $3.83 (Ref. 55) and 8.2 billion lb of all 
cheeses (excluding cottage cheeses) were produced (Ref. 45), for total 
consumer expenditure of $31.4 billion) the costs incurred from fluid UF 
milk are likely to be low because standardized cheeses do not tend to 
have credence attributes. Credence attributes are characteristics that 
consumers are willing to pay more for, even though they are not 
detectable after consumption (e.g., ``dolphin-safe'' tuna). The growth 
in the dairy products over the past 20 years has been largely 
attributed to increased demand for pizza and fast food products that 
contain cheese, particularly Mozzarella and American cheese (Ref. 56). 
These are not the varieties of cheese that tend to be associated with 
cheese connoisseurs who demand purity in cheese ingredients. There is 
no evidence that consumers place a premium on cheeses made under the 
existing definition, in particular because cheese made with UF 
technology must have the same physical and chemical properties as 
cheese made under the existing milk definition and because an unknown 
quantity of cheeses produced in the United States are already made 
using UF technology under the alternate make procedure provisions.
    The U.S. dairy market is regulated under both Federal and State 
regulations. The U.S. Government provides price supports for domestic 
milk production under the USDA's Commodity Credit Corp. A potential 
drop in the demand for milk as cheese producers switch to fluid UF milk 
could result in the market price dropping below the support price, thus 
forcing the government to purchase a larger amount of milk. However, 
fluid UF milk is produced by separating the components of milk. 
Therefore, any decrease in the domestic demand for milk resulting from 
the production and sale of fluid UF milk will be off-set by a decrease 
in the supply of milk, as dairies ultrafilter some of their milk 
instead of selling it directly. As a result, the quantity of milk 
purchased by the government is left unchanged. Stated another way, if 
cheese producers purchase fluid UF milk instead of other milk, the 
demand for milk from cheese producers will fall, while the demand for 
fluid UF milk from cheese producers will rise. As a result, the dairy 
processors who find it profitable to do so will decrease their supply 
of milk and instead ultrafilter the milk before they sell it to the 
cheese producer. If no dairy processors find it profitable to 
ultrafilter their milk before selling it, then cheese producers will 
have no choice but to purchase milk, again

[[Page 60763]]

leaving the amount purchased by the government unchanged.
    In addition, the U.S. Government provides export subsidies under 
the Dairy Export Incentive Program. Fluid UF milk is less expensive to 
transport than milk under the standard definition of milk in cheese, 
leading to fears that expanding the use of fluid UF milk may increase 
imports and further decrease the demand for domestic milk. As of the 
first 9 months of 2002, all UF milk imported into the United States was 
in a dry powder form categorized as MPC (Ref. 57). Therefore, allowing 
the use of fluid UF milk as an ingredient in the standard of identity 
of cheese should not cause foreign-produced UF milk to replace domestic 
milk in cheese production or cause U.S. Government purchases under the 
Commodity Credit Corp. to rise.
Option 2: Allow fluid and dry UF milk in standardized cheese production
    This option would allow UF milk either in fluid or spray-dried 
form. Dry UF milk is often referred to as MPC, though the definition of 
MPC is not consistently used and sometimes includes other dried 
filtered or concentrated milk products. This option differs from the 
baseline and Option 1 by substituting dry UF milk for NFDM or fluid UF 
milk as an ingredient in standardized cheeses.
    Benefits of Option 2: The protein composition of dry UF milk ranges 
from 42 percent to 80 percent (Ref. 40), depending on the degree of 
concentration. In addition, as the protein concentration increases, the 
lactose content decreases from 46 percent to just 4.1 percent at the 
highest concentrations. Therefore, the supplementation of cheese milk 
with dry UF milk during the manufacturing process produces even larger 
yield increases per vat than fluid UF milk or NFDM, thus further 
spreading out fixed costs (labor, equipment, physical facility) over 
more total weight of cheese. Given these larger cheese yield increases 
over fluid UF milk, it is safe to assume that the total yearly savings 
from using dry UF milk would exceed $172 million. In addition, the 
amount of rennet and starter cultures which are added to cheese milk 
can be reduced due to the higher solids content in the cheese milk. The 
rough figure of $11 million savings in coagulant usage annually 
calculated in Option 1 is applicable here as well.
    Calculating the net social benefits to implementing UF technology 
requires subtracting out the private costs to firms of making the 
necessary capital investments. Similar to fluid UF milk, dry UF milk 
production occurs at the processing stage, usually at manufacturing 
plants or dairy cooperatives, so we assume no capital investment in 
equipment by the cheese producer is needed to take advantage of dry UF 
technology for low concentration UF milk replacement. Cheese producers 
can simply replace a portion of milk with dry UF milk purchased from a 
dairy processor without purchasing new equipment.
    The costs of implementing dry UF technology varies among different 
types of dairy processors and will depend on their current production 
technology. If a dairy processor already produces UF milk and NFDM, 
there is no additional cost to allowing the extended definition of milk 
in standardized cheese. If a processor collects milk from fewer than 
100 cows, it may not be economically feasible to implement the UF 
process, making dry UF milk production impossible even if the dairy 
processor has appropriate drying technology. If a dairy processor 
collects milk from 100 to 4,000 cows and is not currently producing UF 
milk, then the cost of implementing a UF system ranges from $175,000 to 
$350,000, depending on the size of the plant. If a processor collects 
milk from 4,000 or more cows and is not currently producing UF milk, 
then the cost of purchasing UF equipment ranges from $350,000 for a 
recirculating system to $1,372,500 for a single-pass system. Using the 
same method as Option 1, the total one time capital expenditure for 
dairy processors who sell their products to cheese producers would be 
$118 to $237 million. If the dairy processor does not own a spray 
dryer, additional capital costs would be necessary, on the order of 
$750,000 (Ref. 58). If half of all 703 dairy plants had to purchase 
this equipment, the one-time capital expenditure would grow by $264 
million for a total of $382 to $501 million. Given that the UF 
equipment depreciates over 7 to 14 years (Ref. 1), we estimated the 
annualized cost over a 10-year period. With a 3-percent interest rate, 
the annualized cost ranges from $45 to $59 million. With a 7-percent 
interest rate, the annualized cost ranges from $54 to $71 million. The 
annualized cost ranges indicate the capital expenditure ranges based on 
the equipment capacity needs described previously in this document.
    Similar to NFDM, spray-drying UF milk significantly increases the 
shelf-life of the milk. Using such milk powders can eliminate the 
natural daily and seasonal variation that occurs in milk composition 
(by standardizing the ratio of casein to fat). In addition, the ability 
to store dry UF milk allows the cheese producer to offset the 
volatility of fresh milk prices (Ref. 9) and be better able to balance 
seasonal imbalances than milk or fluid UF milk.
    The transportation and storage costs associated with dry UF milk 
are lower than either milk or fluid UF milk due to the removal of 
approximately 95 percent of the water, lactose, and ash (Ref. 40) 
during the ultrafiltration and subsequent drying processes. The 
moisture content of dry UF milk is similar to that of NFDM; therefore, 
it is reasonable to assume that shipping and storage costs would also 
be similar for replacing NFDM with dry UF milk in protein 
standardization. If NFDM is not being used for protein standardization, 
then dry UF milk could offer substantial benefits compared to the 
transportation and storage of milk, possibly reducing these costs up to 
95 percent.
    A review of the literature found no manufacturers of dry UF milk in 
the United States; however, informal conversations with industry 
representatives revealed one joint venture in New Mexico that currently 
produces dry UF milk and possibly another firm in New York (Ref. 59). 
Little is known about the cost of producing dry UF milk, and why there 
is little to no U.S. production is a matter of some debate. The price 
floor set by the U.S. Dairy Price Support Program for NFDM is often 
cited as the cause. At the current levels of government purchase prices 
for milk protein, U.S. manufacturers of dry UF milk products would 
obtain the same or lower return per pound of protein than they would 
for producing NFDM. Given the higher manufacturing costs associated 
with UF technology, dairy producers in the United States are often 
better off producing NFDM and selling it to the government than 
producing dry UF milk products for cheese and other food uses (Ref. 
60). Foreign firms who currently export dry UF milk to the United 
States have greater incentive to open their own plants in the United 
States, as it would reduce their transportation and tariff costs.
    Costs of Option 2: There are no health costs associated with the 
lower production costs of cheese made with fluid or dry UF milk.
    If consumers prefer cheese made under the existing milk definition 
and if they purchase cheese made from dry UF milk believing it to be 
made from milk under the existing definition, there will be a small 
cost incurred by the consumer. However, even though the total dollar 
amount spent on cheese is large (about $31.4 billion in 2000) the costs 
incurred from dry UF milk are likely to be low because standardized 
cheeses do not tend to have credence

[[Page 60764]]

attributes and there is no evidence that consumers place a premium on 
cheeses made under the existing definition. Cheese made with UF 
technology must have the same physical and chemical properties as 
cheese made under the existing definition of milk within the cheese 
standards.
    There is some concern over whether allowing dry UF milk (presumably 
imported from other countries) in the definition of milk in cheese 
would displace purchases of other dairy substitutes that are 
domestically produced. A drop in the demand for milk or NFDM as cheese 
producers switch to purchasing dry UF milk could result in the market 
price dropping below the support price, thus forcing the government to 
purchase a larger amount of milk. In addition, since dry UF milk is 
much less expensive to transport than milk and even fluid UF milk, 
expanding the use of dry UF milk may increase imports and further 
decrease the demand for domestic milk.
    As in the case with fluid UF milk, if domestic production of dry UF 
milk increases as a result of the change in definition, any decrease in 
the domestic demand for milk resulting from the production and sale of 
dry UF milk would be offset by an decrease in the supply of milk, as 
dairies ultrafilter and dry some of their milk instead of selling it 
directly. As a result, the quantity of milk purchased by the government 
would be left unchanged. However, unlike fluid UF milk, dry UF milk is 
imported from other countries with no restrictions on the quantity and 
under a very low tariff rate (Ref. 60). The U.S. Government does not 
directly support the price of dry UF milk under the Credit Commodity 
Corp., purchases; however, if foreign-produced dry UF milk is 
substituted in production for NFDM and other milk products, increases 
in dry UF milk imports would cause government purchases of dairy 
products to increase. If, on the other hand, allowing dry UF milk to be 
used in the production of standardized cheese causes domestic 
manufacturers of NFDM to produce dry UF milk instead, the amount of 
government purchases of NFDM may actually decrease as resources shift 
to the new product.
    The inconsistency with which the term MPC is used makes it 
difficult to discern how much foreign-produced dry UF milk is being 
imported into FDA's Operational and Administrative System for Import 
Support (OASIS) database includes MPC as a separately identifiable 
product; however, many dried dairy substances other than dry UF milk 
are also included in this category, including milk protein isolate, 
whey protein concentrate, whey protein isolate, casein, milk protein 
stabilizer, emulsifier or binder, peptones, and total milk proteinate. 
Without a standard definition for MPC it is not clear that even imports 
labeled specifically as MPC are 100 percent dry UF milk.
    In his analysis of MPC imports and the commercial disappearance of 
NFDM, Jesse (Ref. 60) separated the concentrated milk protein imports 
into the following four categories: MPC, Casein-MPC, Casein, and 
Caseinates/Other Casein Derivatives. Then, looking only at the category 
of MPCs, imports increased steadily between 1989 and 1997, at a rate of 
about 4,200 metric tons per year. From 1998 through 2000, imports 
started growing even more rapidly, with an average rate of growth at 
18,000 metric tons per year (Ref. 60). However, 2001 and 2002 saw a 
reversal of this trend, with imports falling from 52,900 metric tons in 
2000 to 28,500 metric tons in 2001 (Ref. 57). Estimates of 2002 imports 
were expected to total about 35,000 tons, about a 23 percent increase 
(Ref. 60). A news release published after the second quarter of 2003 by 
the National Milk Producers Federation states that MPC imports were up 
39 percent from the first half of 2002 and approaching year 2000 levels 
(Ref. 61).
    The impact of these imports increases in significance as USDA 
purchases more NFDM under the Commodity Credit Corp. The USDA had 1.2 
billion lb of NFDM in warehouses, and program cost overruns were almost 
$3 billion more than its original $1.3 billion estimate in mid-2003 
(Ref. 62). The negative impact on dairy production in the United States 
attributable to the MPC imports is uncertain, according to Jesse (Ref. 
60) somewhere between ``an amount much smaller than government 
purchases'' of NFDM to an amount that ``exceeds government purchases, 
and that excess cheese supplies augmented by MPC and other milk 
proteins have depressed the cheese market.'' He estimated displacement 
of NFDM into government purchases at almost 430 million lb in 2002, 
though he added that his estimates ``very likely err on the high 
side.'' Bailey (Ref. 56), who separated ``dry whey'' and ``casein'' 
from MPCs, looked at this question from a cost angle. He estimated that 
MPC imports between 1996 and 2000 increased the cost the dairy price 
support program by about $572 million (Ref. 56).
Option 3: Allow all filtration methods that result in a fluid milk 
product to be used in standardized cheese production
    This option would allow fluid UF milk as well as milk processed 
with other filtration technologies, most notably microfiltration, as 
long as no nonmilk derived ingredients are added in the preparation of 
the liquid concentrates. This option differs from the baseline by 
permitting the substitution of fluid UF and MF milk for NFDM. This 
technology and the resulting product, sometimes referred to as Native 
Milk Casein Concentrates, is not currently available. However, the 
availability of the ingredient may be driven by outside food 
manufacturers who fractionate milk proteins to harvest milk serum 
proteins leaving the native milk casein concentrate for sale to cheese 
manufacturers in the near future (Ref. 9).
    Benefits of Option 3: The benefits from allowing fluid MF milk as 
an ingredient in cheese manufacture are similar to the benefits from 
allowing fluid UF milk due to similar levels of protein, lactose, and 
moisture (Ref. 63) (see table 1 of this document). There are other 
potential benefits from fluid MF milk that fluid UF milk does not 
offer. First, microfilters have larger pore structures than 
ultrafilters, allowing more whey proteins to pass through the membrane. 
If the cheese producers are purchasing MF milk, they will have less 
whey to remove in later steps of the cheese-making process. Second, 
some industry experts believe that MF is the new direction of cheese 
fortification process because it has the potential for continuous 
cheese making without vats for more varieties of cheese (Refs. 9 and 
64).
    Costs of Option 3: Because fluid MF milk is not yet available to 
cheese makers, it is difficult to determine how the costs would differ 
from NFDM. Because of the similar process to producing fluid UF milk, 
the costs are assumed to also be similar to Option 2.
Option 4: Allow all filtration methods that result in a fluid or dried 
milk product to be used in standardized cheese production
    This option would allow milk used in the production of cheese to be 
supplemented with UF milk as well as milk forms derived from other 
filtration technologies, most notably microfiltration, as long as no 
nonmilk derived ingredients had been added in the preparation of these 
liquid or dried concentrates. This option differs from the baseline by 
substituting both fluid and dry UF and MF milk for NFDM as the protein 
standardization ingredient. As with fluid MF milk, this technology and 
the resulting product, sometimes referred to as Native Milk Casein 
Concentrates, is not currently available. However, the availability of 
the ingredient may be driven by outside

[[Page 60765]]

food manufacturers who fractionate milk proteins to harvest milk serum 
proteins, leaving the native milk casein concentrate for sale to cheese 
manufacturers in the near future (Ref. 9).
    Benefits of Option 4: The benefits of allowing fluid or dry MF milk 
as an ingredient in cheese build on the benefits of Option 3, which 
allows for fluid MF milk. In addition to those benefits, allowing dry 
MF milk has decreased transportation and storage costs similar to NFDM 
and dry UF milk.
    Costs of Option 4: Because neither fluid nor dry MF milk is 
available to cheese producers, we are unable to estimate how costs 
would differ from NFDM. Dry MF milk, being similar in manufacture to 
dry UF milk, would be subject to similar costs, including foreign trade 
and domestic purchase adjustments.
Option 5: Allow all milk and products obtained from milk to be used in 
cheese production, in agreement with the Codex general standard for 
cheese
    This option would allow milk to be manufactured with ``milk and/or 
products obtained from milk'' and would mirror the Codex general 
standard for cheese (Ref. 25). This option differs from the baseline by 
allowing any milk derived ingredient to be used as either the sole 
ingredient or the protein-standardizing replacement ingredient in 
cheese production. This option would include isolates of casein that 
contain up to 94 percent protein and little to no lactose. These 
isolates are not currently manufactured in the United States, but have 
been used in other countries as a fortification ingredient (Ref. 9). 
This option would also allow for dry blends of different milk derived 
ingredients, including NFDM, dry UF milk, isolated casein, and whey 
protein concentrate.
    Benefits of Option 5: The benefits to opening the standard to all 
``milk and/or products obtained from milk'' are not certain, but would 
allow cheese producers full freedom in choosing inputs to maximize 
their own production yields and profits.
    Costs of Option 5: The costs to opening the standard to all ``milk 
and/or products obtained from milk'' are not certain. There may be 
domestic and international market adjustments leading to U.S. 
Government purchases of domestic dairy products.

D. Summary of Costs and Benefits

    The total annual costs and benefits from amending the definition of 
milk used to produce standardized cheeses are uncertain, though FDA 
does not have concerns from a food safety standpoint. The uncertainty 
stems from several diverse factors:
     The number of plants that would implement UF or other 
filtration technology,
     The number of plants that already use UF technology,
     The number of plants that already use spray-drying 
technology,
     The size of the plants that would decide to invest in new 
technology,
     The percent of milk that cheese producers would replace 
with UF milk in cheese making, and
     Whether UF milk replaces milk or NFDM in the production 
process
    Table 2 of this document highlights the quantified annual costs and 
benefits of Options 1 through 5 using the assumptions and calculations 
described in the text.

                  Table 2.--Costs and Benefits Summary
------------------------------------------------------------------------
                        Option 1  Option 2  Option 3  Option 4  Option 5
------------------------------------------------------------------------
Annualized              $14-$28   $45-$59   Unknown   Unknown   Unknown
Investment               million   million
                         \1\       \1\
                        $17-$34   $54-$71
                         million   million
                         \2\       \2\
------------------------------------------------------------------------
Yield Increase          $172      $172      Unknown   Unknown   Unknown
                         million   million
------------------------------------------------------------------------
Transportation Savings  < $9 to   > $9 to   Similar   Unknown
                         $24       $24       to
                         million   million   Option
                                   Similar   2
                                   to
                                   Option
                                   1
------------------------------------------------------------------------
Rennet & Starter        $11       $11       Unknown   Unknown   Unknown
 Savings                 million   million
------------------------------------------------------------------------
Benefits (net savings   $164-$19  $133-$16  Unknown   Unknown   Unknown
 in production costs)    3         2
                         million   million
                         \1\       \1\
                        $158-$19  $121-$15
                         0         3
                         million   million
                         \2\       \2\
------------------------------------------------------------------------
Government Programs     No        Potentia  Unknown   Unknown   Unknown
                         increas   l for
                         e in      increas
                         governm   e in
                         ent       governm
                         purchas   ent
                         es or     purchas
                         trade     es of
                         impacts   NFDM
------------------------------------------------------------------------
Costs (change in        None      Uncertai  Unknown   Unknown   Unknown
 government program                n
 costs)
------------------------------------------------------------------------
\1\ At 3 % interest.
\2\ At 7 % interest.

    FDA does not currently have a best estimate on the cost savings of 
this proposed rule and seeks comment on all areas of uncertainty listed 
previously in this document. FDA believes Options 1 and 2, if 
implemented, would lead to social benefits potentially as high as $190 
million at a 7 percent annualized investment rate ($193 million at 3 
percent) and $153 million ($162 million at 3 percent), respectively. 
Options 3 through 5 are difficult to quantify based on the smaller 
amount of research into new filtration and separation technologies in 
the dairy industry. These options lead to increasingly greater 
flexibility for cheese producers to maximize their own production 
yields and profits and have the potential to provide benefits to the 
cheese industry in the future.

IV. Small Entity Analysis

    FDA has examined the economic implications of this proposed rule as 
required by the Regulatory Flexibility Act (5 U.S.C. 601-612). If a 
rule has a

[[Page 60766]]

significant economic impact on a substantial number of small entities, 
the Regulatory Flexibility Act requires agencies to analyze regulatory 
options that would lessen the economic effects of the rule on small 
entities. FDA finds that this proposed rule would have a significant 
economic impact on a substantial number of small entities.
    The Small Business Administration (SBA) considers a dairy 
manufacturer, which includes cheese manufacturers, to be small if it 
employs fewer than 500 workers. Table 3 of this document lists the 
dairy manufacturing statistics by employment size from the U.S. Census 
Bureau's 1997 Economics Census for the three industries most likely to 
be impacted by this proposed rule. The total number of firms listed in 
table 3 of this document is different from earlier parts of the 
analysis because the earlier estimates were derived from 2002 USDA data 
but the most recent Economic Census data available is for 1997.

                           Table 3.-Dairy Manufacturing Statistics by Employment Size
----------------------------------------------------------------------------------------------------------------
                                                           Number of Firms with Less   Percent of Industry that
                                 Total Number Of Firms        than 500 Employees             is ``Small''
----------------------------------------------------------------------------------------------------------------
Cheese Manufacturing                              524                         518                          98.9
----------------------------------------------------------------------------------------------------------------
Fluid Milk Manufacturing                          612                         605                          98.9
----------------------------------------------------------------------------------------------------------------
Dry, Condensed, and                               213                         208                          97.7
 Evaporated Dairy
 Manufacturing
----------------------------------------------------------------------------------------------------------------
Source: U.S. Census Bureau, 1997 Economic Census June 24, 1999 Manufacturing--Industry Series.

    Based on the SBA definition of small business for the dairy 
manufacturing industries, almost all dairy and cheese manufacturers 
qualify. However, Blayney and Manchester found that large dairy 
manufacturing companies and cooperatives, those percent with food and 
nonfood sales in 1998 of $800 million or more, accounted for almost 70 
percent of the industry (Ref. 65). Of this 70 percent, large 
proprietary companies accounted for 42 percent and large cooperatives 
for 27 percent. The remainder of the industry was divided between 
smaller companies, including cooperatives (Ref. 65).
    The dairy industry in the United States exhibits substantial 
economies of scale and, historically, small dairy farms have found ways 
of combining their resources to be able to compete in the industry. The 
1960s saw a wave of mergers and consolidations, leading to almost a 
complete conversion to ``bulk handling and processing'' of milk at 
plants in the 1970s. This trend has continued with ever-decreasing 
numbers of processors handling ever-increasing volumes of milk (Ref. 
47).
    FDA believes that if cheese manufacturers demand UF milk, dairy 
cooperatives will adjust in order to keep themselves and their 
individual members viable in the market. In 1997, the last year the 
USDA did a comprehensive survey of dairy cooperatives, dairy 
cooperatives handled 83 percent of all milk delivered to plants and 
dealers in the United States, and 98 percent of the milk received by 
cooperatives came directly from member producers (Ref. 53). These 
cooperatives are diverse in size, but the average handles 564 million 
lb annually, well above the 2.2 million lb requirement of production 
from 100 cows. According to the National Milk Producers Federation 
(NMPF) Web site, the average U.S. dairy cow produces about 7 gallons of 
milk per day (Ref. 66). To calculate the minimum weight to make UF 
technology financially feasible, we multiplied 100 cows by 7 gallons 
per day by 365 days per year to get 255,500 gallons per year. We then 
multiplied the product by 8.62 lb per gallon (NMPF Web site) to get 
2,202,410 lb per year. FDA seeks comment on the financial burden 
investing in UF technology imposes on dairy processors and cheese 
manufacturers, particularly small entities.
    In addition, small milk operations combined in cooperatives may be 
able to gain additional benefits from UF technology if they are able to 
market their products in a larger geographic region as a result of the 
lower shipping costs. This issue may be important if dairies develop in 
remote locations around the country as Mermelstein (Ref. 48) has 
suggested, or if there is a geographical shift in the production of 
either cheese or its components. Milk production in the West, as a 
percentage of total U.S. production, has increased, and there is some 
concern that Midwestern cheese producers will become ``milk-starved'' 
(Ref. 49). National Agricultural Statistics Services data over the past 
9 years has shown a significant increase in milk production in the 
West, up to 38 percent of the U.S. total in 2001 and 2002. However, 
these data also show a significant increase in cheese production in the 
Western States over this same time period, up to 37 percent in 2001 and 
38 percent in 2002 (Ref. 67). The significantly lower hauling costs for 
filtered milk may enable small milk processors and cheese producers to 
ship ingredients over longer distances to meet manufacturing needs.

V. Unfunded Mandates

    Title II of the Unfunded Mandates Reform Act of 1995 (Public Law 
104-4) requires cost-benefit and other analyses before any rule making 
if the rule would include a ``Federal mandate that may result in the 
expenditure by State, local, and tribal governments, in the aggregate, 
or by the private sector, of $100,000,000 or more (annually adjusted 
for inflation) in any 1 year.'' The current inflation-adjusted 
statutory threshold is $113 million. FDA has determined that this 
proposed rule does not constitute a significant rule under the Unfunded 
Mandates Reform Act.

VI. Small Business Regulatory Enforcement Fairness Act of 1996 (SBREFA) 
Major Rule

    The SBREFA (Public Law 104-121) defines a major rule for the 
purpose of congressional review as having caused or being likely to 
cause one or more of the following: an annual effect on the economy of 
$100 million; a major increase in cost or prices; significant adverse 
effects on competition, employment, productivity, or innovation; or 
significant adverse effects on the ability of United States-based 
enterprises to compete with foreign-based enterprises in domestic or 
export markets. In accordance with the SBREFA, the Office of Management 
and Budget (OMB) has determined that this proposed rule is a major rule 
for the purpose of congressional review.

[[Page 60767]]

VII. Federalism

    FDA has analyzed this proposed rule in accordance with the 
principles set forth in Executive Order 13132. FDA has determined that 
the rule would have a preemptive effect on state law. Section 4 (a) of 
the Executive Order requires agencies to ``construe * * * a Federal 
Statute to preempt State law only where the statute contains an express 
preemption provision, or there is some other clear evidence that the 
Congress intended preemption of State law, or where the exercise of 
State authority conflicts with the exercise of Federal authority under 
the Federal statute.'' Section 403A of the act (21 U.S.C. 343-1) is an 
express preemption provision. Section 403A(a)(1) provides that:
     * * * no State or political subdivision of a State may directly 
or indirectly establish under any authority or continue in effect as 
to any food in interstate commerce-(1) any requirement for a food 
which is the subject of a standard of identity established under 
section 401 that is not identical to such standard of identity or 
that is not identical to the requirement of section 403(g). * * *
This proposed rule makes changes to the general provisions related to 
the standards of identity for cheeses and related cheese products. 
Although this rule would have a preemptive effect in that it would 
preclude States from promulgating requirements for standardized cheese 
and cheese products that are not identical to the standards as amended 
by this proposal, this preemptive effect is consistent with what 
Congress set forth in section 403A of the act.
    Section 4(c) of the Executive Order further requires that ``any 
regulatory preemption of State law shall be restricted to the minimum 
level necessary'' to achieve the regulatory objective. Under section 
401 of the act (21 U.S.C. 341), ``[w]henever in the judgment of the 
Secretary such action will promote honesty and fair dealing in the 
interest of consumers, he shall promulgate regulations fixing and 
establishing for any food * * * a reasonable definition and standard of 
identity. * * *'' Further, section 4(e) provides that ``when an agency 
proposes to act through adjudication or rulemaking to preempt State 
law, the agency shall provide all affected State and local officials 
notice and an opportunity for appropriate participation in the 
proceedings.'' FDA is providing an opportunity for State and local 
officials to comment on this rulemaking. For the reasons set forth 
above, the agency believes that it has complied with all of the 
applicable requirements under the Executive order.
    In conclusion, FDA has determined that the preemptive effect of the 
proposed rule would be consistent with Executive Order 13132.

VIII. Environmental Impact

    We have determined under 21 CFR 25.32(p) that this action is of the 
type that does not individually or cumulatively have a significant 
effect on the human environment. Therefore, neither an environmental 
assessment nor an environmental impact statement is required.

IX. Paperwork Reduction Act of 1995

    FDA tentatively concludes that this proposed rule contains no 
collection of information. Therefore, clearance by OMB under Paperwork 
Reduction Act of 1995 is not required.

X. Comments

    Interested persons may submit to the Division of Dockets Management 
(see ADDRESSES) written or electronic comments regarding this document. 
Submit a single copy of electronic comments or two paper copies of any 
mailed comments, except that individuals may submit one paper copy. 
Comments are to be identified with the docket number found in brackets 
in the heading of this document. Received comments may be seen in the 
Division of Dockets Management between 9 a.m. and 4 p.m., Monday 
through Friday.

XI. References

    The following references have been placed on public display in the 
Division of Dockets Management (see ADDRESSES) and may be seen by 
interested persons between 9 a.m. and 4 p.m., Monday through Friday. 
(FDA has verified the Web site addresses, but FDA is not responsible 
for any subsequent changes to the Web sites after this document 
publishes in the Federal Register.)
    1. Cheryan, M., Ultrafiltration and Microfiltration Handbook, 2d 
ed., CRC Press LLC, Boca Raton, FL, chapters 1 and 3 and pp. 349-
369, 1998.
    2. Rosenberg, M., ``Current and Future Applications for Membrane 
Processes in the Dairy Industry,'' Trends in Food Science and 
Technology, 6:12-19, 1995.
    3. Smith, K.E., Background on Milk Protein Products, Dairy 
Proteins, prepared by the Wisconsin Center for Dairy Research and 
the Wisconsin Milk Marketing Board, 2001.
    4. GAO report, ``Dairy Products: Imports, Domestic Production, 
and Regulation of Ultra-filtered Milk,'' March, GAO-01-326, 2001.
    5. Letter to FDA from Mr. Ted Jacoby, Jr., T.C. Jacoby & 
Company, Inc., May 1, 1996
    6. Letter to Mr. Ted Jacoby, Jr., T.C. Jacoby & Company, Inc., 
from FDA, October 21, 1996.
    7. Letter to Mr. F. Tracy Schonrock, USDA from FDA, October 21, 
1999.
    8. Letter to Mr. Clay Hough, International Dairy Foods 
Association from FDA, April 6, 2005.
    9. Babano, D.M., `` Making Cheese From Higher Solids Milk: 
Advantages and Pitfalls,'' Proceedings of the 37th annual Marschall 
cheese seminar, Visalia, CA, 2000.
    10. Mistry, V.V. and J-L. Maubois, ``Application of Membrane 
Separation Technology to Cheese Production,'' Cheese: Chemistry, 
Physics and Microbiology : General Aspects, 3rd edition (Volume 1), 
Eds: Fox, P.F., P.L.H McSweeney, T.M. Cogan, and T.P. Guinee, 
Elsevier Academic Press, London. pp. 261-285.
    11. Acharya, M.R. and V.V. Mistry, ``Comparison of Effect of 
Vacuum-Condensed and Ultrafiltered Milk on Cheddar Cheese,'' Journal 
of Dairy Science, 87:4004-4012, 2004.
    12. Hydamaka, A.W., R.A. Wilbey, M.J. Lewis, et al., 
``Manufacture of Heat and Acid Coagulated Cheese From Ultrafiltered 
Milk Retentates,'' Food Research International, 34:197-205, 2001.
    13. Rodriguez, J., T. Requena, J. Fontecha, et al., ``Effect of 
Different Membrane Separation Technologies (Ultrafiltration and 
Microfiltration) on the Texture and Microstructure of Semihard Low-
Fat Cheeses,'' Journal of Agricultural and Food Chemistry, 47:558-
565, 1999.
    14. Kosikowski, F.V., A.R., Masters, and V.V. Mistry, ``Cottage 
Cheese From Retentate-Supplemented Skim Milk,'' Journal of Dairy 
Science, 68:541-547, 1985.
    15. Kealey, K.S. and F.V. Kosikowski, ``Cottage Cheese From 
Ultrafiltered Skim Milk Retentates in Industrial Cheese Making,'' 
Journal of Dairy Science, 69:1479-1483, 1986.
    16. Johnson, M., ``Standardization of Milk Using Cold 
Ultrafiltration Retentates for the Manufacture of Swiss Cheese,'' 
Unpublished data, 2004.
    17. Guinee, T.P., P.D. Pudja, W.J. Reville, et al., 
``Composition, Microstructure and Maturation of Semi-Hard Cheeses 
From High Protein Ultrafiltered Milk Retentates With Different 
Levels of Denatured Whey Protein,'' International Dairy Journal, 
5:543-568, 1995.
    18. Oommen, B.S., V.V. Mistry, and M.G. Nair, ``Effect of 
Homogenization of Cream on Composition, Yield, and Functionality of 
Cheddar Cheese Made From Milk Supplemented With Ultrafiltered 
Milk,'' Lait, 80:77-91, 2000.
    19. Poduval, V.S. and V.V. Mistry, ``Manufacture of Reduced Fat 
Mozzarella Cheese Using Ultrafiltered Sweet Buttermilk and 
Homogenized Cream,'' Journal of Dairy Science, 82:1-9, 1999.
    20. Raval, D.M. and V.V. Mistry, ``Application of Ultrafiltered 
Sweet Buttermilk in the Manufacture of Reduced Fat Process Cheese,'' 
Journal of Dairy Science, 82:2334-2343, 1999.
    21. Raphaelides, S., K.D. Antoniou, and D. Petridis, ``Texture 
Evaluation of Ultrafiltered Teleme Cheese,'' Journal of Food 
Science, 60:1211-1215, 1995.
    22. Lelievre, J. and R.C. Lawrence, ``Manufacture of Cheese From 
Milk Concentrated by Ultrafiltration,'' Journal of Dairy Research, 
55:465-478, 1988.
    23. Spangler, P.L., L.A. Jensen, C.H. Amundson, et.al., ``Gouda 
Cheese Made

[[Page 60768]]

From Ultrafiltered Milk: Effects of Concentration Factor, Rennet 
Concentration, and Coagulation Temperature,'' Journal of Dairy 
Science, 73:1420-1428, 1990.
    24. Eckner, K.F. and E.A. Zottola, ``The Behavior of Selected 
Microorganisms During the Manufacture of High Moisture Jack Cheeses 
From Ultrafiltered Milk,'' Journal of Dairy Science, 74:2820-2830, 
1991.
    25. Codex General standard for Cheese, Codex Stan A-6-1978, Rev. 
1-1999, amended 2003.
    26. Codex Group Standard for Cheeses in Brine, Codex Stan 208-
1999, amend. 1-2001.
    27. Codex International Individual Standard for Cottage Cheese, 
Including Creamed Cottage Cheese, Codex Stan C-16-1968.
    28. Codex International Individual Standard for Cream Cheese 
(Rahmfrischkase), Codex Stan C-31-1973.
    29. Codex International Standard for Extra Hard Grating Cheese, 
Codex Stan C-35-1978.
    30. Codex Group Standard for Unripened Cheese Including Fresh 
Cheese, Codex Stan 221-2001.
    31. Codex General Standard for Named Variety Process(ed) Cheese 
and Spreadable Process(ed) Cheese, Codex Stan A-8(a)-1978.
    32. Codex General Standard for Process(ed) Cheese and Spreadable 
Process(ed) Cheese, Codex Stan A-8(b)-1978.
    33. Codex General Standard for Process(ed) Cheese Preparations 
(Process(ed) Cheese Food and Process(ed) Cheese Spread), Codex Stan 
A-8(c)-1978.
    34. Codex Standard for Whey Cheese, Codex Stan A-7-1978, Rev. 1-
1999.
    35. Miller, G.D., J.K. Jarvis, and L.D. McBean, Handbook of 
Dairy Foods and Nutrition, second edition. CRC Press LLC, Boca 
Raton, FL, pp. 6-16, 2000.
    36. Rattray, W. and P. Jelen, ``Protein Standardization of Milk 
and Dairy Products,'' Trends in Food Science and Technology, 7:227-
234, 1996.
    37. Dairy Management, Inc., ``Opportunities for Membrane 
Filtration of Milk. Innovations in Dairy,'' Dairy Industry 
Technology Review, 2000.
    38. Mehaia, M.A. and S.M. El-Khadragy, ``Physicochemical 
Characteristics and Rennet Coagulation Time of Ultrafiltered Goat 
Milk,'' Food Chemistry, 62:257-263, 1998.
    39. Green, M.L., J. Scott, M. Anderson, et al., ``Chemical 
Characterization of Milk Concentrated by Ultrafiltration,'' Journal 
of Dairy Science, 51:267-278, 1984.
    40. Wisconsin Center for Dairy Research and the Wisconsin Milk 
Marketing Board, Dairy Proteins. Accessed September 26, 2003 (http://www.cdr.wisc.edu/CDRWEBPa.nsf), 2001.
    41. American Dairy Products Institute, ``Dry milk products--
Utilization & Production Trends,'' 2002.
    42. Moran, J.W., G.W. Trecker, and S.P. Monckton, Continuous 
Manufacture of Process Cheese, US Patent 6183805, 2001.
    43. Voutsinas, L.P., C.M. Katsiari, C.P. Pappas, et al., 
``Production of Brined Soft Cheese From Frozen Ultrafiltrered 
Sheep's Milk. Part 1. Physicochemical, Microbiological and Physical 
Stability Properties of Concentrates,'' Food Chemistry, 52:227-233, 
1995.
    44. Johnson and Wendorff, ``Making Cheese With UF Raw Milk,'' 
Dairy Pipeline, 10(2):5-7, 1998.
    45. USDA, National Agricultural Statistics Services, Dairy 
Products 2002 Summary, 2003.
    46. USDA, ``Milk Production,'' National Agricultural Statistics 
Services, Agricultural Statistics Board, released February 14, 2003.
    47. Liebrand, C., ``Structural Change in the Dairy Cooperative 
Sector, 1992-2000,'' USDA, Rural Business-Cooperative Service, RBS 
Research Report 187, 2001.
    48. Mermelstein, N.H., ``Concentrating Milk,'' Food Technology, 
56(3):72-74, 78, 2002.
    49. Fassbender, R. Using Cold Ultrafiltered Milk in 
Cheesemaking, Proceedings of the 38th Annual Marschall Cheese 
Seminar, Visilia, CA, accessed July 16, 2003 (http://marschall.com/marschall/proceed/index.htm), 2001.
    50. Dairy Management, Inc., ``Nonfat Dry Milk Ingredients,'' 
DMI-A8040-0500-01, 2000.
    51. Slack, A.W., C.H. Amundson, and C.G. Hill, Jr., ``On-farm 
Ultrafiltration of Milk: Part 2. Economic Analysis,'' Process 
Biochemistry, 17:23-33, 1982.
    52. Sheehan, J., Memo to file, September 1, 2005.
    53. USDA, Cooperatives in the dairy industry. Rural Business--
Cooperative Service, Cooperative Information Report 1, Section 16, 
released July 2002.
    54. Barske, L.J., Milk hauling charges in the upper Midwest 
marketing area, May 2001, Federal Milk Market Administrator's 
Office, Staff Paper 02-01, 2002.
    55. USDA, Economic Research Service, ``Food Marketing and Price 
Spreads: Farm to Retail Price Spreads for Individual Food Items,'' 
briefing room, accessed October 17, 2003 (http://ers.usda.gov/briefing/foodpricespreads/spreads/table1.htm).
    56. Bailey, K.W., ``U.S. Market Structure: The Dairy Industry in 
the 21st Century,'' Paper presented at the 66th Annual Meeting of 
the International Association of Milk Control Agencies, Calgary, 
Alberta, Canada, July 14-17, 2002.
    57. Bailey, K.W., ``Milk Protein Concentrate Imports: 
Implications for the North American Dairy Industry,'' Advances in 
Dairy Technology, 15:137, 2003.
    58. The Economics of Drying, On-line article on the Chemical 
Engineer's Resource Center, accessed September 30, 2003 (http://geocities.com/chemforum).
    59. Sheehan, J., Memo to file, September 1, 2005.
    60. Jesse, E., ``U.S. Imports of Concentrated Milk Proteins: 
What We Know and Don't Know?'' Marketing and Policy Briefing Paper 
80, 2003.
    61. NMPF News Release Latest import watch newsletter sees MPC 
imports increase, Surge in Milk Protein Concentrate, Casein Again 
Points Out Need for Tariff Bills, accessed September 9, 2003 (http://www.aae.wisc.edu/future/), 2003.
    62. Kampert, P., ``Yogurt Proposal Has Dairy Farmers Sour, They 
Say Rules on Imported Protein A Threat,'' Chicago Tribune, September 
14, 2003.
    63. Dairy Management, Inc., ``Opportunities for Membrane 
Filtration of Milk,'' Innovations in Dairy, January 2000.
    64. Berry, D., Microfiltering Milk in the February 2001 ed. of 
Lab Talk, accessed May 20, 2002. (http://www.dairyfoods.com/articles/2001/0201/0201lab.htm), 2001.
    65. Blayney, D.P. and A.C. Manchestern, Large companies active 
in changing dairy industry. Food Review 23(2):8-13, 2000.
    66. NMPF FAQ. 2003. Accessed October 6, 2003 (http://www.nmpf.org/faq), 2003.
    67. USDA, National Agricultural Statistics Services, Dairy 
Product 1994 through 2003 and Milk Production 1994 through 2003.

List of Subjects in 21 CFR Part 133

    Cheese, Food grades and standards, Food labeling.
    Therefore, under the Federal Food, Drug, and Cosmetic Act and under 
authority delegated to the Commissioner of Food and Drugs and re-
delegated to the Director of the Center for Food Safety and Applied 
Nutrition, it is proposed that 21 CFR part 133 be amended as follows:

PART 133--CHEESES AND RELATED CHEESE PRODUCTS

    1. The authority citation for 21 CFR part 133 continues to read as 
follows:

    Authority: 21 U.S.C. 321, 341, 343, 348, 371, 379e.
    2. Section 133.3 is amended by revising paragraphs (a) and (b) and 
by adding new paragraphs (f) and (g) to read as follow:


Sec.  133.3  Definitions.

    (a) Milk means the lacteal secretion, practically free from 
colostrum, obtained by the complete milking of one or more healthy 
cows, which may be clarified and may be adjusted by separating part of 
the fat therefrom; concentrated milk, reconstituted milk, and dry whole 
milk. Water, in a sufficient quantity to reconstitute concentrated and 
dry forms, may be added. For the purposes of this part, wherever the 
term ``milk'' appears in the individual standards for cheeses and 
related cheese products, ultrafiltered milk as described in paragraph 
(f) of this section, may be used.
    (b) Nonfat milk means skim milk, concentrated skim milk, 
reconstituted skim milk, and nonfat dry milk. Water, in a sufficient 
quantity to reconstitute concentrated and dry forms, may be added. For 
the purposes of this part, wherever the term ``nonfat milk'' appears in 
the individual standards for cheeses and related cheese products, 
ultrafiltered nonfat milk as described in paragraph (g) of this 
section, may be used.
* * * * *

[[Page 60769]]

    (f) Ultrafiltered milk means raw or pasteurized milk that is passed 
over one or more semipermeable membranes to partially remove water, 
lactose, minerals, and water-soluble vitamins without altering the 
casein:whey protein ratio of the milk and resulting in a liquid 
product.
    (g) Ultrafiltered nonfat milk means raw or pasteurized nonfat milk 
that is passed over one or more semipermeable membranes to partially 
remove water, lactose, minerals, and water-soluble vitamins without 
altering the casein:whey protein ratio of the nonfat milk and resulting 
in a liquid product.

    Dated: October 7, 2005.
Leslye M. Fraser,
Director, Office of Regulations and Policy, Center for Food Safety and 
Applied Nutrition.
[FR Doc. 05-20874 Filed 10-18-05; 8:45 am]
BILLING CODE 4160-01-S