United States Department of Agriculture Food Safety and Inspection Service Office of Policy and Program Development Washington, DC  20250-3700

REGULATIONS AND DIRECTIVES DEVELOPMENT

The U. S. Department of Agriculture (USDA), Food Safety and Inspection Service (FSIS) has responsibility for the appropriate labeling of raw meat/poultry products as mandated under the Federal Meat Inspection Act and Poultry Products Inspection Act.

Regulations published in January 1993 established a voluntary nutrition labeling program for single ingredient raw meat and poultry products, and allowed quantitative nutrition information to be supplied by means of point-of-purchase materials. Every two years, FSIS must survey retailers for significant participation in the Voluntary Nutrition Labeling Program. A total of 45 major cuts of meat and poultry have been identified to measure voluntary nutrition labeling participation. (See pp. 37 for a listing of these foods and an example of the nutrition information.)

Regulations published in 1994 made safe handling instructions mandatory on all raw meat and poultry product labeling. The scope and design of the nutrition labeling survey includes data which estimates the prevalence of stores that are providing safe handling instructions for raw meat and poultry items packaged at the retail level.

A nationally projectable survey was conducted in June 1995 to measure compliance with these requirements. On the nutrition labeling portion of the survey, participation by retailers was determined to be "significant". At that time, survey criteria encompassed both "new" and "old" formats for nutrition information at the point-of-purchase. The current survey, however, considers only "new" materials (located within the appropriate department) as the current measure of compliance. The June 1995 survey also reported safe handling instructions for raw meat/poultry to be less than the mandatory 100%.

This report summarizes the results of the Nutrition Labeling Safe Handling Instructions Survey for Raw Meat and Poultry conducted in December 1996, in order to comply with FSIS regulations that require a review of the program every two years.                                                        

[Back to Table of Contents]

Store samples are designed to be representative of the larger universes from which they are derived. The extent to which data and assumptions extracted from this sub-sample deviate from the larger actual universe produces a "sampling error". The size of the sample, relative to its data universe, predetermines the precision (i.e., degree of certainty or sampling error) of estimates governing the final evaluation of survey data.

The United States Department of Agriculture/Food Safety and Inspection Service (USDA/FSIS) determined that a scientifically drawn nationally representative sample of approximately 2,000 retail food stores should be selected to obtain the information necessary to assess compliance with the guidelines for the voluntary nutrition labeling of raw meat and poultry. A sample of approximately 2000 stores provides a relatively narrow margin of error around an observed compliance level. For example, for a sample of 2000, the error for an identified compliance level of 50% would be a maximum of plus or minus 4 percentage points, with a degree of statistical confidence of 95%. In other words, for a compliance level of 50%, we would be 95% certain that any and all other assessments of compliance (using the same sample size and sample design) would provide estimates falling somewhere between 46% and 54%. Furthermore, as the percent of compliance increases, the percent of uncertainty decreases, with additional increases in sample size achieving only small, and diminishing, reductions in uncertainty.

While sample size determines the precision of survey estimates, intricacies in the sample design determine how closely the sample represents the total population. In order to ensure representativeness to the overall national population of food retailers, USDA/FSIS include four characteristics in the selection of the sample:

• Store Sales Volume
• Store Type
• State
• County Size

Two levels representing average annual sales were established in the sample design to segment the sale volume of retailers. The volume factor is indicative of both the size of the store and the portion of the market it serves. The two annual sales levels are:

• Large Stores ($2 million or more in sales)
• Medium/Small Stores (sales between $500,000 and $2 million)

Store Type was based upon definitions, which placed each food retailer in one of two categories:

• Chain (four or more stores under common ownership)
• Independent (an independently operated store)

The inclusion of regional variation indicators, State and County Size, provided the means of representing highly urbanized, urbanized, rural and very rural county factors from all over the country in the survey sample. All states in the continental United States were include in the sample design, as well as four levels of county size based upon Metropolitan Areas that were established by the Office of Management and Budget. County size categories included:

• B Counties - All counties not included in A that are either over 150,000 population or in CMSAs or MSAs with population of at least 150,000.

• C Counties - All counties not included in A or B that are either over 40,000 population or in MSAs with over 40,000 population.

• D Counties - All other counties.

Data collection for the assessment of the prevalence of nutrition labeling information compliance was completed by Retail Diagnostics, Inc., an independent market research contracting firm located in Oradell, New Jersey.

RDI conducts a monthly syndicated observation service in a sample of food stores. The retail food store listings used by the contractor are comprehensive and subject to a continual updating process which reflects store openings, take-overs, and other developments. To obtain this sample, the RDI sample was sub-sampled as needed, and additional sample stores selected for those components not ordinarily covered by RDI’s syndicated survey. A scientifically drawn, nationally representative sample of retail stores based upon sales volume, store type, state, and county size, was selected by RDI. A sample of 2000 food retailers was designated to accommodate the survey criteria and reflect the test design. (See Table 1 and Table 2.) A breakdown of the store sample by County Type within State is included on page 33 of this report.

After receiving training in the contract requirements and data collection, field representatives were asked to enter and survey each of the 2,000 retail food stores in the sample. Data was collected during a two-week period beginning in mid December 1996, within the two reassessment period required by regulations. Field representatives inspected raw meat and raw poultry departments to determine the on-site status of nutrition labeling information for these products. Vehicles displaying information (e.g., signs, posters, brochures, notebooks, pamphlets, etc.) were studied and relevant data were recorded on a form, custom designed by RDI specifically for use in this survey. Data were reviewed for accuracy and completeness, entered into a computer, tabulated by the contractor, and delivered to USDA/FSIS.

Data descriptive of compliance are reported on two bases: Store Count and All Commodity Volume. Store Count represents the actual number of sampled stores visited in the survey. AACV data, on the other hand, are weighted estimates that represent annual store sales volumes and reflect the percent of the market serviced. Furthermore, ACV data approximates, more representatively than Store Count, the percent of the population exposed to the nutrition labeling information. While USDA/FSIS determined that substantial compliance would be met if at least 60% of the food retailers sampled provided nutrition labeling (as specified in the guidelines) for at least 90 percent of the foods they sell, the agency strongly believes that ACV data provides valuable information and should also be reported.

For example, a store count compliance estimate of 60% means that 60% of the stores sampled provided nutrition labeling information. An ACV compliance estimate of 60% is interpreted to mean that the stores providing nutrition labeling information sell 60% of all products sold in all stores. A second, equally appropriate interpretation of a 60% ACV compliance estimate is that at least three out of five of all consumers are being exposed to the nutrition labeling information.

[Back to Table of Contents]

The aggregate percentages for previous compliance determinations that were derived from Nutrition Labeling Information Survey of June 1995 were deemed significant and substantial for raw meat/poultry (66.5% Store Count/72.2% ACV) with regard to the 60% compliance goal and the presence of both "old" and "new" nutrition information vehicles. In June 1995 vehicles with the "new" nutrition information had just recently been printed/manufactured and had just begun to be distributed to stores via the efforts of the Food Marketing Institute and other organizations. Therefore, not all stores had access to the "new" vehicles at that time. Fair credit was given to any store, which displayed a vehicle, "old" or "new" as an indicator of intent to comply. For assessment purposes, a store which displayed the "old" format vehicles was no less compliant than a store which displayed the "new" format vehicles. When discounting the use of "old" vehicles, compliance was much lower (53.7% Store Count / 59.1% ACV) for all stores in the survey.

The criteria for the December 1996 survey were shaped by different circumstances. By the time of the December 1996 survey, the "new" format vehicles had been in distribution for a period of time deemed sufficient for all stores to have had access to them. Therefore, the December 1996 survey criteria restricted crediting compliance to only those situations where the "new" vehicles were in use. If a store used only the "old" vehicles, it was regarded as non-complying. Therefore, a comparison of compliance results between June 1995 and the current survey necessarily reflects the different criteria used to assess the data.

The percent of stores which were using "new" vehicles in June 1995 (53.7%) is slightly lower than the percent of stores using "new" vehicles during December 1996 (57.7%). The present survey did not include a measure of the number of stores still using "old" format vehicles. Stores which did not comply in this survey either had no nutrition information present in the store or had only "old" format vehicles at the time of the survey. The 8.8 percentage point decrease in compliance represents a decline in voluntary participation in the program by food retailers.

The December 1996 survey results for raw meat/poultry indicates that voluntary nutrition labeling compliance by food retailers has dropped from 66.5% in June 1995 to 57.7% in December 1996, or a drop of almost 9 percentage points. (See Table 3.) On a store count basis, the current compliance level of 57.7% slightly exceeds the 53.7% benchmark for the "new" vehicles present 18 months ago. Based upon ACV (All Commodity Volume) the margin widens by 11 percentage points to 60.9% in December 1996 versus 72.2% in June 1995. (See Table 4.) On an ACV basis, December 1996 compliance just achieves the target goal of 60%; on a store count basis, December 1996 compliance still falls short of the target goal by over 2 percentage points.

Examination of the data in greater detail shows that large chain stores continue to outperform independents by a wide margin (64.2% versus 44.8% on a store count basis and 64.3% versus 48.5% on an ACV basis). Similarly, large stores continue to outperform medium/small stores by a similar margin (61.7% versus 39.7% on a store count basis and 61.9% versus 38.0% on an ACV basis). (See Table 5.) A/B counties only slightly outpaced C/D counties in terms of overall compliance (57.9% versus 57.1% on a store count basis and 61.9% versus 58.3% on an ACV basis). (See Table 6.)

In order to meet the compliance standard a store needed accurate nutrition labeling in some form for 90% of all the major cuts of raw meat and poultry stocked in that store. Looking at the compliance range detail (Table 7) it is interesting to note that 32.3% of the stores were not in compliance for any of their raw meat/poultry products. The remaining 10% of the stores had sporadic nutrition information, which never achieved the required level of covering 90% of the survey items stocked in their stores.

A look at the broad categories of raw meat/poultry reveals that Beef/Veal, Pork/Lamb, and Chicken/Turkey all achieved similar levels of compliance (61.9%, 61.0%, and 63.1% respectively based on store count). (See Table 8.) The net compliance level of all raw meat/poultry is generally lower (57.7%) due to the fact that a store could be in compliance on beef/veal and not pork/lamb or chicken/turkey or any other combination thereof. The net compliance was determined by taking all of the survey items stocked in the store and determining if a 90% level was achieved across all of these items. The date indicates that there were instances of individual category compliance within stores where the store did not achieve net compliance encompassing all of these items.

The most common vehicle for displaying labeling information when present continues to be the poster (91.5% of all complying raw meat/poultry labeling). (See Table 11, Table 12, and Table 13.) This figure is significantly larger than the corresponding figure from June 1995 (66.3%) and may represent a trend towards utilizing this method for providing nutrition information at the retail level. However, it is also important to note that more stringent criteria were in effect in the December 1996 survey which may have some effect upon the reported percentages. In June 1995, other forms of nutrition labeling (such as "label affixed to package") were recorded even if posters were also present. In the current survey, other forms of nutrition labeling were only recorded if the primary forms (poster, panel, pamphlet, brochures, and notebooks) were not present. Therefore, a comparison of the percentages between the two studies must be made with an awareness of the different methodologies and the subsequent differences in the base numbers from which the percentages were derived.

The measure for compliance for safe handling instructions labeling required that the store have the appropriated "safe handling instructions" label affixed to each and every item stocked in the fresh meat/poultry category, including but not restricted to, the 45 items included in the nutrition labeling portion of the survey. The presence of this labeling is mandatory and the target goal is 100%.

Eighteen months ago in the June 1995 survey, a safe handling compliance level of 92.2% was achieved at the aggregate level. The data for the current survey indicates that the level of compliance in this mandatory area has remained essentially the same, rising slightly to 93.3%. (See Table 18 and Table 19.) All of the meat and poultry categories performed within a 1 to 1.5% range of their previous June 1995 results. Large chain stores remained essentially the same (+0.5%), with large independents and medium/small independents exhibiting only slightly better increases (+1.7% and 2.0% respectively).

[Back to Table of Contents]

[Back to Table of Contents]

[Back to Table of Contents]

[Back to Table of Contents]

Table 9

[Back to Table of Contents]

Table 10

[Back to Table of Contents]

Table 12

[Back to Table of Contents]

[Back to Table of Contents]

Table 15

[Back to Table of Contents]

Table 16

[Back to Table of Contents]

Table 17

[Back to Table of Contents]

[Back to Table of Contents]

Table 20

BY MEAT CATEGORY

SUMMARY OF SAFE HANDLING LABELING COMPLIANCE
BY MEAT CATEGORY

[Back to Table of Contents]

SUMMARY OF SAFE HANDLING LABELING COMPLIANCE
BY MEAT CATEGORY

[Back to Table of Contents]

USDA NUTRITION LABELING/SAFE HANDLING
INFORMATION STUDY DECEMBER 1996

SAMPLE COMPOSITION BY COUNTY SIZE WITHIN STATE

P E R C E N T O F  S T A T E  S A M P L E
STATE	A	B         B	C         C	D	TOTAL	SAMPLE         SIZE
ALL STATES	35.0	29.6	17.7	17.7	100.0	2000
ALABAMA	5.0	57.5	20.0	17.5	100.0	40
ARKANSAS	.	40.0	20.0	40.0	100.0	25
ARIZONA	58.6	17.2	17.2	6.9	100.0	29
CALIFORNIA	77.6	15.9	4.0	2.4	100.0	208
COLORADO	45.8	12.5	16.7	25.0	100.0	24
CONNECTICUT	27.3	72.7	.	.	100.0	22
DC	100.0	.	.	.	100.0	5
DELAWARE	100.0	.	.	.	100.0	4
FLORIDA	46.1	34.3	13.7	5.9	100.0	102
GEORGIA	41.4	19.0	13.8	25.9	100.0	58
IOWA	.	34.8	26.1	39.1	100.0	23
IDAHO	.	30.0	30.0	40.0	100.0	10
ILLINOIS	67.1	12.3	11.0	9.6	100.0	73
INDIANA	18.2	43.2	18.2	20.5	100.0	44
KANSAS	25.0	25.0	10.0	40.0	100.0	20
KENTUCKY	10.5	34.2	15.8	39.5	100.0	38
LOUISIANA	.	56.4	23.1	20.5	100.0	39
MASSACHUSETTS	55.6	27.8	16.7	.	100.0	36
MARYLAND	71.1	.	13.2	15.8	100.0	38
MAINE	.	30.0	50.0	20.0	100.0	10
MICHIGAN	44.4	30.6	9.7	15.3	100.0	72
MINNESOTA	33.3	18.5	18.5	29.6	100.0	27
MISSOURI	47.7	9.1	13.6	29.5	100.0	44
MISSISSIPPI	.	37.0	18.5	44.4	100.0	27
MONTANA	.	.	54.5	45.5	100.0	11
NORTH CAROLINA	.	56.4	30.8	12.8	100.0	78
NORTH DAKOTA	.	.	57.1	42.9	100.0	7
NEBRASKA	.	58.3	.	41.7	100.0	12
NEW HAMPSHIRE	44.4	.	33.3	22.2	100.0	9
NEW JERSEY	94.5	.	5.5	.	100.0	55
NEW MEXICO	40.0	.	40.0	20.0	100.0	10
NEVADA	.	54.5	27.3	18.2	100.0	11
NEW YORK	62.6	25.2	9.2	3.1	100.0	131
OHIO	36.3	38.6	16.3	8.8	100.0	80
OKLAHOMA	.	51.7	17.2	31.0	100.0	29
OREGON	.	66.7	20.0	13.3	100.0	30
PENNSYLVANIA	51.2	30.0	16.5	2.2	100.0	91
RHODE ISLAND	.	100.0	.	.	100.0	7
SOUTH CAROLINA	.	62.5	20.0	17.5	100.0	40
SOUTH DAKOTA	.	.	50.0	50.0	100.0	8
TENNESSEE	.	62.0	6.0	32.0	100.0	50
TEXAS	48.5	26.5	10.6	14.4	100.0	132
UTAH	.	75.0	.	25.0	100.0	12
VIRGINIA	28.1	45.6	8.8	17.5	100.0	57
VERMONT	.	.	63.6	36.3	100.0	19
WASHINGTON	53.7	24.4	12.2	9.8	100.0	41
WISCONSIN	29.3	24.4	26.8	19.5	100.0	41
WEST VIRGINIA	.	42.9	57.1	.	100.0	14
WYOMING	.	.	.	100.0	100.0	7

Send mail to webmaster  with questions or comments about this web site.
Last modified: November 25, 2002