Packaging as an O ﬄ ine Method to Share Information: Evidence from the Food and Beverage Industry in the Republic of Korea

: With the growing participation by diverse stakeholders in the total ﬂow of products, as seen with supply chains and logistics, it is becoming increasingly complicated to decide what information is to be shared and who is to be a partner. The purpose of this study is to explore the role of packaging as an o ﬄ ine method to share information in the total channel. This is because packaging improves communication among stakeholders and is normally part of the ﬁrst contact between them and the product. Thus, it has the strategic potential to share product information that meets stakeholders’ needs. To accomplish this objective, we built a research framework that depicts four hypotheses and tested it with structural equation modeling (SEM). Data were collected by surveys and measured for statistical analysis. After identifying the role of packaging, we showed nine speciﬁc related variables and the information’s perceived e ﬀ ects on stakeholders and their directions and relative values. This will help future researchers to discuss packaging’s extended roles, the needs of information separation, and its priority to be shared to help executives develop packaging strategies as an o ﬄ ine means to share information. Additionally, as packaging is considered to be an information generator, it gives participants the opportunity to extend its roles and to assign additional information to the product along the o ﬄ ine ﬂow of goods from manufacturers to end users.


Introduction
Information sharing is a strategy to integrate stakeholders in the total channel, including supply chains and logistics [1]. It is essential for sharing useful and effective information with partners [2]. With the growth in participation by diverse stakeholders, it is becoming more complicated to determine which information is to be shared and who to include as a partner [3]. However, both costs and risks stemming from increasing complexity are limitations of information-sharing strategies [4], and stakeholders' profits are occasionally not practical or fall short of their expectations [5,6]. Packaging is, meanwhile, an essential strategy [7], and its positive impacts are comprehensively provided through logistic processes in supply chains [8][9][10][11] and on the efficiency of manufacturing, distribution, and goods handling [12]. By contrast, offline communication is achieved with packaging, which improves communication among stakeholders [13,14] because it is normally part of the first contact between them and the product.
The purpose of this study is to explore the role of packaging as an offline method to share information in the total channel. To accomplish this objective, it is necessary to separate packaging from manufacturing and treat it as a generator of product information. For example, the handling of goods is product information, which is sharing within stakeholders rather than manufacturing information. This distinction in information enables diverse stakeholders to assign additional information to packaging along the offline flow of goods, ranging from manufacturers to end users. This helps us to answer the research questions: What is packaging's role in the sharing of information? What information is necessary? Who shares it with whom? And to what extent is shared information helpful and effective?
This research was comprised of four sections. For the first section, we built a research framework that depicts our hypotheses. For the second section, data were collected by surveys and measured for statistical analysis. For the third section, we tested the research framework with structural equation modeling (SEM) and described our results. In the last section, we conclude and discuss our findings, their implications, and the study's limitations and recommend further research.

Literature Review and Hypotheses Development
Information sharing produces more profit for stakeholders [15] and increases the cooperativity of their relationships [16]. Sharable information includes necessities such as tactical, strategic, and decisive information [17]; confidential ideas, plans, and processes [18]; and general data and knowledge [19][20][21]. Delivery information helps manufacturers reduce overages [22] and includes product quantities, delivery type, routes, shipment rates, transportation equipment-use plans, traceability, and loading capacity [23]. Sharing of scheduling and routing improves the efficiency of delivery, rapidly meeting customers' needs [24,25]. This sharing is more profitable to customers than to raw material suppliers [26,27] and improves delivery performance with positive communication among stakeholders [28]. Traceability is particularly essential information to be shared [29]. Manufacturing information affects forwarding agents who generate the delivery information's effect on customers and covers production planning, lead time, orders, demand data/forecasts, and procurement [30,31], as well as sales volume, inspection date, price, status, delay [32,33], daily orders [34], and made-to-order items [35]. Order sharing strengthens stakeholders' relations [36], demand forecast sharing produces practical benefits [37,38], and lead time sharing helps supply chain management to be more effective [39,40]. It is not profitable, however, to share lead time without customers' needs [41]. The sharing of manufacturing information is explored with sales and operations planning [42] and with a resource-based view [43]. Based on these considerations, we formulated the following hypotheses: H1: Delivery information has positive effects on customers who produce customer information.
H2: Manufacturing information has positive effects on forwarding agents who produce delivery information.
Packaging design is represented by an integrated management model [44], and logistics-driven packaging is identified and its strategic potential described [45] and explored for supply chain needs [46]. This makes for efficient transportation [47] and a sustainable supply chain regarding transformation aspects [48]. Its system is examined for logistical and environmental performance [49], and returnable packaging is helpful for reverse resource exchanges [50][51][52]. Information is shared online using information and communication technologies (ICT) [53][54][55][56]. As for manufacturers, both manufacturing and product information are produced independently. However, the two types of information play different roles within total channels. Packaging is not affected by manufacturing but, rather, they are separate from each other. Therefore, well-defined packaging information is critical to meet stakeholders' expectations and achieve efficient supply chains [57]. Its extent is examined in the secondary packaging [58] and affects consumer perceptions of labels and brand names [59][60][61]. The use of a food consumption database is collected differently by the contact consumer [62]. There is a difference between how males and females perceive information from healthy food packaging [63]. The choice of cosmetic creams, although they are not a food or beverage, is influenced by product composition, product information, and brand image [64]. Both the customers' demographic changes and the growth in demand for healthy food have resulted in smaller packs, information about product origin, and quality/ecofriendly certification for personal needs, influencing manufacturer behaviors. This trend illustrates the indispensable role of packaging [65] and its promotion of competitiveness [66]. Including product information on packaging communicates that information to customers [67], shares it with other stakeholders, and helps companies' logistical and supply chain needs by using ICT [68][69][70]. Based on these considerations, we formulated the following hypotheses: H3: Manufacturing information has positive effects on packaging that produces product information.
H4: Customer information has positive effects on packaging that produces product information.
The research framework was set up with the above hypotheses, which are based on results of the previous literatures, as shown in Figure 1. needs, influencing manufacturer behaviors. This trend illustrates the indispensable role of packaging [65] and its promotion of competitiveness [66]. Including product information on packaging communicates that information to customers [67], shares it with other stakeholders, and helps companies' logistical and supply chain needs by using ICT [68][69][70]. Based on these considerations, we formulated the following hypotheses: H3: Manufacturing information has positive effects on packaging that produces product information.
H4: Customer information has positive effects on packaging that produces product information.
The research framework was set up with the above hypotheses, which are based on results of the previous literatures, as shown in Figure 1.

Operational Definition of Variables
All 25 variables were divided into four latent types: manufacturing information, delivery information, customer information, and product information. The 25 observed variables were generated from manufacturers, forwarding agents, customers, and packaging. Operational definitions of all variables are shown in Table 1.

Ma1
Production planning Ma2 Lead time Ma3 Order processing Ma4 Rate of production Ma5 Defective products Ma6 Production cost

De1
Consolidation service De2 Delivery scheduling De3 Delivery type

Operational Definition of Variables
All 25 variables were divided into four latent types: manufacturing information, delivery information, customer information, and product information. The 25 observed variables were generated from manufacturers, forwarding agents, customers, and packaging. Operational definitions of all variables are shown in Table 1.

Latent Variables Observed Variables Definition
Manufacturing information

Ma1
Production planning Ma2 Lead time Ma3 Order processing Ma4 Rate of production Ma5 Defective products Ma6 Production cost

De1
Consolidation service De2 Delivery scheduling De3 Delivery type De4 Transport path De5 Traceability and confirmation of delivery

Latent Variables Observed Variables Definition
Customer information Average income Cu5 Consumption patterns Product information Packaging material Pr3 Manufacturer Pr4 Security indication Pr5 Quality certification Pr6 Handling of goods Pr7 Ecofriendly certification Pr8 Packing unit/size Pr9 Labeling and advertising

Sampling and Data Collection
To conduct this research, we designed a questionnaire (see Appendix A). Respondents were categorized into manufacturers, forwarding agents, customers, and packagers, each of which has a different role in producing information. Packagers' tasks are particularly separate from those of manufacturers because manufacturers are involved in generating information from Ma1 (production planning) to Ma6 (production cost), whereas packagers are involved from Pr1 (origin) to Pr9 (labeling and advertising). Characteristics of the respondents are shown in Table 2. We informed the questionnaire respondents of the objective of our survey on the first page. They were asked to rank the 25 variables, from De1 (consolidation service) to Cu5 (consumption patterns), on a five-point Likert-type scale ranging from 1 = "of no importance" to 5 = "of major importance." The response rate was 21.4% (240/1121).

Hoelter's Critical N
Hoelter's critical N index focuses on the adequacy of the sample size rather than the model fit [71,72]. An adequate sample size is shown in Table 3. A Hoelter 0.05 value indicates that a sample size exceeding 88 is adequate, and a Hoelter 0.01 value indicates that a sample size exceeding 104 is adequate. The sample size of 240 is adequate for this research.

Reliability and Validity
Each Cronbach's alpha of the variables is greater than 0.7, and each average variance extracted (AVE) is greater than 0.5, as shown in Table 5. Both reliability and validity are confirmed with the results. Nine observed variables were selected to explain the four latent variables of manufacturing information, delivery information, customer information, and product information. Regression weights were used: Ma1 was weighted by 0.1, De1 was weighted by 1.0, Cu1 was weighted by 0.2, and Pr6 was weighted by 0.7.
Manufacturing information is mostly explained by Ma1 (production planning) and Ma2 (lead time) of six observed variables. Delivery information is mostly explained by De1 (consolidation service) and De5 (traceability and confirmation of delivery) from five observed variables. Customer information is mostly explained by Cu1 (age) and Cu2 (gender) from five observed variables. Product information Sustainability 2019, 11, 6327 6 of 13 is mostly explained by Pr5 (quality certification), Pr6 (handling of goods), and Pr7 (ecofriendly certification) from nine observed variables.

Correlations
The latent variables have no correlations with each other. The result of correlation analysis is shown in Table 6.

Hypothesis Test
Delivery information from forwarding agents is helpful to customers [23,29]. Manufacturing information is helpful to forwarding agents [7,30,31,33] and is effective for packaging agents [9,13,14]. Packaging is positively affected by customer information such as age and gender [57,65]. Each hypothesis is accepted based on the results of the hypothesis test, as shown in Table 7. The fit indices were chosen using three principles [75][76][77][78][79][80]. First, the index cannot be easily changed by sample size. Second, the index can indicate the model parsimony. Third, the index cannot be easily changed with many kinds of estimations. The recommended indices depending on these principles are shown in Table 8. These indices fit the recommended threshold values. Hence, the model is acceptable. Although these indices are acceptable, the reliability and validity of the model have to be preferentially verified, as shown in Table 5.

Effects of Relationships
Standardized coefficients have the same correlations (value 1.0) in structural equation modeling. There is no effect when a coefficient is 0.0 and there is a strong effect on the causal relationship with a high coefficient. The standardized coefficient is used to examine its relative weight among variables. Total effects are shown in Figure 2 and Table 9.

Effects of Relationships
Standardized coefficients have the same correlations (value 1.0) in structural equation modeling. There is no effect when a coefficient is 0.0 and there is a strong effect on the causal relationship with a high coefficient. The standardized coefficient is used to examine its relative weight among variables. Total effects are shown in Figure 2 and Table 9. Delivery information has a direct effect (0.196) on manufacturers and no indirect effect (0.000). Customer information has an indirect effect (0.075) on manufacturers and a direct effect (0.381) on forwarding agents. Product information has both a direct effect (0.245) and an indirect effect (0.032) on manufacturers, and its total effect comes to 0.277, as well as an indirect effect (0.163) on forwarding agents and a direct effect (0.428) on customers. Based on the results, delivery information is effective for manufacturers. This appears to be based on the fact that delivery information represents observed variables like consolidation service, traceability, and confirmation of delivery. Customer information is more effective on forwarding agents than on manufacturers, giving forwarding agents the insight to consider customer information (age and gender) in primary delivery service factors. Product information is more effective on customers than on forwarding agents. Even if it is limited to Pr5, Pr6, and Pr7, this result provides deeper insight into manufacturers than forwarding agents [13,14] and customers [57,67]. This is because manufacturers play decisive roles in making product information, more strategic and more effective with by collaborating with packagers, including from Pr1 to Pr9.   Delivery information has a direct effect (0.196) on manufacturers and no indirect effect (0.000). Customer information has an indirect effect (0.075) on manufacturers and a direct effect (0.381) on forwarding agents. Product information has both a direct effect (0.245) and an indirect effect (0.032) on manufacturers, and its total effect comes to 0.277, as well as an indirect effect (0.163) on forwarding agents and a direct effect (0.428) on customers. Based on the results, delivery information is effective for manufacturers. This appears to be based on the fact that delivery information represents observed variables like consolidation service, traceability, and confirmation of delivery. Customer information is more effective on forwarding agents than on manufacturers, giving forwarding agents the insight to consider customer information (age and gender) in primary delivery service factors. Product information is more effective on customers than on forwarding agents. Even if it is limited to Pr5, Pr6, and Pr7, this result provides deeper insight into manufacturers than forwarding agents [13,14] and customers [57,67]. This is because manufacturers play decisive roles in making product information, more strategic and more effective with by collaborating with packagers, including from Pr1 to Pr9.

Discussion and Conclusions
This research aimed to explore the role of packaging as an offline method to share information within the total channel, including supply chains and logistics. The four-hypotheses research framework was tested with 240 sets of data collected from respondents in the food and beverage industry in the Republic of Korea.

Findings and Implications
Based on the observed role of packaging, we found nine specific variables and the information's perceived effects on stakeholders, their directions, and relative values. This helps future researchers discuss packaging's extended role, the need for information separation, its priority for sharing, and executives' development of a packaging strategy as an offline way to share information. Findings and implication are as follows.
What is packaging's role to share information? Packaging delivers product information and shares it with stakeholders in an offline total channel. When packaging is considered as a generator of product information, it gives diverse participants ways to extend its role and assign additional information. The distinction of information helps packaging play a role in assisting online information sharing as well as the offline flow of goods from manufacturers to end users. This improves stakeholders' competitive edge and answers the next questions directly or indirectly.
Which information is necessary? To answer this, we categorized information to be shared among stakeholders into four latent variables and explained them with 25 observed variables. Only nine observed variables were chosen and used to examine the research framework because the other 16 did not show normality. Manufacturing information was represented with production planning and lead time, delivery information with consolidation service and scheduling, customer information with age and gender, and product information with quality certification, handling of goods, and ecofriendly certification.
Who shares information with whom? To answer this, it was necessary to clarify the direction of the information flow within stakeholders. For this, we introduced packaging as an offline product information generator into the research framework. This allowed us to test relations between manufacturers and packaging, packaging and customers, and manufacturers and forwarding agents. Customer information is more effective for forwarding agents than for manufacturers, and delivery information from forwarding agents is helpful to manufacturers. This gives forwarding agents the insight to consider the customer as the primary service factor. Furthermore, packaging generates more effective information for customers than for forwarding agents. This gives a deeper insight into manufacturers rather than forwarding agents as they play decisive roles in making product information more strategic and more effective with the packers' collaboration.
Finally, to what extent is shared information helpful and effective? To answer this, the research framework was composed of four hypotheses to describe information effects on stakeholders and was tested using structural equation modeling. The effects are perceived by respondents and their values are relative. In terms of total effects, they are enumerated in rank: product information to customers (0.428), customer information to forwarding agents (0.381), product information to manufacturers (0.277), delivery information to manufacturers (0.196), product information to forwarding agents (0.163), and customer information to manufacturers (0.75).

Limitations and Further Research
The 25 variables were described as sharable information and assessed for normality in statistical analysis. Nine variables were chosen as the other 16 did not show normality. It is conceivable that they might be chosen for further research and, accordingly, the factor score weights of the chosen variables were added, as listed in Appendix B. Standardized coefficients of effects were used to compare them with each other. The comparison of estimated effects could not help but be limited to the use of standardized coefficients and to the scope of the food and beverage industry in the Republic of Korea. Non-standard coefficients of effects are listed in Appendix C. This enables future researchers to compare these results with their own future results. Both the information priority and its relative effectiveness, with additional comments, could depend on industry scope, country, and relationship levels among stakeholders.

De1
Consolidation service De2 Delivery scheduling De3 Delivery type De4 Transport path De5 Traceability and confirmation of delivery

Ma1
Production planning Ma2 Lead time Ma3 Order processing Ma4 Rate of production Ma5 Defective products Ma6 Production cost

Pr1
Origin Pr2 Packaging material Pr3 Manufacturer Pr4 Security indication Pr5 Quality certification Pr6 Handling of goods Pr7 Ecofriendly certification Pr8 Packing unit/size Pr9 Labeling and advertising Customer information Cu1 Age Cu2 Gender Cu3 Residence Cu4 Average income Cu5 Consumption patterns Appendix B