FTC Staff Comment Before the Department of Agriculture, Food and Inspection Service Concerning Cooked Sausage Standard

Po I , cy V I , "-- -- ,
~oom 3812-South Bu;1din~
'£2 JU~ ;Itm
ff~/113
i·r;~lMJSSmN 	AUTHORllf,U
Comments of the Bureaus of Economics,
Consumer Protection and Competition
of the Federal Trade Commission'
Submitted to the
Food Safety and Inspection Service,
Department of Agriculture
In Response to a Request for
Comments on its Proposal to
Amend the Cooked Sausage Standard
Docket No. 85-009£
[51 FR 42239, 52 FR 2416]
Attention: Ms. Linda Carey
Room 3168
South Agriculture Building
Washington, D.C.
~ 2 JUN t987
• These comments represent the views of the above named Bureaus and do not
necessarily 	represent 	the 	views 	of 	the 	Commission 	or 	of 	any 	individual
Commissioner. 	The Commission has, 	however, 	voted 	to authorize 	the staff
to submit these comments.

I. Introduction
In its Notice of Proposed Rulemaking, the Food Safety and Inspection
Service (FSIS) has requested comment on its proposal to amend the standard
of 	identity 	for 	frankfurters 	and 	cooked 	sausage 	products.} 	The 	proposed
amendment would give producers flexibility to substitute added water for fat.
This 	flexibility 	would 	enable 	manufacturers 	to 	produce 	low 	fat 	cooked
sausages 	that 	are 	moist 	enough 	to 	be 	palatable. 	FSIS 	has 	also 	solicited
comment on 	whether there is a "market failure that creates a need 	for 	the
frankfurter 	standard 	and 	on 	the 	benefits 	and 	costs 	to 	industry 	and
consumers associated with eliminating the standard."
The 	FSIS 	proposal 	represents 	an 	improvement 	over 	the 	current, 	more
rigid, standard. 	However, we also believe that if there is a market failure,
then a mandatory disclosure of fat content 2 would be preferable to both the
current and 	proposed standards. 	In the case of cooked sausages, disclosure
is 	likely 	to 	provide 	greater 	net 	benefits 	than 	an 	identity 	(or 	"recipe")
standard 	because 	provision 	of 	the 	information 	appears 	generally 	to 	be
inexpensive and because consumer tastes vary.
We begin our main comment (in section II) with a statement of why the
FTC staff has an interest in this matter. 	Then, in sections III and IV, 	we
discuss 	the 	types of potential market "failures" 	that recipe standards were
} The eXIstmg standard limits fat and added water content to 30% and
10% respectively. 	The proposed standard would still limit fat content to 30%
but would permit 	fat 	plus added 	water content to total 	to 40%. 	Details of
the current and proposed standards are given beJ.ow in sections VI and VII.
2 Under 	some 	circumstances, 	mandatory 	disclosure 	of 	protein 	content
could be beneficial. 	See footnotes 29 and 34 below.

intended 	to 	redress, and 	their 	root 	causes. 	Assuming 	that 	an 	unregulated
market 	would 	fail, 	Section 	V 	assesses 	the 	relative 	costs 	and 	benefits 	of
identity standards and 	mandatory disclosure requirements and considers the
conditions under which each approach would be preferable. 	Following this
general 	inquiry, 	we 	discuss 	the 	current 	and 	proposed 	cooked 	sausage
standards 	10 	sections 	VI 	and 	VII, 	respectively.
summarized in section VIII.
Our 	conclusions 	are
The issue of whether recipe standards are effective goes far beyond the
current 	proposal 	reiating 	to 	cooked 	sausages. 	More 	than 	30 	years 	ago, 	it
was estimated that over 50% of consumer expenditures on food went toward
products 	covered 	by 	FDA 	or 	USDA 	identity 	standards. 3 	Since 	that 	time
many 	new 	standards 	have 	been 	added 	and 	few 	withdrawn, 	so 	identity
standards 	apply 	to 	hundreds 	of 	billions 	of dollars 	of 	food 	products 	each
Identity 	standards, 	though 	they 	may 	provide "benefits, 	may 	also
reduce innovation and competition in 	the food 	industry. 	If such standards
are no longer necessary or if less costly alternative remedies can 	be found,
then 	their 	elimination 	or 	replacement 	could 	provide 	important 	benefits 	to
consumers.
3 In 	their 	article 	"Like 	Mother 	Used 	to 	Make: 	An 	Analysis 	of 	FDA
Food Standards of Identity", Richard A. Merrill and Earl M. Collier cite an
estimate published in 1954. 	See Columbia Law Review, 1974, p. 561.
.c Consumers 	spent 	over 	$474 	billion 	on 	food 	(including 	restaurant
expenses) in 1985. 	Economic Report of the President, 1986, p. 268.
2

II. FTC Staff Interest
The FTC's interest in food standards and labeling stems from its general
statutory 	obligations 	to 	promote 	competition 	and 	to 	prevent 	false 	and
deceptive 	advertising. 	Under 	sections 	5 	and 	12 	of 	the 	Federal 	Trade
Commission 	Act 	(15 	U.S.c. 	45 	et 	seq.), 	which 	prohibit 	false, 	deceptive 	or
unfair practices in or affecting commerce, the FTC has jurisdiction over the
I
advertising of 	food, 	and 	has concurrent jurisdiction 	with 	FDA 	and 	USDA
over the labeling of food. 	The FTC also has statutory authority to enforce
a 	number 	of laws 	that 	mandate disclosure 	(Federal 	Cigarette Labeling and
Advertising 	Act, 	the 	Truth 	In 	Lending 	Act, 	the 	Energy 	Policy 	and
Conservation Act [appliance labeling]), has itself promulgated disclosure rules
(Octane 	Rating, 	R-Value 	Rule, 	Care 	Labeling) 	and 	enforces 	several 	laws
rela ting to standard-setting (the Wool 	Products Labeling Act, the Magnuson-
Moss Warranty - FTC Improvement Act). 	In implementing its responsibilities,
the 	FTC 	staff 	has 	developed 	considerable 	economic 	expertise 	In
understanding 	the 	roles of advertising and 	labeling in 	providing consumers
with reliable product information and in the roles that might be played by
disclosure 	or standards 	in situations 	where 	the 	market 	will 	fail 	to 	provide
adequate information without regulation. s
S FTC staff's 	research 	includes W. Jacobs et aI., "Improving Consumer
Access 	to 	Legal 	Services: The 	Case 	for 	Removing 	Restrictions on 	Truthful
Advertising 	(I984); 	R. 	Bond 	et 	aI., 	"Effects of 	Restrictions 	on 	Advertising
and 	Commercial 	Practice in 	the Professions: the Case of Optometry" (1980);
Bureau 	of 	Consumer 	Protection, 	FTC, 	"Drug 	Product 	Selection" 	(1979); 	A.
Masson 	and 	R. Steiner, 	"Generic Substitution 	and 	Prescription 	Drug 	Prices:
Economic 	Effects 	of State 	Drug Substitution 	Laws" (1985); 	M. 	Frankena 	et
aI., "Alcohol 	Advertising, Consumption, and 	Abuse" (1985); M. 	Lynch et aI.,
"Experimental 	Studies 	of 	Markets 	with 	Buyers 	Ignorant 	of 	Quality 	Before
Purchase: When Do 'Lemons' Drive out High Quality Products?" (1986).
3

III. Potential Market Failure
A 	recipe 	standard 	prescribes 	that 	certain 	ingredients 	in 	minimum 	or
maximum 	proportions 	be 	present 	in 	a 	named 	product, 	includes 	a 	list 	of
optional ingredients, and may also prescribe the way in which the ingredients
may be manufactured and combined. 	A product may be sold under the name
designated 	by 	the 	identity 	standard 	if, 	and 	only 	if, 	it 	conforms 	to 	the
standard. 	Moreover, 	the 	sale 	of 	any 	food 	which 	"purports 	to 	be" 	or 	is
"represented 	as" 	the 	standardized 	food 	is 	prohibited 	unless 	it 	conforms 	to
the standard. 6
Recipe standards (such as cooked sausage standard) appear to have been
motivated 	by 	three 	concerns. 	The 	main 	concern 	was 	with 	deceptive
"economic 	adulteration.,,7 	It 	was 	feared 	that 	unregulated- producers 	would
substitute new and cheaper ingredients in 	traditional foods, and "pass them
6 See section 	403(g) 	of 	the 	Federal 	Food, 	Drug 	and 	Cosmetic 	Act 	of
1938. 	For 	example, 	for 	many 	years 	prior 	to 	the 	passage 	of 	the 	this 	Act,
Quaker 	Oats 	had 	marketed a 	product 	named "Quaker Farina 	Wheat 	Cereal
Enriched With Vitamin D." 	In 1938, the FDA adopted two standards: "plain
farina" 	and 	"enriched farina." 	Quaker's product did 	not conform 	to either
standard and its production and sale were prohibited since it purported to be
enriched 	farina, 	despite 	the fact 	its 	product was wholesome and 	truthfully
labeled. 	Quaker 	appealed 	all 	the 	way 	to 	the 	Supreme 	Court 	and 	lost
[Federal Security Administrator v. Quaker Oats Co., 318 U.S. 218 (1943)].
7 For 	a 	clear, 	and 	very 	interesting, 	account 	of 	some 	"adulteration"
problems 	encountered 	in 	the 	early 	part 	of 	this 	century, 	see 	Carl 	Alsberg,
"Economic 	Aspects 	of 	Adulteration 	and 	Imitation," 	Ouarterlv 	Journal
Economics, November, 1931, 1-33. 	Alsberg (a chemist) had been Chief of the
Burea u of Chemistry at USDA, the group responsible for enforcing the 1906
Food and Drug Act. 	He later was the first Director of the Food Institute at
Stanford University.
4

off" as traditional staples to unsuspecting consumers. 8 	This concern will be
elaborated more fully below.
A second and closely related concern was that producers might add new
ingredients to traditional products and that these products might ultimately
prove 	to 	be 	unsafe, even if 	the producers were 	not attempting 	to pass off
their 	products 	as 	something 	they 	were 	not.9 	Whatever 	was 	true 	in 	the
past, safety considerations no longer provide a rationale for recipe standards.
Prior 	to 	1958, 	the 	burden 	was 	on 	FSIS 	or 	FDA 	to 	show 	tha t 	a 	suspect
untested 	ingredient 	was 	"poisonous 	or 	deleterious." 	In 	1958, 	however,
Congress amended the Pure Food and Drug Act to require manufacturers to
obtain 	prior 	approval 	for 	all 	food 	additives, 	whether 	for 	standardized 	or
non-standardized 	foods. lO 	Although 	the 	existence 	of 	a 	recipe 	standard 	is
now 	irrelevant 	to 	preventing 	the 	use 	of 	inherently 	unsafe 	ingredients, 	a
,
standard 	may still 	be 	relevant in 	limiting the use of ingredients which are
safe 	when 	consumed 	in 	mOderation, 	but 	which 	pose a 	health 	risk 	to 	some
consumers 	when 	consumed 	in 	larger 	amounts. 	Fat, 	for 	example, 	is 	an
important source of metabolic energy, and certain fatty acids are necessary
for good health. 	Excessive fat consumption, however, is widely believed to
increase the risk of heart disease.
The third concern was that producers would add insignificant amounts of
8 See H. Thomas 	Austern, "The F-O-R-M-U-L-A-T-I-O-N of Mandatory
Food Standards," Food Drug Cosmetic Law Journal, September, 1971, pp. 380-
382 and passim. (reprinted from the Dec. 1947 issue).
9 Merrill & 	Collier point out that although Congress conceived of food
standards 	primarily 	as 	a 	means 	of 	combatting 	economic 	adulteration, 	in
practice "it is difficult to distinguish sharply between pocketbook and health
interests of consumers" 	op. cit., fn. 14, p. 564.
10 Collier & Merrill, op. cit., p. 600.
5

nutrients or other seemingly desirable ingredients and 	then exaggerate their
importance 	and 	deceive 	consumers 	into 	paying 	premium 	prices 	greatly
exceeding 	the 	value 	of 	the 	extra 	ingredients. 11 	This 	argument 	seems 	to
have little current relevance for justifying recipe standards, since it assumes
that manufacturers could label and advertise additives in a misleading fashion
despite 	FSIS, 	FDA 	and 	FTC 	regulations 	that 	prohibit 	false 	and 	deceptive
claims. 	In 	any 	event, 	the 	argument 	has 	no 	relevance 	to 	fat 	and 	added
water, which are not ingredients manufacturers would want to exaggerate.
Under current conditions, then, recipe standards must be assessed 	as a
means of combating the potential failure of the market to prevent two types
of problems: (I) the deceptive sale, under traditional names, of products that
are in fact cheap imitations, and (2) the sale of products that, unknown 	to
the 	buyer, 	contain 	high 	concentrations 	of 	ingredients 	that 	may 	pose
increased 	health 	risks 	for 	some 	consumers. 	Absent 	a 	cooked 	sausage
standard, or alternative remedy, consumers might unknowingly buy sausages
with 	more 	than 	30% fat. 	If consumers could not detect the additional fat,
producers 	would 	have 	an 	incentive 	to substitute 	fat 	for 	meat, since 	fat 	is
cheaper than meat. 	Consumers who regard more fat as a "bad" might suffer
injury 	not 	only 	to 	their 	health 	but 	also 	to 	their 	wealth, 	if 	less 	fatty
sausages could be supplied at the sanle or a lower price.
IV. Causes of Potential Market Failure
Both 	of 	these 	concerns 	r~quire 	that 	buyers 	be 	ignorant 	of 	the
presence 	or 	quantity 	of 	undesirable 	added 	or 	substituted 	ingredients.
11 Collier and Merrill, ibid., pp. 597-599.
6

Although 	buyer 	ignorance 	of 	some 	product characteristics 	before, 	or even
after, 	purchase 	can 	lead 	to 	market 	failure, 	in 	many 	cases 	explicit 	and
implicit private contracts have evolved that overcome such buyer ignorance.
These include seller or brand name reputation and seller warranties. 	So long
as 	buyers can 	and 	sometimes do 	learn 	the overall quality 	of an 	item 	they
have purchased from a known seller, and sellers need good "word of mouth"
or repeat business to survive, reputation can overcome problems that would
otherwise 	arise. 12 	Also, 	new 	(and 	established) 	sellers 	can 	attract 	trial
customers by offering "money back if not satisfied" or other guarantees.
However, 	not all 	characteristics can 	be assessed 	through experience 	in
this way. 	For example, consumption will not reveal that an ingredient added
to 	help 	preserve 	a 	traditional 	food 	may 	also 	increase 	the 	risk 	that 	the
consumer 	will 	so~e day 	develop cancer, and 	it 	might 	not 	reveal 	that 	one
brand 	of 	hot dogs contains 	more 	fat 	than another. 	Though 	manufacturers
often 	disclose 	and 	tout 	the 	desirable 	credence 	qualities 	of 	their 	products,
one 	would 	not 	expect 	them 	to 	advertise 	negative 	characteristics 	directly.
12 Economists 	divide 	product 	characteristics 	into 	three 	categories.
"Search" characteristics are those that can be verified by sensory inspection;
"experience" characteristics can be 	verified 	in 	the course 	of 	normal 	use or
consumption; 	and 	"credence" 	characteristics 	are 	those 	that 	can 	only 	be
discovered 	by 	costly, 	special 	examination. 	For 	the 	first 	two, 	see 	Phillip
Nelson, 	"Information and Consumer Behavior", Journal of Political Economy,
March/April 	1970,78,311-329. 	For 	credence 	characteristics, 	see 	Neil
Borden, The Economics of Advertising, Irwin, 1942, pp. 33', 425 and 429 and
Michael Darby and Edi Karni, "Free Competition and the Optimal Amount of
Fraud", Journal of Law & Economics, XVI (1), April 1973,67-88.
For 	some 	recent 	economic 	theories 	of 	the 	role 	of 	reputation 	in
experience 	good 	markets, 	see 	Benjamin 	Klein 	and 	Keith 	B. 	Leffler,
"Non-Governmental Enforcement of Contracts: The Role of Market Forces in
Guaranteeing 	Quality," 	Journal 	of 	Political 	Economy, 	August 	1981, 	89,
615-641, and Carl Shapiro, "Consumer Information, Product Quality and Seller
Reputation," 	Bell 	Journnl 	of 	Economics, 	Spring 	1982, 	13, 	20-35. 	For 	an
economic theory of warranties, see Sanford J. Grossman, "The Informational
Role of Warranties and Private Disclosure about Product Quality," 	Journal of
Law and Economics, December 1981,24,461-484.
7

They 	may 	do so 	indirectly, 	however.. 	If 	products 	with 	less 	of a 	negative
characteristic can be manufactured, then some firms may produce them and
advertise 	their 	advantage 	("no 	preservatives," 	"less 	fat 	in 	our 	franks").
Nevertheless, 	if 	the 	negative characteristic cannot 	be 	varied (cholesterol 	in
eggs), 	and 	if 	there 	are 	no close substitutes, 	then 	the 	market will 	probably
fail 	to 	provide 	sufficient 	information. 13 	Incentives 	to 	build 	a 	good
reputation might remedy a market failure caused by credence characteristics,
but only if a significant number of consumers take the time and trouble to
become informed concerning them.
In 	summary, 	if 	the 	market 	suffers 	from 	the 	two 	potential 	problems
discussed 	above, 	then 	the 	root 	cause 	of 	both 	is 	lack 	of 	consumer
information. 	Market failure is most likely to occur with respect to credence
characteristics, 	particularly 	those 	that 	are 	intrinsic 	to 	the 	product 	and 	so
difficult 	or 	impossible 	to 	vary. 	Under 	these 	conditions, 	some 	form 	of
government 	regulation 	may 	have 	benefits 	for 	consumers 	that 	exceed 	the
associated costs.
To 	some 	degree, 	the 	fat 	content 	of 	sausages 	may 	be 	a 	credence
characteristic. 	Consumers 	may 	not 	be 	able 	to 	distinguish, 	by 	taste 	or
,
appearance, 	between 	sausages 	with 	very 	different 	fat 	contents. 14 	The
13 See 	Robert 	Pitofsky, 	"Mandated 	Disclosure 	in 	the 	Advertising 	of
Consumer Products", and 	Richard Posner, "The Federal Trade Commission's
Mandated-Disclosure 	Program: 	A 	Critical 	Analysis," 	and 	their 	dialogue, 	in
Business 	Disclosure--Government's 	Need 	to 	Know, 	edited 	by 	Harvey
Goldschmid, 	New 	York: 	McGraw-Hili, 	1979, 	311-376. 	For 	a 	thorough
discussion 	of 	the 	peculiarities 	of 	the 	market 	for 	consumer 	product
information, 	see 	Howard 	Beales, 	Richard 	Craswell 	and 	Steven 	Salop, 	"The
Efficient 	Regulation 	of 	Consumer 	Information," 	Journal 	of 	Law 	and
Economics, XXIV(3), Dec. 1981.
14 Some USDA surveys showed that "high levels of fat contribute little
to 	taste 	value. 	Other 	factors 	such 	as 	seasoning 	appear 	to 	be 	of 	greater
importance," 	according 	to 	testimony 	of 	Dr. 	Leighty, 	thcn 	Chicf 	of 	the
8

average fat content of frankfurters rose 	from 	19% in 	1937 to 20% in 	1950
and to 33% in 1969.IS 	Protein content fell from 20% in 1937 to 14% in 1950
and to 11% in 1969. 16 	According to one writer,
"The 	trend 	to 	increased 	levels 	of 	fat 	was 	accelera ted 	by 	new
processing techniques imported from Germany and the discovery of
new 	fat 	emulsifiers 	by 	Oscar 	Meyer 	and 	Company. 	These
developments 	made 	it 	possible 	to 	incorporate 	vastly 	increased
amounts of fat (some products tested 	had as much as 51 % percent),
without disturbing the appearance or taste of the sausage. 	Because
sausage 	production 	is 	a 	competitive 	industry, 	sausage 	makers 	who
cut 	production 	costs 	by 	increasing 	the 	fat 	and 	lowering 	the 	meat
content forced other firms to follow suit."17
The above is essentially a description of a type of market failure known as
a 	"lemons" 	market 	by 	economists. 1S 	Because 	consumers 	cannot 	distinguish
between high and low fat hotdogs, they will not pay premium prices to those
Technical Services Branch, USDA, Public Hearings on Fat Content of Cooked
Sausage, 	June 	18, 	1969. 	Quoted 	in 	Harrison 	Wellford, 	Sowing 	the 	Wind,
Grossman, 	1972, p. 93. 	Surveys and 	test 	tastes commissioned 	by 	two 	meat
trade 	associations 	suggested, 	however, 	consumers 	had 	differing 	preferences
and that some of the observed 	variation 	in actual hotdog fat content (from
14% to 51%) was due to manufacturers trying to satisfy different tastes. 	For
more detail on this hearing, see Wellford, pp. 87-96.
15 Figures (rounded 	to the nearest whole number) taken 	from a 	Memo
from R.H. Alsmeyer, Head, Consumer & Marketing Standards Group, USDA, to
\V.J. Milnor, Acting Director, Technical Services Division, USDA, November
29, 1968. 	As quoted by Wellford, op. cit., p. 89.
16 ibid, p. 90.
17 ibid., p. 90.
18 See 	George 	A. 	Akerlof, 	"The 	Market 	for 	Lemons: 	Quality
Uncertainty 	and 	the Market 	Mechanism," 	Quarterlv 	Journal 	of 	Economics,
August 	1970, 84, 488-500, and 	M. Lynch, 	R. Miller, C. 	Plott and 	R. 	Porter,
Experimental 	Studies 	of 	Markets 	with 	Buyers 	Ignorant 	of 	Qualitv 	Before
Purchase: When Do "Lemons" Drive out High Quality Products?, Report to the
Federal Trade Commission, September, 1986.
9

producers who in fact provide lower fat hotdogs. 	The end result is a market
with only high fat content hotdogs selling at competitive prices.
It is clear, however, that the lemons model does not fully apply to the
cooked sausage market. 	As mentioned above, both in the past and to some
extent 	now, 	frankfurters 	with 	differing 	fat 	contents 	have 	been 	and 	arc
offered on 	the market. 	An 	informal survey of local supermarkets showed
..
that, 	of 	those 	labeled, 	most 	meat 	franks 	contained 	about 	30% 	fat, 	that 	is,
the maximum amount allowed under the current standard. 	This is what one
would expect 	to observe in a 	lemons market. 19 	However, the supermarkets
also 	carried 	chicken 	and 	turkey 	franks. 	Manufacturers 	of 	chicken 	and
turkey franks, which are not subject to a recipe standard, generally provide
a 	complete 	nutritional 	disclosure 	on 	their 	package 	labels, 	advertise 	their
products as 	"lower 	fat," 	and 	state that 	their 	products 	have 	at 	most 20-25%
fat 	compared 	to 	the 	30% 	fat 	standard 	for 	meat 	franks. 	Clearly, 	these
producers believe that lower 	fat provides a competitive advantage for 	their
products 	and 	they 	voluntarily 	provide 	fat 	content 	information 	about 	their
own products relative to those of their competitors. 2o
19 There 	was 	one 	exception. 	Eckrich 	"Lean 	Supreme 	Jumbo 	Franks"
contained 23% fat. 	Compared to the same company's "regular" franks, their
lean franks contain about 4 more grams of meat and 4 less grams of fat per
56.7 gram frank. 	Contrary to what one might expect, the lean franks were
lower 	in 	price 	than 	the 	regular 	franks. 	Prices, 	for 	"beef 	franks" 	with
identical 	fat 	and 	protein 	contents, 	varied 	substantially 	from 	$1.58 	to $2.49
per 	pound 	at 	the 	same 	outlet 	on 	the 	same 	day 	(April 	3D, 	1987; 	Giant,
Arlington). 	Thus, some consumers are willing 	to 	pay a 	premium 	for some
brands of hotdogs even though their nutritional content is the same as store
brands. 	In a 	pure lemons 	market, no one pays a 	premium 	because no one
can distinguish the quality of one brand from another.
20 	Although 	FSIS 	does 	not 	require 	nutritional 	labeling 	even 	when 	a
low fat claim is made for chicken franks, labels still 	must be approved for
use. 	Some of 	the 	voluntary 	disclosures may have been 	influenced 	by 	FSIS
labeling 	requirements 	for 	meat 	franks, 	even 	though 	they 	do 	not 	formally
apply. 	We also note that chicken franks are substantially cheaper than meat
franks. 	Prices varied 	between 89 and 99 cents per pound in the same store
10

Thus, 	although 	there 	is some 	evidence 	that 	a 	completely 	unregulated
cooked sausage market might fail 	because of credence aspects, the evidence
with 	respect 	to 	the 	existing 	market 	is 	mixed 	and 	scanty. 	There 	does 	not
appear to have been any extensive study of tnis market either prior to the
1969 recipe standard or thereafter. 	Nor has there been any careful study of
how 	the absence of a 	recipe standard affects the operation of 	the chicken
frank market. 	It is clear, however, that the perception that the unregulated
sausage market would or had failed 	played an 	important role in generating
the current regulations.
Even if we knew that an unregulated cooked sausage market would fail
due to lack of consumer information about fat content, the question of how
best to remedy the failure would still require resolution. 	A recipe standard
as currently 	used is one possibility, but there are other possibilities as well.
These 	include 	mandatory 	disclosure 	requirements 	or 	a 	regulation 	allowing
manufacturers 	to 	choose 	between 	conforming 	to 	a 	recipe 	standard 	or
providing disclosure. 	Recipe standards and mandatory disclosure requirements
clearly 	have 	different 	costs and 	may 	produce 	different 	benefits. 	We 	next
discuss circumstances under which one or the other may be more efficient.
V. Recipe Standards or Disclosures?
Recipe standards have advantages relative to disclosure requirements in
situations 	where 	it 	is 	difficult 	or 	expensive 	to 	provide 	consumers 	with
relevant 	information, 	and 	where 	consumer 	preferences 	are 	known 	and 	are
similar 	for 	all 	consumers. 	On 	the 	other 	hand, 	when 	disclosures 	are
and 	on 	the 	same 	day 	that 	meat 	franks 	varied 	between 	$1.58 	to 	$2.49 	per
pound.
II

relatively 	easy 	to 	make 	and 	consumer 	tastes 	are 	varied 	and 	changing,
disclosures 	may 	provide 	important 	benefits 	that 	cannot 	be 	attained 	with
recipe standards.
A. Comparative Benefits
The main advantage of a recipe standard relative to disclosure is that a
standard saves consumers 	the 	time and effort it takes to 	learn 	how to 	use
the 	information 	disclosed 	and 	the 	time 	it 	takes 	to 	read 	and 	compare
disclosures. 	A 	standard 	can 	eliminate 	"undesirable" 	foods 	(those 	that
knowledgeable consumers would not buy) when It is difficult or expensive for
consumers 	to 	become 	knowledgeable. 	Disclosure 	may 	be 	expensive 	either
because consumers would have to educate themselves to use it or because it
is 	physically difficult to make the disclosure. 	Many 	consumers understand
the significance of caloric intake for body weight, and many are aware of a
positive 	relationship 	between 	cholesterol 	and 	heart 	disease. 	Fewer,
presumably, 	would 	be 	aware 	of 	the 	increased 	risk 	of 	cancer 	from 	natural
toxins, such 	as hydrazines in 	raw mushrooms, or safrole in cinnamon leaf,
nutmeg and 	pepper. 21 	Effective disclosure in the latter cases might require
educational 	messages 	in 	addition 	to 	simple 	content 	disclosure. 	Disclosure
may be physically difficult for foods sold to be consumed away from home.
For example, ingredient or nutritional disclosure on the sausage package will
not 	be 	seen 	by 	a 	person 	who 	buys 	a 	hot 	dog 	at 	a 	ballpark, 	nor 	by 	a
customer who consumes unpackaged foods in a restaurant. 22
21 See 	Richard 	Wilson 	and 	E.A.C. 	Crouch, 	"Risk 	Assessment 	and
Comparisons: An Introduction," 	Science, v. 236, April 17, 1987, pp.267 - 280.
22 If reputation 	is 	not a sufficient 	incentive 	for 	concession owners at
ballparks 	or 	restaurant 	owners, 	then 	recipe 	standards 	may 	prevent 	the
deceptive sale of food consumed away from home. 	However, a general recipe
12

Recipe standards might also have advantages over disclosure if consumer
tastes 	are 	known 	and 	vary 	little. 	If a 	substance is 	known 	to 	be 	harmful
even in very small concentrations (e.g., arsenic), then clearly a standard that
prohibits 	its 	use 	In 	any 	recipe 	is 	more 	efficient 	than 	mandating 	its
disclosure. 	Or, if (to take another example) no consumer would knowingly
buy a sausage with more than 30% fat (at a price that covers costs), then a
standard that mandates a maximum fat content of 30% would save consumers
the time and trouble it would take to compare recipes across sausage brands
to avoid such sausages. 	Of course, it would not take long for producers to
learn 	what 	the standard setter is 	presumed 	to 	know: namely, 	that sausages
with 	more 	than 	30% fat 	do not sell. 	So disclosure 	would 	also be a 	viable
remedy in this case.
If tastes vary, disclosure may provide important benefits that a recipe
standard 	will 	not. 	For 	example, 	suppose 	consumers 	have 	different
preferences for fat in sausages, and are willing to trade-off fat for money at
different 	rates. 	An 	obvious advantage 	of disclosure 	over 	a 	standard 	that
imposes a ceiling of 30% on 	fat content 	is that, with 	disclosure, consumers
would be able to buy cheaper but fattier sausages, if they wished to do so.
A 	less 	obvious advantage 	is 	that disclosure 	might enable consumers to buy
low fat sausages that might not be marketed under a recipe standard. 	Under
a 	recipe standard, consumers 	may 	not 	be aware of 	the fat 	content of any
gIven 	brand 	of 	sausages 	unless 	the 	manufacturer 	discloses 	it. 	If 	no
manufacturer chooses to disclose fat content, then all sausages might contain
standard 	seems 	to 	be an 	inefficient 	way 	to deal 	with 	a 	problem 	that 	only
Occurs when the consumer docs not see the package label. 	A more efficient
solution would be to apply the recipe standard only to food consumed away
from home.
13

the maximum 30% fat, since these will be the cheapest to produce. 23 	Thus,
as compared to a recipe standard, disclosure may benefit consumers whether
they prefer more or less than the maximum fat allowed by the standard.
More generally, use of recipe standards requires that regulators decide
on 	which 	food 	characteristics 	are 	desirable, 	on 	how 	such 	characteristics
should be traded off against each other or against undesirable characteristics
(fat may taste good and provide nutrition, but too much may be unhealthy),
and 	on 	how 	all 	characteristics should 	be 	traded 	off 	against 	money 	(fat 	is
cheap). 	Moreover, 	since 	consumer 	tastes 	vary, 	the 	regulator 	must 	also
decide 	which 	consumer 	tastes 	should 	be 	satisfied 	and 	which 	not. 	As
ingredient prices, technology and tastes change, the standard setter must re-
evalua te all 	these decisions. 	These are 	f ormida ble problems. 	A standard
setter will always be forced to adopt an arbitrary "bright line" standard such
as 	"cooked 	sausages should 	never 	contain 	more 	than 	30o/~ fat."24 	Such 	an
arbitrary standard will lead to the loss of valuable product diversity, unless
the standard is so loosely set that few knowledgeable consumers would want
to purchase products that did not comply with the standard.	I
Disclosure regulations, in contrast, enable informed consumers to trade
off 	desirable 	and 	undesirable 	characteristics 	according 	to 	their 	individual
23 If some consumers are 	willing to 	pay as 	much, or, more, 	for 	lower
fat 	sausages, one would 	expect some' producers to 	voluntarily disclose 	their
product's 	(low) 	fat 	content. 	It 	is 	not 	known 	if, 	and 	to 	what 	extent,
producers advertised 	their products as "lower fat" 	during 	the 	1950 to 	1969
period 	in 	which 	the 	fat content of 	hot dogs increased 	by 50%. 	A careful
study 	of 	this period could shed 	much 	light on 	whether 	and 	how a 	market
failure might require some regulatory rcmedy.
24 Standards for similar products may bc very different. 	For examplc,
though "hot dogs," "franks," etc., may contain no mOrc than 30% fat, "Polish
sausage" 	may 	contain 	up 	to 	50% and 	"smoked 	sausagc" 	has 	no 	fat 	contcnt
limitations whatsoever.
14

preferences.
B. Comparative Costs
The major costs of recipe standards are: (1) they may decrease or retard
desirable 	innovation, 	and 	(2) 	they 	may 	be 	expensive 	to 	administer.
Mandatory 	disclosures, 	on 	the 	other 	hand, 	will 	not 	retard 	appropriate
innovation and entail little administrative cost. 	Their major costs appear to
be the foregone benefits of whatever messages would have appeared in their
stead and, perhaps, the cost of better quality control for manufacturers. 25
Recipe standards may reduce innovation and 	retard 	the rate at which
innovations are 	introduced. 26 	When 	a 	recipe standard 	applies, a 	firm 	that
has 	found 	a 	new 	and 	cheaper 	way 	to 	manufacture 	an 	equally 	nutritious
product covered by a recipe standard cannot market it until the old standard
has been amended or revoked or a new one promulgated. 	This may entail a
long and 	arduous process, especially 	if the effort is opposed 	by 	firms 	that
expect 	to 	be 	injured 	by 	a 	change 	in 	the 	existing 	standard. 27 	Ice 	cream
manufacturers, for example, who sought 	to amend 	the recipe to allow non-
25 The 	latter 	cost 	depends, 	in 	part, 	on 	the 	amount 	of 	sampling
variation 	permitted 	relative 	to 	the 	average 	percentages 	disclosed 	on 	the
label. 	We have no information on whether this is a significant problem for
cooked sausage manufacturers.
26 See 	Merrill 	& 	Collier, 	ibid., 	pp. 	602-603, 	and 	Steven 	Golden, 	"The
Effects of Government Policies on Technical Innovation in the Food Industry:
A Industry Perspective," in Critical Food Issues of the Eighties, edited by M
Chou and D. Harmon, Pergamon, 1979.
. I
27 The "safe and suitable" approach used by FDA in the breaded shrimp
standard, 	among 	others, 	reduces 	the 	cost 	of 	lost 	innovation 	by 	allowing
manufacturers 	flexibility 	with 	respect 	to 	new 	ingredients, 	which 	they 	may
use so 	long 	as thcy 	are 	safe and 	appropriate. 	See 	Robert 	Schaffner, "The
Effects of Government Policies on Technical Innovation in the Food Industry:
A Government Perspective," in Critical Food Issues of the Eighties, edited by
M. Chou and D. Harmon, Pergamon, 1979.
15

dairy 	substitutes 	for 	milk 	in 	ice 	cream 	were 	opposed 	by 	the 	Dairy
Association. 	Ingredient 	producers 	may 	have 	a 	vested 	interest 	not 	only 	in
preserving existing standards, but in creating new standards that require the
use 	of 	their 	ingredients. 	The 	Dairy 	Association 	has 	(unsuccessfully)
petitioned to amend 	the pizza standard to require real as opposed 	to ersatz
cheese in frozen pizzas.
The cost of promulgating or changing recipe standards may be large both	,
for the taxpayer and for the firms involved. 	Mandatory disclosure standards
also 	entail 	costs 	(e.g., 	the 	loss 	of 	information 	that 	would 	otherwise 	have
appeared 	on 	the 	label), 	but 	these 	costs 	may 	be 	small 	compared 	to 	those
incurred 	for 	recipe 	standards. 	For 	example,28 	in 	1958 	Proctor 	& 	Gamble
(P&G) 	began 	to 	market a 	new 	peanut 	butter called 	"Jif." 	Unlike 	the 	two
leading 	peanut 	butter 	brands 	of 	the 	time, 	"Skippy" 	and 	"Peter 	Pan", 	Jif
contained a 	blend of hydrogenated non-peanut oil in addition to peanut oil.
The new mixture made lif highly smooth and spreadable and P&G hoped this
innovation would attract a large market share. 	In 1959, the FDA proposed,
for 	the 	first 	time, 	a 	recipe 	standard 	for 	peanut 	butter. 	The 	proposed
standard would have precluded the marketing of lif under the name "peanut
butter." 	A legal battle involving the three major manufacturers and the FDA
ensued. 	The case was finally finished eleven years after the original FDA
proposal. 	The case ended in a victory for P&G; lif was peanut butter, but
"Skippy" 	and 	"Peter 	Pan" 	were 	not.
reformulate their successful products.
The 	two 	leading 	firms 	had 	to
In summary, recipe standards have advantages In remedying information
problems in situations where informing consumers IS very costly and where
28 Merrill & Collier, OD. cit., pp. 585-591.
16

consumer tastes are very similar. 29 	However, disclosure 	regulations provide
potentially 	important 	benefits 	that 	are 	not 	likely 	to 	be 	available 	under 	a
recipe standard. 	Moreover, recipe standards entail costs in terms of reduced
innovation and 	high administrative costs that are not entailed by disclosure
regulations. 	Therefore, 	the 	use 	of 	recipe 	standards 	should 	be 	limited 	to
cases where disclosure is clearly not sufficient or where it is infeasible.
We turn 	now 	to the question 	of whether 	the cooked sausage standard
should be amended or replaced by a disclosure requirement.
VI. Current Cooked Sausage Standard
FSIS regulation 319.180 deals with 	red meat "cooked sausage" products,
including 	frankfurters, 	franks, 	furters, 	hot 	dogs, 	wieners, 	viennas, 	bologna,
knockwurst, and similar 	products3o• 	It provides 	that such 	products contain
at 	most 	10% added 	water, and at 	most 30% fat. 31
water that is mechanically added to the product.
The "added 	water" 	is
In 	addition, 	FSIS 	has 	issued 	various 	administrative 	labeling 	policy
29 As 	previously 	indicated, 	where 	a 	substance 	is 	known 	to 	be
significantly 	harmful, 	then 	a 	standard 	that 	bans 	its 	use 	is 	clearly 	more
efficient than a mandatory disclosure of its presence.
30 "Franks" 	that 	are 	not 	red 	meat 	products 	are 	not 	covered 	by 	this
standard. 	FSIS has a separate "chili frank" standard. 	There is no "chicken
frank" standard.
31 Indirectly, 	the 	standard also 	requires 	that 	sausage 	products contain
at least 11.5% protein. Meat or meat byproducts and certain other sources of
protein 	must 	account 	for 	at 	least 	57.5% of 	the 	sausage (binders 	can 	be as
much as 3.5%). 	The regulation constrains the type of meat that may be used
in such a manner that the ratio of protein to water must be at least one to
four. 	Thus the sausage must contain at least 11.5% protein. 	Package labels
may show a smaller protein percentage because of downward rounding.
17

memoranda 	that establish guidelines for cooked sausage 	products 32• 	Under
present guidelines, a 	product that conforms to the standard can be called a
"lite" frank if it contains at most 22.5% fat. 33
If the 	manufacturer substitutes water for 	fat (thus exceeding the 	10%
maximum added water allowed under the standard), then he cannot call the
product 	a 	"frank" or 	a 	"low 	fat 	frank", 	or even 	describe 	it 	as 	a 	"low 	fat
frank with 15% added water." 	If he wants to use the word "frank", he must
label the product "imitation frank". 	Alternatively, he can make up a name
for 	the 	product 	(such 	as 	"lite 	link"), 	and 	describe 	it 	on 	the 	label. 	For
example, 	he 	might call 	it a "beef, water and 	isolated 	soy 	protein 	product".
The 	American 	Meat 	Institute 	(AMI) 	claims 	(understandably) 	that 	"such
nomenclature 	is 	unreasonably 	burdensome 	and 	has 	acted 	to 	inhibit 	the
marketing and sale of new, innovative products,"34
VII. The Proposed Rule and Alternatives
The present standard permits the manufacture of franks with less than
30% 	fat, 	but 	AMI 	asserts 	that 	the 	10% 	added 	water 	limitation 	makes 	it
difficult 	to 	produce a 	low 	fat sausage moist 	enough 	to 	be 	palatable at an
32 In 	the 	remainder 	of this 	comment 	we 	will 	(in 	most cases) 	use 	the
term "frank" to substitute for the more inclusive term "cooked sausage".
33 Similar 	guidelines 	exist 	for 	"low 	salt" 	and 	"very 	low 	salt" 	franks.
The fat of a 	frank 	can only be reduced by substituting meat for fat or by
substituting mechanically added water for fat. 	The current standard permits
the first, 	but not the second. 	As noted above, Eckrich markets a 	lower fat
frank, 	which 	conforms 	to 	the 	current 	standard 	and 	contains 	morc 	protein
than their regular frank.
3. See the discussion of AMI's petition by FSlS, 51 FR 42239.
18

acceptable cost. 	The FSIS proposed standard would 	require cooked sausage
products- to contain at most 30% fat and at most 40% fat and added water
combined. 	Thus, 	water could 	be substituted 	for 	fat 	at 	the 	manufacturer's
discretion. 	The 	proposed 	new 	standard 	would 	permit 	manufacturers 	to
produce 	palatable lower fat 	franks containing the same amount of 	protein,
without 	having 	to market 	them 	under a 	pejorative 	name. 	Most agree 	that
the availability of palatable and equally 	nutritious 	but lower fat 	franks 	is
desirable. 	Since 	the 	proposed 	standard 	will 	encourage 	desirable 	product
variety, it is superior to the current standard.
AMI 	proposes 	that 	FSIS 	promulgate a 	new 	and 	separate standard 	for
"lite" 	cooked 	sausages, 	rather 	than 	allow 	more 	flexibility 	under 	a 	single
standard. 35 	Under their proposal, FSIS would set "bright line" standards of
22.5% 	maximum 	fat 	and 	11.5% 	minimum 	protein content 	for 	lite 	franks. 36
The AMI proposal would also encourage the production and marketing of low
fat, 	palatable 	and 	equally 	nutritious 	franks, 	since 	it 	would 	allow
manufacturers 	to substitute 	mechanically 	added 	water 	for 	fat. 	Lower 	fat
franks, containing the same amount of protein as under the current sta!1dard,
could 	be produced at the same cost as regular franks. 	Therefore, the AMI
proposal also represents an improvement over the current standard.
On 	the 	other 	hand, 	the 	AMI 	proposal 	would 	not 	relieve 	the 	current
difficulties 	of 	producing 	low 	cost 	franks 	with 	fat 	content 	in 	the 	range
35 Currently 	there 	is 	no "lite" sausage standard, 	but 	FSIS does 	have a
an administrative guideline for the use of the word "lite" on the label. 	See
above, p. ] 8.
36 The 	AMI 	proposed standard 	would allow 	producers to 	mechanically
add 	water up to 20% of the product's weight. 	Since the ratio of protein to
naturally contained water is about one to four in meat, the 	11.5% minimum
protein content 	implies 	that at 	least 	57.5% (11.5% protein 	plus 46% 	natural
water) of the sausage will 	be meat or meat 	byproducts. 	If the fat content
is 22.5%, then mechanically added 	water will be 20%, or double the amount
allowed under the current standard. 	19

, .
between 22.5% and 30%. 	Such franks would not conform to AMI's proposed
"lite" standard, 	and 	to 	be 	palatable 	they 	would 	require 	more 	added 	water
than allowed under the current standard. 	Such franks would conform to the
FSIS proposed standard, and could have the fat content on their labels, but
still 	could 	not 	use 	the 	words 	"lite" 	or 	"light" 	under 	the 	AMI 	proposal. 31
Therefore, the FSIS proposal allows for greater frank 	variety than the AMI
proposal. 	Moreover, accepting the AMI approach seems to suggest that FSIS
should 	initiate a new rulemaking each time a different "lite" or low calorie
bright 	line 	standard 	is 	proposed 	for 	any 	product 	under 	its 	jurisdiction.
Surely less expensive alternatives exist.
While the FSIS proposal appears superior to the AMI proposal, even the
FSIS 	proposed 	standard 	will 	entail 	reduced 	product 	variety, 	reduced
innovation, and substantial administrative costs. 	These costs will be incurred
despite 	the 	lack 	of detailed 	evidence 	that, 	without 	a 	standard, 	the cooked
sausage market 	would 	fail 	to provide the variety of franks 	that consumers
are willing to pay for. 	In our view, if any regulation is required to prevent
a 	market 	failure, 	mandatory 	disclosure 	of 	fat 	content 	would 	provide
consumers larger net benefits than the FSIS proposed standard. 38
37 Under current FSIS policies, franks 	with 	fat content in 	the 25% to
30% range, would 	not be allowed 	to use the terms "lite" or "light" on their
labels.
38 As we noted above, the current standard indirectly sets a mlOlmum
protein content for franks. 	Absent the standard, if consumers were unaware
of 	protein 	content, firms 	might 	produce lower 	protein 	franks 	because 	they
are 	cheaper 	to 	produce. 	Since 	protein 	is 	a 	desirable 	characterisitic, 	we
might expect firms to disclose protein content voluntarily. 	Some do now, 	It
is possible, however, that consumers who were previously unaware of the low
protein content of franks would react adversely to all franks once disclosure
is 	made. 	In 	that case, 	no 	manufacturer 	would 	voluntarily 	disclose 	protein
content. 	Thus, 	one 	could 	conceivably 	make 	a 	case 	for 	the 	mandatory
disclosure of protein in addition to fat content.
20

An alternative remedy that 	might be worth consideration 	would 	be to
give 	producers 	the option 	of 	either 	meeting 	the 	revised 	FSIS standard, or
making 	a 	clear 	and 	conspicuous 	disclosure 	of 	fat 	content. 	Makers 	of
products 	that 	conform 	to 	the 	standard 	would 	not 	have 	to 	make 	the
disclosure. 	Firms that choose to depart from the presumptive standard could
market non-conforming products if they made the disclosure.
VIII. Conclusions
The standard 	proposed 	by 	FSIS is superidr 	to the current standard 	in
that it would encourage the production and marketing of franks that are 	as
palatable 	and 	nutritious 	as 	those 	conforming 	to 	the 	present 	standard 	and
that 	contain 	less 	fat. 	Although 	the 	proposed standard 	is 	an 	improvement,
our analysis indicates that disclosure may be better than any recipe standard.
Mandatory 	disclosure 	of 	fat 	content 	appears 	to 	be 	a 	sufficient 	remedy 	to
prevent any market failure that may exist due to consumer ignorance. 	Such
disclosure 	also seems superior 	to any of 	the proposed 	standards. 	It 	would
encourage valuable product variety and innovation and would probably reduce
administrative 	costs. 	It 	would 	also 	facilitate 	consumer 	search 	and 	choice
among 	cooked 	sausage 	products 	by 	providing an 	easy 	way 	to 	compare 	the
content 	of 	various 	products. 	In 	particular, 	disclos'ure 	would 	more
effectively 	facilitate 	shopping 	for 	low 	fat 	sausages 	than 	would 	a 	recIpe
standard. 	Another 	alternative, 	worth 	consideration, 	is 	to 	give 	producers 	a
choice of either meeting the revised FSIS standard or disclosing fat content.
21