Companies Opinions and Acceptance of Global Food Safety Initiative Benchmarks after Implementation

Transcription

1 1660 Journal of Food Protection, Vol. 75, No. 9, 2012, Pages doi: / x.jfp Copyright G, International Association for Food Protection Companies Opinions and Acceptance of Global Food Safety Initiative Benchmarks after Implementation PHIL CRANDALL, 1 * ELLEN J. VAN LOO, 1,2 CORLISS A. O BRYAN, 1 ANDY MAUROMOUSTAKOS, 3 FRANK YIANNAS, 4 NATALIE DYENSON, 4 AND IRINA BERDNIK 3 1 Department of Food Science, University of Arkansas, Fayetteville, Arkansas 72704, USA; 2 Department of Agricultural Economics, Bioscience Engineering Faculty, Ghent University, Coupure links 653, B-9000 Ghent, Belgium; 3 Agricultural Statistics Laboratory, University of Arkansas, Fayetteville, Arkansas 72701, USA; and 4 Walmart Stores, Bentonville, Arkansas 72716, USA MS : Received 16 December 2011/Accepted 7 May 2012 ABSTRACT International attention has been focused on minimizing costs that may unnecessarily raise food prices. One important aspect to consider is the redundant and overlapping costs of food safety audits. The Global Food Safety Initiative (GFSI) has devised benchmarked schemes based on existing international food safety standards for use as a unifying standard accepted by many retailers. The present study was conducted to evaluate the impact of the decision made by Walmart Stores (Bentonville, AR) to require their suppliers to become GFSI compliant. An online survey of 174 retail suppliers was conducted to assess food suppliers opinions of this requirement and the benefits suppliers realized when they transitioned from their previous food safety systems. The most common reason for becoming GFSI compliant was to meet customers requirements; thus, supplier implementation of the GFSI standards was not entirely voluntary. Other reasons given for compliance were enhancing food safety and remaining competitive. About 54% of food processing plants using GFSI benchmarked schemes followed the guidelines of Safe Quality Food 2000 and 37% followed those of the British Retail Consortium. At the supplier level, 58% followed Safe Quality Food 2000 and 31% followed the British Retail Consortium. Respondents reported that the certification process took about 10 months. The most common reason for selecting a certain GFSI benchmarked scheme was because it was widely accepted by customers (retailers). Four other common reasons were (i) the standard has a good reputation in the industry, (ii) the standard was recommended by others, (iii) the standard is most often used in the industry, and (iv) the standard was required by one of their customers. Most suppliers agreed that increased safety of their products was required to comply with GFSI benchmarked schemes. They also agreed that the GFSI required a more carefully documented food safety management system, which often required improved company food safety practices and increased employee training. Adoption of a GFSI benchmarked scheme resulted in fewer audits, i.e., one less per year. An educational opportunity exists to acquaint retailers and suppliers worldwide with the benefits of having an internationally recognized certification program such as that recognized by the GFSI. * Author for correspondence. Tel: ; Fax: ; crandal@uark.edu. The mission of the Global Food Safety Initiative (GFSI) is to provide continuous improvement in food safety management systems that will increase consumers confidence in the delivery of safe food worldwide (2). GFSI was initially released in May 2000, following extensive consultations among major international retailers and regulatory bodies in a unified response to a number of food safety scares. GFSI provides real-time information to manage food safety, minimize risks from foodborne pathogens, manage costs associated with establishing a dynamic and effective food safety culture, and maintain consumer confidence in our food and retail industries (2, 3). Numerous conflicting and competing industry specific regulations and food safety standards have caused confusion, imposed trade barriers, and may have unnecessarily increased the costs of food (9). GFSI has made a concerted effort to benchmark existing international food safety standards against expert-derived, unifying food safety criteria. Each approved GFSI food safety scheme is now benchmarked against a common criterion to create a harmonized and verified control system to produce and deliver safe foods. The GFSI vision is once certified, accepted everywhere (2), thus requiring only one benchmarked food safety scheme for each company under the GFSI framework. The hope is that the GFSI benchmarked scheme will improve food safety and enhance cost efficiency in the supply chain by reducing the duplication of third party audits or additional customer audits, which can add unnecessarily to the price of food. However, no scientific studies exist on the impact of GFSI certification. Many retailers, including Walmart Stores (Bentonville, AR), have agreed to accept any of the GFSI benchmarked schemes, and more food service, retail, and manufacturing companies are following this approach. In February 2008, Walmart announced that all of their private (store) brand suppliers and some of their national brand suppliers would be required to become certified based

2 J. Food Prot., Vol. 75, No. 9 COMPANIES OPINIONS OF GLOBAL FOOD SAFETY INITIATIVE 1661 TABLE 1. Twelve GFSI-recognized schemes from the GFSI Guidance Document, version 5 a Manufacturing schemes Primary production schemes Primary and manufacturing scheme British Retail Consortium (BRC) Global Standard, version 5 Dutch Hazard Analysis Critical Control Points (HACCP) (option B b ) Canada Good Agricultural Practices Global Good Agricultural Practices Integrated Farm Assurance Scheme, version 3 Food Safety System Certification (FSSC) Safe Quality Food (SQF) 1000, level 2 Global Aquaculture Alliance (GAA) Best Aquaculture Practices (GAA Seafood Processing Standard) Global Red Meat Standard, version 3 International Food Standard, version 5 SQF 2000, level 2 Synergy PrimusGFS a GFSI Guidance Document, version 6 has been released recently (January 2011) (3). b The Dutch HACCP scheme was previously recognized in the GFSI Guidance Document, version 5 but has not been resubmitted to GFSI for benchmarking because of the new more stringent requirements in version 6 (5). on one of the GFSI standards (11) (Table 1). The current study was conducted to evaluate the impact of on companies that selected and became certified based on one of the accepted GFSI benchmarked schemes by July 2009 (11). Opinions of suppliers and distributors regarding the implementation of GFSI benchmarked schemes were scored. Measures included the perceived benefits (for food safety), cost and time to become compliant, the effect on the number of third party audits, and the motivation to become GFSI compliant. Because the GFSI benchmarked schemes are gaining international attention and GFSI can be important for companies needing to comply with the U.S Food Safety Modernization Act (FSMA) (12), studies are needed to evaluate the impact of the implementation of GFSI benchmarked schemes from the perspective of the food suppliers who are trying to meet these standards. Results can be used to drive continuous improvement in food safety by communicating the benefits and opportunities of GFSI to stakeholders, i.e., industry, regulatory agencies, academia, and consumers. MATERIALS AND METHODS Questionnaire development. An in-depth review of the literature was conducted to provide background information about GFSI certification, its benefits and constraints, and the investment required to become certified for a GFSI recognized scheme. Only a limited number of scientific studies have been published regarding GFSI certification. Some documents about food safety certification published by the various agencies contain definitive findings on the impact of food safety certifications such as Integrated Farm Assurance (IFS) or on certification based on a GFSI-recognized scheme; however, no references were included to support their conclusions (4, 5). The survey questions for the present study were developed based on the reported findings about GFSI, our previous surveys (1, 8, 13, 14), and input from knowledgeable food safety professionals from academia, quality managers working in the food industry, and professional auditors for food safety certification programs. The questions were beta tested with a diverse group of suppliers (n ~ 17) who produce food products from various food categories and represent companies of various sizes. These suppliers filled out the draft questionnaire and gave further suggestions for improving the questionnaire in follow-up phone conversations. The final set of questions was reviewed by the retailer, the Institutional Review Board of the University of Arkansas, and food safety specialists. Topics of the survey questions included (i) suppliers individual food safety culture before and after embracing one of the benchmarked GFSI schemes, (ii) companies motivations for becoming certified for one of the GFSI schemes and information about their selection of a certain benchmark GFSI provider, (iii) cost associated with becoming GFSI compliant, and (iv) leading indicators of benefits of complying with GFSI benchmarked schemes. A complete copy of the survey is included here as an Appendix. Survey procedures. This nationwide survey was launched using the online survey software Qualtrics ( com), which allows a surveyor to launch and monitor the survey and answer questions from the respondents, all on-line. As a privately conducted survey, only those suppliers receiving an invitation were allowed to provide input. For online surveys, 7 to 10 days appears to be a sufficient amount of time for completion (6). As suggested by Instructional Assessment Resources (6), two reminders were sent during the 3-week survey period. A total of 309 national suppliers were first contacted by e- mail sent to the supplier list provided by the retailer. Initially, a cover letter from the retailer encouraging the suppliers to participate, a copy of the questions to be answered on-line, and the URL directing the user to the Qualtrics Web site were ed to one representative of each company. Typically, the responses were coordinated by the national account manager, national sales manager, vice president, director of quality assurance, or in some cases the business owner. Technical input was encouraged from GFSI benchmark certified practitioners and quality assurance personnel at each company. Demographics. Based on the foods produced, respondents were assigned by the surveyors to one of three risk categories based on published risk categorizations from the U.S. Food and Drug Administration Food Safety and Inspection Service (Washington, DC) and Food Standards Australia New Zealand (Canberra,

3 1662 CRANDALL ET AL. J. Food Prot., Vol. 75, No. 9 TABLE 2. Demographic information for the companies that supplied products Company type No. (%) of companies Annual sales ($) 0 50 million 29 (17) million 29 (17) million 27 (16) million 40 (23).500 million 49 (28) Food safety risk of products Low 83 (48) Medium 65 (37) High 26 (15) Australia; Wellington, New Zealand). The levels of risk were (i) high: raw red meat and poultry, raw cured and fermented meat and fish, and ready-to-eat meals, sandwiches, and desserts; (ii) medium: raw prepared products, fully cooked meat, poultry, or fish products, fresh and prepared fruits, vegetables and nuts, dairy, eggs, and confectionery; and (iii) low: the remainder of the products with low water activity or that were shelf stable. All three levels of food safety risk were represented by the products chosen for the survey (Table 2). To estimate and compare the different supply companies with respect to their size, the total annual sales range of each company was requested as a demographic response in the form of a multiple choice question (Q9). Five ranges were given:,$50 million, $51 to $100 million, $101 to $200 million, $201 to $500 million, and.$500 million. Based on the annual sales, a fairly even spread of company sizes was obtained (Table 2). This information allowed us to compare companies based on size and risk level of foods. Plant information. Information at the lowest level, the individual manufacturing plant, Q4, included the GFSI benchmarked scheme that is currently used by an individual plant. To assess the time required to become GFSI certified, suppliers were asked to provide start dates and completion dates for the GFSI certification process at each individual manufacturing facility. To determined whether GFSI certification reduced the number of third-party audits, the respondents were asked to list the annual number of audits for each plant before and after being certified (Q6). Many of the suppliers surveyed worked with more than one manufacturing plant to supply foods to Walmart. Companies with more than five plants selected their top five plants based on either the volume of product or the amount of sales to Walmart. Opinion about GFSI certification. To get a better understanding of the suppliers opinions about GFSI certification, response choices on a 5-point Likert scale (1, strongly disagree; 5, strongly agree) were provided to ascertain the respondent s level of agreement with statements about the costs of becoming GFSI certified (in terms of capital and time investments), potential benefits from GFSI certification (such as food safety benefits and reducing amount of audits), and the reasons for becoming GFSI certified (Q1). The survey also included two ranking questions: reasons for becoming GFSI certified (Q3) and reasons why a particular GFSI scheme was chosen over other possible schemes (Q5). These questions were developed based on results from the preliminary questionnaire with the beta testers, who were selected to be representative of the test population as a whole. In the beta testing phase, an open-ended response question was used for asking why the respondent became GFSI certified. Based on the written responses of the beta testers, a summary list of the most likely motivating factors for becoming certified was created for the national survey. Costs of becoming GFSI compliant. The questionnaire was originally designed to be both qualitative and quantitative, allowing estimates of the actual cost of becoming certified in one of the GFSI benchmarked schemes. However, feedback from the beta testers about their lack of historical data on food safety expenses made a quantitative approach impossible at this time. In Q1, two statements about the costs were included so suppliers could express their opinions. One question concerned the cost (in time and capital) of becoming GFSI compliant (Q1a), and the second concerned any increase in production costs due to GFSI requirements (Q1j). Statistical analysis. Data were analyzed with JMP (release 9.0, SAS Institute, Cary, NC), including both univariate and bivariate analyses. Correlations among continuous variables also were calculated and summarized. A one-way analysis of variance was performed to compare the summary statistics for all continuous variables for each group. Groups were defined based on categories of product safety risk and annual sales. Significance was determined based on an alpha level of Multiple comparison tests, such as Tukey s honestly significant difference and Hsu s multiple comparisons with the best, were used to identify significant differences between means. Paired t tests were performed to evaluate differences among means, comparing responses before and after certification. The categorical platform in JMP also was used to produce tables and graphs of multiple categorical responses overall and for each group with the appropriate chi-square tests. Various other exploratory graphs were created with Graph Builder in JMP to emphasize similarities and differences among distributions and summary statistics across groups. RESULTS AND DISCUSSION Demographic delineators. A total of 309 suppliers were contacted, and 174 total responses were received (56% of those contacted). The majority of the suppliers (98%) were primarily food manufacturers; only two suppliers were primarily distributors, and one supplier was primarily a grower. Because there were too few distributors and growers, this information could not be used to categorize the sample; therefore, responses from these three suppliers were pooled with those of the manufacturers. Based on annual sales, respondents represented a fairly even spread of company sizes (Table 2). The largest number of responses (49) was from companies with annual sales of $500 million or more, and the fewest number of responses (27) was from companies with annual sales of $101 to $200 million, similar to those from companies that sold in the lowest two ranges ($0 to $100 million; 29 responses each). Each of the 174 suppliers was assigned by the surveyors to one of the three categories of product. We used this risk designation to study differences among suppliers based on their product risk categories. Most respondents were in the low risk category (48%), followed by medium risk (37%), and high risk (15%) (Table 2).

4 J. Food Prot., Vol. 75, No. 9 COMPANIES OPINIONS OF GLOBAL FOOD SAFETY INITIATIVE 1663 TABLE 3. Level of agreement with statements regarding companies opinions about GFSI certification (Q1) Q1 part Statement Mean score a d Several of our customers require a GFSI certification to meet their requirements 4.18 A o Due to the implementation of a GFSI standard, our food safety management system is better documented 3.95 AB k Our company sees GFSI standards as beneficial to improving the safety of the products we produce 3.90 AB i Our company adopted a GFSI requirement because it is a WM requirement 3.81 BC a Our company has had to make significant additional investments 3.79 BCD c Our company has had to make changes to our food safety management system to meet the GFSI requirements 3.69 BCD e Several of our customers recognize a GFSI certification but do not require GFSI certification 3.64 BCD b Adoption of a GFSI standard has enhanced our company s ability to comply with food safety regulations 3.63 BCD l Due to GFSI requirements, we have more employee food safety training 3.48 CDE m Adoption of a GFSI standard has further enhanced our ability to produce safe food 3.44 DE g Adoption has resulted in less duplication of effort in the need to comply with the multiple third-party audits 3.18 EF n Due to the implementation of a GFSI standard, we have less product errors/defects 3.01 F j GFSI requirements have added significant costs to the production of our products 2.98 F f The amount of third-party audits has been reduced due to adoption of GFSI standards 2.91 FG h Adoption of a GFSI standard has resulted in a net savings on the total amount we spend on redundant thirdparty audits 2.61 G a Scores: 1, strongly disagree; 2, disagree; 3, neither agree nor disagree; 4, agree; 5, strongly agree. Means (n ~ 174) followed by different letters are significantly different (P, 0.05). Pooled responses on companies opinions about GFSI certification. Question 1 contained 15 statements (1a through 1o) asking for the respondents opinions about different statements. The level of their agreement with these statements was based on a 5-point Likert scale (1, strongly disagree; 5, strongly agree) (Table 3). Statements concerning GFSI and third party audits. Three statements were included to evaluate the perceived effect of GFSI on third-party audits (Q1f, Q1g, and Q1h). Our hypothesis was that third-party audits would be reduced after implementation of a GFSI benchmarked scheme. However, the suppliers neither agreed nor disagreed that GFSI reduced third-party audits (mean ~ 2.91). They also neither agreed nor disagreed that adaptation of a GFSI benchmarked scheme resulted in less duplication of efforts to comply with multiple third-party audits (mean ~ 3.18). Suppliers disagreed with the statement that after becoming GFSI compliant, less money was spent on third-party audits (mean ~ 2.61). Statements concerning GFSI, food safety, and quality. Six statements (Q1b, Q1c, Q1k, Q1l, Q1m, and Q1o) addressed the suppliers opinions on the effects of GFSI implementation on the safety of their food products. Respondents agreed that the implementation of a GFSI standard resulted in a better documented food safety management system (mean ~ 3.95), and they perceived that GFSI standards helped them improve the safety of their products (mean ~ 3.90). Most of the suppliers agreed that their company had to make changes to their food safety management system to meet GFSI requirements (mean ~ 3.69) and that their ability to comply with food safety regulations was enhanced by GFSI implementation (mean ~ 3.63). Other leading indicators of beneficial effects on food safety included more employee food safety training (mean ~ 3.48). The suppliers also indicated that GFSI enhanced their ability to produce safe food products (mean ~ 3.44). Thus, most suppliers agreed that implementation of a GFSI benchmarked scheme benefited the safety of their food products, which is a positive leading indicator for adopting a GFSI benchmarked scheme. Statements concerning GFSI recognition and requirements. Three statements (Q1d, Q1e, and Q1i) evaluated the suppliers opinions on GFSI requirements and recognitions. Suppliers agreed that several of their customers required GFSI certification (Q1d, mean ~ 4.18). Most suppliers indicated that they adopted a GFSI benchmarked scheme because it was a Walmart requirement (mean ~ 3.81) and that several of their customers recognized a GFSI benchmark certification but did not require it (mean ~ 3.64). Statements concerning GFSI and cost. Two statements reflected the suppliers opinions about the cost of becoming GFSI compliant. The statement adopting GFSI benchmarked scheme required our company to make significant additional investments in capital and staff time in food safety was agreed to by most suppliers (mean ~ 3.79). However, when suppliers were asked to respond to the statement GFSI requirements have added significant costs to the production of our products, the mean score was 2.98, indicating that the suppliers neither agreed nor disagreed that GFSI requirements added significant costs to the production of their products. Question 2 asked the respondents to classify the investments the company made to comply with additional GFSI requirements. Possible responses ranged from little or no additional investment to significant capital and staff time investment. Responses were fairly evenly clustered around the midpoint, labeled investments in critical infrastructure upgrades and minimal expenditures (27%); 22% of respondents gave this statement a score of 2 and 26% gave it a score of 4. Only 18% declared that they had made significant

5 1664 CRANDALL ET AL. J. Food Prot., Vol. 75, No. 9 FIGURE 1. Number of respondents who chose each reason for becoming GFSI certified (n ~ 174) (Q3). Responses: a, meet customers requirement to continue or start doing business; b, reduce number of third-party audits; c, remain competitive in the industry; d, enhance quality assurance; e, enhance food safety; f, improve reputation; g, improve documentation of food safety management system; h, develop more efficient and dynamic control of food safety hazards. capital investments, and 8% reported little or no additional investment. Risk categories. The overall responses were separated by the level of risk associated with the food produced (high, medium, or low) (Table 4). No differences in responses to most statements (scores) were noted among the three different risk groups. Suppliers in the high-risk category neither agreed nor disagreed (3.08) that GFSI adoption enhanced their ability to produce safe foods; their mean score was not significantly different from that of the medium risk producers (3.43), who somewhat agreed with the statement. Low-risk suppliers also slightly agreed (3.55), which was not significantly different from the score for the medium-risk suppliers. This finding may indicate that the group of suppliers that produce higher risk foods had already made significant investments of capital and time to establish well-documented food safety systems as part of their hazard analysis critical control point plans or other regulation-mandated programs. The high-risk group had significantly higher agreement with the statement several of our customers recognize a GFSI certification but do not require GFSI certification (4.12) than did the medium-risk (3.62) and low-risk (3.52) group. Thus, the suppliers that agreed most strongly with the statement were those that supplied the higher-risk foods to this large retailer and probably the same or very similar foods to other customers, who would probably be aware of the additional risks. These other customers were apparently aware of GFSI certification but had not required that their suppliers become compliant. Several significant differences among the responses to statements Q1d, Q1e, and Q1b were noted when sorted by annual company sales (Table 4). The smallest companies (annual sales of less than $50 million) agreed less strongly with the statement several of our customers require GFSI (Table 4). Significant differences but no clear pattern were noted in responses to adoption of a GFSI standard has enhanced our company s ability to comply with food safety regulations. These responses may have been more influenced by the distribution of risk groups (unevenly spread over the size range of companies). When responses were compared by risk group, there were no significant differences. When all companies were combined, the majority (52%) agreed that they were better able to comply with food safety regulations, and another 13% strongly agreed. Drivers for becoming GFSI certified. Respondents were asked to rank possible drivers for becoming GFSI certified from most important (1) to least important (8) (Q3). The most important reason for becoming GFSI certified was to meet customers requirements to continue their business relationship (response 3a) (Fig. 1). The second and third most important reasons to comply with a GFSI scheme were to enhance food safety (response 3e) and to enhance quality assurance (response 3d). When suppliers were divided according to food safety risk, suppliers of the highest risk foods ranked response 3a, meet customer s requirement to continue doing business almost unanimously (96%) as the most important reason; 74% of medium-risk suppliers and 77% of low-risk suppliers also chose response 3a as the most important reason for becoming GFSI certified. Supplier preferences for GFSI benchmarked schemes. Of the 12 benchmarked schemes recognized by GFSI at the time of this survey, only 6 were being used by the suppliers. One of the main benefits of using an initial beta test group was to develop a set of descriptors for the national questionnaire. Respondents were asked in an openended question to list two or three of the most important reasons why one benchmark certifier was selected over the others. These responses were used in the national survey when the participants ranked different possible reasons for selecting a particular GFSI scheme (Q5), and respondents

6 J. Food Prot., Vol. 75, No. 9 COMPANIES OPINIONS OF GLOBAL FOOD SAFETY INITIATIVE 1665 TABLE 4. Level of agreement with statements regarding companies opinions about GFSI by food safety risk level and of company size Mean scores a according to: Risk level Annual sales Q1 part Statement Low (n ~ 83) Medium (n ~ 65) High (n ~ 26) $0 $50 M (n ~ 29) $51 $100 M(n ~ 29) $101 $200 M(n ~ 27) $201 $500 M(n ~ 40).$500 M (n ~ 49) Total (n ~ 174) d o k i a c e b l m g n j f h Several of our customers require a GFSI certification to meet their requirements Due to the implementation of a GFSI standard, our food safety management system is better documented Our company sees GFSI standards as beneficial to improving the safety of the products we produce Our company adopted a GFSI requirement because it is a WM requirement Our company has had to make significant additional investments Our company has had to make changes to our food safety management system to meet the GFSI requirements Several of our customers recognize a GFSI certification but do not require GFSI certification Adoption of a GFSI standard has enhanced our company s ability to comply with food safety regulations Due to GFSI requirements, we have more employee food safety training Adoption of a GFSI standard has further enhanced our ability to produce safe food Adoption has resulted in less duplication of effort in the need to comply with the multiple third-party audits Due to the implementation of agfsistandard,wehave less product errors/defects GFSI requirements have added significant costs to the production of our products The amount of third-party audits has been reduced due to adoption to GFSI standards Adoption of a GFSI standard has resulted in a net savings on the total amount we spend on redundant third-party audits B 4.24 A 4.26 A 4.25 A 4.29 A B 3.62 B 4.12 A 3.66 AB 3.21 B 3.78 A 3.75 A 3.73 A BC 4.1 A 3.59 BC 3.78 AB 3.35 C A 3.43 AB 3.08 B a Scores: 1, strongly disagree; 2, disagree; 3, neither agree nor disagree; 4, agree; 5, strongly agree. Means followed by different letters are significantly different (P, 0.05).

7 1666 CRANDALL ET AL. J. Food Prot., Vol. 75, No. 9 TABLE 5. Frequency of use of GFSI benchmarked schemes among food production plants Scheme No. (%) ofplants SQF 2000 (level 2 or higher) 221 (54) British Retail Consortium (BRC) Global Standard, version (37) International Food Standard (IFS), version 5 14 (3) PrimusGFS 11 (3) SQF 1000 (level 2) 5 (1) Food Safety System Certification (FSSC) a 4 (1) Total 406 a Includes International Organization for Standardization (ISO) plus British Standards Institute PAS 220. were given an opportunity to add additional reasons. The most important reason for choosing a certain GFSI benchmarked scheme was because it was widely accepted by customers (50% of respondents, Fig. 2). Only 20% of respondents indicated that their most important reason for choosing a particular scheme was because one or more customers required it. The other three choices in descending order were good reputation (12%), most often used in the industry (10%), and recommendation (7%). Data pooled at the plant level. The 174 surveyed suppliers reported the GFSI schemes of 406 plants; many suppliers gathered products from more than one plant to supply Walmart. In some instances, different plants working with the same supplier implemented different GFSI benchmarked schemes. The most frequent GFSI benchmarked scheme among the 406 plants was the Safe Quality Foods (SQF; at level 2 or higher), representing 54.4% of the plants (Table 5). Within this group, 38% of the plants were SQF 2000 level 2 certified and 62% went beyond the current GFSI requirement and had a SQF 2000 level 3 certification. The second most popular benchmarked scheme, chosen by 37% of the suppliers, was the British Retail Consortium (BRC), version 5. Other schemes such as IFS, PrimusGFS, SQF TABLE 6. Frequency of use of GFSI benchmarked schemes among food suppliers Scheme No. (%) of suppliers Only SQF 2000 (level 2 or higher) 100 (57.8) Only BRC 53 (30.6) Only IFS, version 5 5 (2.9) BRC and SQF 2000 (level 2 or higher) 4 (2.3) Only PrimusGFS 4 (2.3) Only FSSC a 3 (1.7) SQF 2000 and SQF (1.2) Only SQF 1000 (level 2) 1 (0.6) BCR and SQF 1000 (level 2 or higher) 1 (0.6) Total 173 a Includes International Organization for Standardization (ISO) plus British Standards Institute PAS , and Food Safety System Certification (FSSC) each represented 3% or less of the plants in this survey. Data pooled at the supplier level. When evaluating the chosen GFSI benchmarked schemes for the 174 different suppliers, SQF was again the most popular option (Table 6). Most suppliers (96%) were certified for only one of the GFSI benchmarked schemes. Approximately 58% of the suppliers used only SQF 2000 for all of their plants. The second most common scheme was BRC, with 31% of the suppliers having only this certificate. Data pooled at the industry level. To identify the most prevalent GFSI benchmarked schemes in the various food industries, suppliers were categorized according to the type of food product produced (Table 7). In the bakery and meat industries, most plants had BRC certification (54 and 61% of plant for the bakery and meat industries, respectively), followed by SQF 2000 (level 2 or higher; 33 and 37% of the plants for the bakery and meat industries, respectively). In the oil and fat industries, BRC also was FIGURE 2. Ranking of reasons of why a certain GFSI benchmarked scheme was chosen (1, most important; 5, least important) (Q5): a, widely accepted by customers; b, required by a customer; c, recommendation from others; d, good reputation in the industry; e, most often used in our industry.

8 J. Food Prot., Vol. 75, No. 9 COMPANIES OPINIONS OF GLOBAL FOOD SAFETY INITIATIVE 1667 TABLE 7. Frequency of use of GFSI benchmarked schemes among the food production plants categorized according to type of product Type of product Scheme No. (%) ofplants per product type a Bakery BRC Global Standard, version 5 31 (54.4) International Food Standard 5 (8.8) (IFS), version 5 SQF 1000 (level 2) 2 (3.5) SQF 2000 (level 2 or higher) 19 (33.3) Cereals and BRC Global Standard, version 5 1 (6.7) snacks SQF 2000 (levels 2 and 3) 14 (93.3) Confectionery BRC Global Standard, version 5 1 (11.1) IFS, version 5 2 (22.2) SQF 2000 (level 2 or higher) 6 (66.7) Dairy and eggs BRC Global Standard, version 5 6 (15.0) FSSC b 1 (2.5) SQF 2000 (level 2 or higher) 33 (82.5) Dried foods and ingredients BRC Global Standard, version 5 9 (42.9) IFS, version 5 4 (19.0) SQF 2000 (level 2 or higher) 8 (38.1) Drinks BRC Global Standard, version 5 8 (19.0) FSSC b 2 (4.8) SQF 2000 (level 2 or higher) 32 (76.2) Fruits and vegetables BRC Global Standard, version 5 11 (22.9) PrimusGFS 11 (22.9) SQF 2000 (level 2 or higher) 26 (54.2) Meat BRC Global Standard, version 5 46 (61.3) IFS, version 5 1 (1.3) SQF 2000 (level 2 or higher) 28 (37.3) Oils and fats BRC Global Standard, version 5 7 (63.6) FSSC b 1 (9.1) IFS, version 5 2 (18.2) SQF 2000 (level 2 or higher) 1 (9.1) Other SQF 1000 (level 2) 2 (11.1) SQF 2000 (level 2 or higher) 16 (88.9) Total 336 a Number (percentage) of all plants producing that type of food product. b Includes International Organization for Standardization (ISO) plus British Standards Institute PAS 220. TABLE 8. Time to become GFSI compliant for suppliers of different sizes or risk categories (Q4) Company type Mean time to compliance (mo) a Annual sales ($) million 12.6 A million 11.8 AB 0 50 million 9.4 BC million 8.5 C.500 million 8.3C Food safety risk of products Medium Low High 10.4 A 10.2 A 7.0 B a Means followed by different letters are significantly different (P, 0.05). TABLE 9. Number of audits before and after becoming GFSI compliant (Q6) Company type 2 3 yr before complying with GFSI No. of audits After complying with GFSI Difference Overall Food safety risk of products Low Medium High Annual sales ($) 0 50 million a million million million million a Not significantly different from zero. preferred (64% of plants). For plants in the dairy and egg industry, 83% were SQF 2000 certified, and for the confectionery (67%) and drinks (76%) industries, most plants were SQF 2000 certified. For vegetables and fruit, most plants were SQF 2000 certified (54%); however, BRC (23%) and PrimusGFS (23%) also were common. The plants producing dried food and ingredients were nearly equally divided between BRC certified (43%) and SQF certified (38%). Time to become GFSI compliant. Q4 asked the supplier to respond at the individual plant level concerning the GFSI benchmarked scheme selected, the starting month and year, and the date the plant first became certified under this standard. It took an mean ( SD) of 9.7 ( 6.9) months from the start of working toward compliance with a GFSI benchmarked scheme to complete the process. Significant differences in the time to become GFSI compliant were noted based on the size of the company and the level of food safety risk of the foods produced (Table 8). On average, mid-size companies took the longest to become GFSI certified (11.8 to 12.6 months). For small and large companies, 2 to 3 months less were needed to become GFSI compliant compared with mid-size companies (Table 8). Companies producing food with a high food safety risk took only 7 months to become GFSI certified, in contrast to more than 10 months for companies producing foods with low and medium food safety risk (Table 8).These findings may indicate that plants producing higher risk foods were more advanced in their hazard analysis than were plants that produced lower risk foods. Number of audits. Question 6 was designed to collect quantitative data on one of the hypothesized benefits of becoming GFSI compliant, that of reducing the number of individual customer or third-party audits required by customers other than Walmart. Respondents reported an average of 4.71 audits per year before GFSI implementation,

9 1668 CRANDALL ET AL. J. Food Prot., Vol. 75, No. 9 with a lower quartile (25% percentile) of 3 audits per year and an upper quartile (75% percentile) of 6 audits per year (Table 9). After becoming GFSI compliant, the average number of audits decreased to 3.89 audits per year, the quartiles dropped to 2 (lower quartile) and 5 (upper quartile) audits per year. However, as the GFSI approach becomes increasingly more common in the food industry, more companies should become aware of the benefits (7, 11). Thus, the number of customer and third-party audits would continue to drop as more customers to whom the supplier sells accept the results of audits certified according to the GFSI benchmarked scheme used by the supplier. One of the GFSI s goals is certified once, accepted everywhere. The numbers of audits before and after becoming compliant with a GFSI benchmarked scheme were also evaluated by food safety risk group and size of company (Table 9). For all three levels of food safety risk, the number of audits was reduced. The mean number of audits dropped from 4.57 to 3.69 for the low-risk suppliers, from 4.42 to 3.97 for the medium-risk suppliers, and from 5.72 to 4.28 for the high-risk suppliers (Table 9). Depending on the size of the company, the number of audits also differed. For companies with annual sales of less than $50 million, the difference in the number of audits before and after being GFSI compliant was not significant (Table 9). For larger companies, the number of audits decreased after becoming GFSI certified. Once these companies customers become more familiar with the GFSI benchmarked schemes, the number of audits probably will decrease further. One company reported a reduction from 17 audits per year to only 2 after becoming GFSI compliant (15). Future directions. Initially, we hoped to be able to quantify the cost and estimate the dollar benefits from becoming certified based on a GFSI benchmarked scheme. However, in the initial questionnaire given to the beta testers we quickly found that most companies had not put in place cost accounting protocols to capture these costs or estimate their cost-benefit ratios. Therefore, we realized that a qualitative survey, not asking for actual dollar amounts of costs and savings, would be much more appropriate for this survey. A second, follow-up survey is planned after suppliers have been alerted to the kind of quantitative data needed. This survey will be used first with a representative beta test group and then with the national supplier base. This second survey will be more quantitative and will be used to document the actual costs and benefits of adopting the GFSI approach. In conclusion, this study provided significant insight into which GFSI benchmarked schemes were chosen by food suppliers and why, how many months were required to become compliant, the suppliers opinions of the relative costs of becoming compliant, and the leading indicators on which most suppliers could agree. The most common reason suppliers stated for implementing the GFSI system was to meet customers requirements to continue doing business. This reason also given most often for why the suppliers became GFSI certified. Thus, the decision to embrace one of the GFSI benchmarked certified schemes often was not voluntary, which indicates the powerful role collaborative, industry-led initiatives or food industry self-regulation can play in advancing food safety. Other important drivers for becoming GFSI compliant included enhancement of food safety and the need to remain competitive in the industry. In addition to complying with customers requirements, the GFSI benchmarked certification provided other benefits cited by the respondents. Most suppliers agreed that implementation of a GFSI benchmarked scheme enhanced the food safety of their products; thus, this reason is a positive leading indicator for adopting a GFSI benchmarked scheme. Suppliers also agreed that the implementation of a GFSI standard resulted in a better documented food safety management system, which included more through documentation of food safety risk assessments and prevention control plans and monitoring to ensure controls were properly implemented and maintained. These benefits appeared to be consistent with the requirement for food suppliers to implement preventative controls per the U.S. FSMA (10, 12). Most suppliers perceived GFSI standards as beneficial for improving the safety of their products. Most suppliers also agreed that changes to their food safety management systems were needed to meet the GFSI requirements. These changes were considered improvements to the safety of their food products. GFSI compliance further enhanced the ability of these companies to produce and supply safe food. Another outcome was an increase in employee food safety training. Regarding food safety audits, adoption of GFSI benchmarked schemes resulted in fewer audits (one less audit per year). Some suppliers may have interpreted the question more broadly to include all types of third-party audits, not just food safety. The results of this survey suggest that an important educational opportunity exists to acquaint retailers and suppliers worldwide with the savings and benefits associated with a single, internationally recognized food safety certification program such as GFSI. ACKNOWLEDGMENT Support from Walmart Stores is gratefully acknowledged. REFERENCES 1. Crandall, P. G., E. C. Friedly, M. Patton, C. A. O Bryan, A. Gurubaramurugeshan, S. Seideman, S. C. Ricke, and R. Rainey Estimating the demand for organic foods by consumers at farmers markets in northwest Arkansas. J. Agric. Food Info. 11: Global Food Safety Initiative Welcome to the Global Food Safety Initiative. Available at: Accessed 20 June Global Food Safety Initiative Guidance document, 6th ed. Available at: dancedocumentsixthedition.html. Accessed 20 March Global Food Safety Initiative Enhancing food safety through third party certification. Consumer goods forum. Available at: Accessed 20 March Global Food Safety Initiative Latest information on GFSI recognised schemes. Consumer goods forum. Available at: DutchHACCPFINAL.pdf. Accessed 1 March Instructional Assessment Resources Response rates: guidelines for maximizing response rates. Division of Instructional Innovation and Assessment, University of Texas Austin. Available at:

10 J. Food Prot., Vol. 75, No. 9 COMPANIES OPINIONS OF GLOBAL FOOD SAFETY INITIATIVE survey-response.php. Accessed 20 March International Food Standards Food safety certification: a necessary investment. White paper. Released 2 November Available at: ifs_whitepaper_final_9_29_10.pdf. Accessed 20 March Rainey, R., P. G. Crandall, C. A. O Bryan, S. C. Ricke, S. Pendleton, and S. Seideman Marketing locally produced organic foods in three metropolitan Arkansas farmers markets: consumer opinions and food safety concerns. J. Agric. Food Info. 12: Schmidt, R. H Regulatory requirements and HACCP. Available at: AND_HACCP.pdf Accessed 20 June Scott-Thomas, C Food safety law gets $39m funding for Available at: Accessed 20 March Suarez, J. P Wal-Mart becomes first nationwide U.S. grocer to adopt Global Food Safety Initiative standards. Available at: Accessed 20 March U.S. Food and Drug Administration Food Safety Modernization Act (FSMA). Available at: FSMA/ucm htm. Accessed 20 June Van Loo, E. J., V. Caputo, R. M. Nayga, J. Meullenet, P. G. Crandall, and S. C. Ricke The effect of organic poultry purchase frequency on consumer attitudes toward organic poultry meat. J. Food Sci. 75:S384 S Van Loo, E. J., S. C. Ricke, S. R. Milillo, S. Seideman, and P. G. Crandall Consumer food safety perceptions on ready-to-eat deli foods in northwest Arkansas. Food Prot. Trends 30: Wellik, R Global Food Safety Initiative improves organizational culture, efficiency in the food industry. Food Qual. Mag. Available at: Global_Food_Safety_Initiative_Improves_Organizational_Culture_ Efficiency_in_Food.html. Accessed 27 April 2012.

11 1670 CRANDALL ET AL. J. Food Prot., Vol. 75, No. 9 APPENDIX: QUESTIONNAIRE

12 J. Food Prot., Vol. 75, No. 9 COMPANIES OPINIONS OF GLOBAL FOOD SAFETY INITIATIVE 1671

13 1672 CRANDALL ET AL. J. Food Prot., Vol. 75, No. 9