Calorie Labeling in Saudi Arabia

CHAPTER 1

1.0 Introduction

In 2018, the Saudi Food and Drug Authority (SDFA) implemented a policy that required all food facilities in the Kingdom to give details of ingredients and the amount of calorie content I all foods and beverages consumed within the country (Gillett, 2009). According to Murray (2018), this directive applied to all entities handling food, including restaurants, supermarkets, bakeries, shops, government agencies, universities and recreational facilities. Furthermore, there were several legal penalties put in place in case any relevant entity failed to adhere to the new regulation (Murray, 2018). Ideally, as Gillett (2009) observes, the main aim of this regulation was to ensure the highest levels of health, hygiene and well-being of the Saudi population by maintaining healthy food production, distribution and consumption within the Kingdom. The negative health effects of calories have made food labelling, particularly at the point of purchase, an item of great interest to both the government and public health practitioners. Because foods with high-calorie content are associated with adverse health outcomes such as obesity, food facilities have been targeted for various changes in food content. Notably, researchers such as Zarkin et al. (1993) and Centre for Science (2009) suggested that there is an importance in knowing the amount of calorie intake in food because it helps consumers have diets with balanced energy content. However, while consumer surveys have shown that consumer desire to know the amount of calorie content of food at the point of purchase (Technimic Inc, 2009), studies on the actual impact of the labelling indicate mixed findings (Elbel et al. 2009, Tandon et al. 201).

Researchers (e.g. Wansink & Chandon 2006, Lichtman et al. 2009) have speculated that one of the reasons for inconsistent effectiveness of amount of calorie content labelling is that people may not be aware of the value of calories, or they may not have an exact reference for the right amount of calorie intake. However, according to Platkin (2014), nutritionists have widely discussed exercise equivalents as a simpler way of informing the public about the value of calorie intake. Nonetheless, the consumption of fast foods in Saudi Arabia has been associated with higher intake of calorie and greater risk of various health complications such as obesity (Bowman & Vinyard 2004, Mancino et al. 2010). With an increasing number of consumers buying food in fast food restaurants and supermarkets (Liu et al. 2013), policymakers’ attention has been drawn to various regulatory approaches meant to influence consumer behaviour and choice including SDFA’s compulsory calories content labelling on all food products sold within Saudi Arabia. Calorie labelling as a public health measure against poor health outcomes such as obesity is prospectively effective. Roberto et al. (2009) argue that without such labelling, nutritionists find it hard to access and understand calorie content. Furthermore, Girz et al. (2011) insist that most consumers may not be aware that such nutritional information is always available with the purchasing department of food stores or on manufacturers’ websites, and of those who are knowledgeable, only a few percentages may take time to seek this information. But research by Bates et al. (2011) & Block et al. (2013) indicates that consumers tend to underestimate the number of calories in purchased fast foods and that as the total amount calorie in such foods increase, so does the level of underestimation of the number of calories. Moreover, research by Martinez et al. (2013) has found that consumers tend to need nutritional information and that they would want to use such information to make healthier food choices; hence calorie labelling emerges as a good source of such information primarily at the point of food purchase.

Nearly ten years ago, Harnack & French (2008) conducted literature review research to investigate the impact of calorie labels on consumer behaviour, specifically on consumer food choice. The study found a correlation between food choice and calorie labelling, although this correlation was only found to be weak or inconsistent. In that study, only one of the reviewed studies found a positive influence of calorie labelling on food choice, even though even in this study the effects were found to be either inconsistent or marginal in specific categories of food. Furthermore, at the time that the research was conducted, only six experimental studies were in existence because no government had implemented the calorie labelling policy. The authors, therefore, only reviewed a few menu interventions in university and workstation cafeterias as well as circumstantial experiments where the individuals had made hypothetical food choices. Ultimately, it was concluded that the six reviewed studies had various methodological shortcomings, and this called for further researcher conducted under natural settings. Some years after the study by Harnack & French (2008), several jurisdictions have implemented the mandatory calorie labelling policy for food outlets, supermarket and restaurants and several studies have been conducted on the impact of this policy on consumer behaviour. Consequently, four years after Harnack & French (2008) publication, Swartz et al. (2011) published an update of Harnack & French (2008). In the review study by Swartz et al. (2011), only the studies that had calorie labelling at the point of food selection or purchase and used experimental or natural researcher design were included. Hence, the results from studies by Swartz et al. (2011) were drawn from only the studies that drew evidence from non-hypothetical food selections. Ultimately, Swartz et al. (2011) concluded that calorie labelling was not an effective way of influencing consumers into buying food products and brands with low-calorie content.

Apart from the above-highlighted methodological shortcomings and lack of consensus of previous studies on the impact of calorie labelling on consumer choice and purchase behaviour, there seems to be a lack of qualitative empirical evidence on the same topic area. The main aim of the current study is to explore the impact of compulsory calorie labelling directive in the Republic of Saudi Arabia; issued by the SDFA on consumer purchase behaviour. The study will also aim to achieve the following objectives:

1.1 Objectives of Study

i. To explore the impact of mandatory calorie labelling on consumer purchase behaviour

ii. To investigate the impact of mandatory calorie labelling on consumer attitude about food

iii. To examine the diversity of consumer responses of compulsory calorie labelling

1.2 Research Questions

i. What are the impacts of mandatory calorie labelling on consumer purchase behaviour?

ii. What are the effects of compulsory calorie labelling on consumer attitude about food?

iii. How diverse are consumer responses to the compulsory calorie labelling directive in Saudi Arabia?

1.3 Research Hypothesis

H1: Calories displayed affect brand favourability

H0: Calories displayed does not affect brand favourability

1.4 Justification of the Study

The prevalence of obesity and overweight has been on an upward surge in the as few decades especially in countries within the Gulf Cooperation Council, i.e. Saudi Arabia, Qatar, Bahrain and Kuwait (Abdul et al., 2014). According to Ali et al. (2011), the discovery of oil in this region in the 1960s has contributed to significant economic growth leading to an increase in wealth per capita in these countries as well as population growth. But, according to Swinburn (2011), the characterising economic prosperity of this population has contributed to a variety of environmental factors (obesity drivers) that have contributed to an increase in the obesity epidemic. Some examples of obesity drivers include demographic transfers from rural to urban areas, enhanced technology characterised by increased motorisation, and the proliferation of processed food (Abdul et al., 2014). Another exacerbating factor is that countries in the GCC are having among the lowest rates of physical activities globally (Abdul et al. 2014), and populations in these countries have been increasingly adopting a more westernised food culture that is characterised by highly processed food (Montiero et al. 2013, Balbo 2012). The result has been an increased interest of governments towards various public health approaches and strategies aimed at addressing the issue of obesity and overweight among the populations. This interest has especially gained traction considering research evidence (e.g. World Health Organization 2000 & Dietz et al. 2009) indicating that population-based public health approaches and strategies can be useful in controlling or preventing the prevalence of obesity and overweight among a population. Consequently, governments have developed and implemented various obesity prevention strategies targeting the factors that contribute to obesity such as unhealthy lifestyle, environmental factors, as well as socio-economic factors.

Based on multiple public health frameworks, the Saudi Arabian government has implemented various public health strategies aimed at addressing the problem of obesity, with empirical evidence indicating mixed outcomes of these strategies. For instance, according to Alnaami (2016), the government of Saudi Arabia developed an overall strategic plan that took an inter-professional approach towards controlling and preventing obesity and overweight in the Kingdom. But since this strategic plan was implemented, no significant change in obesity prevalence has been reported. But from a more critical perspective, Kizsko (2014) argues that these policies affect the business of food and beverage production and sale, even though there is a lack of scientific evidence highlighting these effects. Hence, conducting this study will help fill the knowledge gap regarding how such policies affect the business operations of food vendors – in the context of consumer behaviour and choice. More importantly, the findings of this study will equip policymakers with more knowledge and information regarding the critical factors they need to consider when developing mandatory regulations, especially after understanding how these regulations affect the business community. The results of this study will also be a useful source of information to the entrepreneurs in the food and beverage production industry, as they may have more ability to predict customer behaviour and how to survive under the new policy.

Chapter 2

2.0 Literature Review

2.1 Conceptual Framework

2.1.1 The concept of Brand favourability

The conceptual framework of brand favourability can be understood from the consumers’ perspective. However, it emanates from another concept named: consumer-based brand equity, which is defined as the effect of a consumer’s brand knowledge on their response to that particular brand and the marketing activities aimed at promoting it (Graham et al, 2017). According to Pettigrew et al (2017), brands are considered to have either negative or positive customer-based brand equity depending on the consumer response to an element of the marketing mix – and the response may either be favourable or unfavourable. Negative brand favourability exists when the customers have negative customer-based brand equity while positive brand favourability occurs when the customers have positive customer-based brand equity (Gorski et al 2018).

2.1.2 The concept of Food labelling

Also termed a nutrition information panel nutrition fact labelling is a kind of labels embedded on food packs as part of public health regulatory policies aimed at providing general information to consumers about the nutritional content of the food product (Pettigrew et al, 2017). According to Watson et al (2014), these labels are for educational purposes and are based on specific dietary contents of the food product. In Saudi Arabia, cafes and restaurants are required to display the nutritional content food labels especially the calorie content – as a move by the Saudi Food and Drug Authority to address obesity as a public health issue by 2030.

2.2 The Historical Background of Food Labelling

The relationship between consumer purchase behaviour and nutritional labels dates back in history and has a variety of theoretical perspectives. For instance, the now-famous product characteristic theory was published more than 40 years ago by Lancaster (1966). The theory suggests that consumers derive more utility from transformed goods compared to non-transformed ones, although some scholars observed that the implementation of the Lancaster Model could still not be achieved straightforwardly due o the difficulties in measuring the attributes of the goods (Silberberg & Suen, 2001). However, the Lancaster model underwent a transformation when marketers had begun to consider various products such as foods not only as consumption goods but also as a bundle of products (Lazardis & Drichouts, 2005). While Nelson (1974) distinguished between the experience attributes and searched attributes of goods, Darby & Karni (1973) developed the concept of credence attribute, which could not be evaluated during the product use of after the product consumption. However, according to Caswell & Mojduszka (1996), placing a nutritional label on the food could help turn the credence attribute of food nutrients into search attributes. Consequently, the regulatory bodies in most countries (e.g., the USA) begun to advocate for the display of nutritional food information as a way of transforming the credence attribute of food nutrition into a search attribute. The increasing trend of nutrition-related health problems globally triggered the consideration of nutritional labeling as a vital part of food purchasing decision by both scientific and non-scientific literature. Earlier on, most empirical application of food lableing was based on Stigler’s approach (Stigler, 1961), even though researchers (e.g., Drichouts et al. 2006) attempted to develop other theoretical frameworks later on.

2.3 Factors Determining the Use of Food Labels

Earlier empirical research has also explored the factors determining consumers’ use of food labels. Primarily, these studies focused on identifying the characteristics of consumers who were likely to use food labelling when making purchase decisions based on Stigler’s cost-benefit approach. Ideally, Stigler (1961) approach held that consumers will always search for nutrition-related information as long as the cost of searching for such information (i.e., the time spent reading the labels) is lower than the benefits of reading such labels (i.e., making healthy food choices). However, most of the studies evaluating consumer’s use of nutritional labels have focused on the general determinants of these labels, while a few (e.g., Bender & Derby, 1992) have focused on the difference between nutrition panels and ingredient list, or evaluated the use of specific nutrition information (e.g., Drichoutis et al, 2005). Furthermore, only one study (Nayga, 1999) has focused on the determinants of beliefs or perceptions of food label usage. Based on the studies by Drichoutis et al. (2005), Nayga (1999) and Nayga (2005), the determinants of nutritional label use can be classified into five major categories, namely involvement of product class, individual characteristics, motivational factors, attitudinal, behavioural and attitudinal factors, and behavioural factors.

2.3.1 Individual Attributes

It is well known that the information search behaviour of an individual is much affected by their unique characteristics. With this regard, age is a common characteristic that has been found to have a significant effect on how an individual uses nutritional labels. For example, earlier research by Burton & Andrew shows that older people consider the labels as more challenging to understand. This finding may explain the results by Bender & Derby (1992) that older people tend to be only interested in the nutritional labels or just the ingredient list. These findings corroborate with the findings of Cole & Balasubramanian (1993) and Kim et al. (2001) that the probability of the use of nutritional labels decreases with an increase in age. However, studies by Coulson (2000) found contrary results.

2.3.2 Attitudinal Behavioural and Situational Factors

With regards to attitudinal behavioural and situational factors, time pressure has been found as a significant determinant of nutritional label use (Feick et al. 1986, Park et al. 1989, and Beatty & Smith 1987). In most of the studies, time pressure as a determinant of information use behaviour is influenced by income and working status. However, other studies have found contradictory results. For instance, Nayga (1999) found that consumers with higher incomes were more likely to agree with the importance of nutritional information, the ease of making food choices based on nutritional information, the importance of relying on the nutritional information rather than one’s knowledge, and that nutritional food labels can be a motivation to try new food products.

2.3.3 Product Involvement Factors

Earlier researchers (e.g., Thayer 1997, Rose 1994) hypothesized that the use of nutritional labels by consumers is affected by the importance placed by consumers on specific food attributes; because the attributes have a significant influence on the food purchase decisions. Furthermore, according to Drichoutis et al. (2005), consumers considering prices as relevant information for food purchase decisions are less likely to use nutritional information as a factor to consider while making such decisions. This can be explained by the fact that consumers who make purchase decisions based on food price are more likely to search for price information as opposed to nutritional details placed on the food labels due to lack of time for multitasking (Nayga et al. 1998; Nayga 2000). Conversely, it is difficult to establish how the importance of taste affects nutritional label use because studies (e.g., Nayga et al. 1998; Drichoutis et al. 2005) have found contradicting results.

2.3.4 Nutritional Knowledge

The influence of dietary knowledge on the use of food label information occurs because consumers with nutritional knowledge efficiently use such information, thereby decreasing the cost of using the labels (i.e., as measured by the time taken to study the labels). Specifically, earlier studies by Bender & Derby (1992) found a relationship between professional nutritional knowledge and the use of consumption of specific food nutrients. Furthermore, according to Moorman & Matulich (1993), consumers who are knowledgeable about nutrition and health are likely to acquire more information from various forms of media sources, including food labels. Other researchers such as Kim et al. (2001) and Szykman et al. (1997) have also found a relationship between nutritional knowledge and label use, even though results by Nayga (2000) found contrary evidence. Nayga’s (2000) findings are supported by the findings of Moorman (1998) that more knowledgeable consumers are less concerned with nutritional information. However, whereas one would ordinarily expect that consumers’ nutritional knowledge significantly influences dietary label use; there is a possibility that the use of food labels can affect consumers’ nutritional knowledge. For instance, reading more nutritional labels can increase the consumer’s nutritional awareness of the products they purchase. With this regard, studies by Drichoutis et al. (2005) indicated that the general use of food labels, as well as the use of ingredients information, enables consumers to acquire more nutritional knowledge.

2.3.5 Motivation

Researchers have proved that motivation to process information is one of the factors influencing the use of nutritional labels to make purchase decisions. For instance, while investigating the influence of enduring motivation on the use of nutritional information, Mooran (1990) found that enduring motivation influenced the consumers’ ability to process information, and this relationship was a strong one. These results corroborate with the findings of Keller et al. (1997) that enduring motivation moderately affects consumer’s perception of product nutritional value. However, the conceptualization and measurement of motivation is an essential factor to consider when evaluating the influence of motivation on the use of nutritional labels. For instance, Moorman (1990) defined motivation as the general use of dietary labels by consumers while shopping. Besides, motivation is depicted as one’s interest in reading health-related information. It is also important to note that Drichoutis et al. ’s (2005) label use questions to participants were similar to those used by Moorman (1990), and therefore cautiousness is needed while comparing these the results of these studies.

2.4 Existing evidence

Researchers interested in consumer psychology have attempted to explore the influence of food labelling on consumer behaviour even though not all of them have had a focus on calorie labelling. However, in this chapter, we comb through existing research to evaluate what is already known about the influence of food labelling on consumer choice. While some sources will act as a background sources to provide general information on this topic area, other will be used exhibit sources to confirm or refute the existence of such a relationship. Furthermore, while most of the reviewed resources were not specific on RAS, the information they provide is useful in developing a background understanding of the relationship between food labelling and consumer behaviour. In a recent study by Gorski et al (2018), the authors intended to explore the impact of Five Font-of-Pack Nutrition Labels policy recommended by The US National Academy of Medicine on consumer purchase behaviour and perceptions. The study involved 1247 adults who randomly participated in online interview surveys. The participants were asked to respond to six different conditions namely: NuVal, Facts Up Front, multiple traffic lights, single traffic light, no label, or a rank of 0-3. The No-Front-of-Package labels acted as a control label, while the single traffic label indicated the amount of calories per serving – with the traffic lights (i.e. green yellow or red) indicate the product’s nutritional quality. Multiple traffic lights indicated the among of calories in every serving label and the traffic light symbols indicated the levels (i.e. low, medium or high) levels of nutrients such as fats, sodium, sugars. Conversely, Facts Up Front were the manufacturers’ labelling as per the food industry regulations, which indicated the content levels of sugars, calories, and sodium per unit serving. Furthermore, NuVal lebel was developed by the authors and indicated the health suitability of the products in a scale of 1 to 100. Lastly, the 3-star labelling indicated the amount of calorie content in a scale of zero to three, as well as the amounts of added sugar, fats and sodium. The researchers evaluated the participants on the accuracy of their label interpretation as well as their purchase intentions. There was a similarity (p = 0.845) hypothetical nutritional quality of all shopping baskets, and the results showed a relationship between nutritional labelling and the ability of customers to identify the nutritional quality of food, even though the quality of label designs determined the outcomes. Furthermore, the results indicated that multiple traffic light labels and NuVals enable consumers have a greater (p < 0.001) ability to identify the healthier foods. Nonetheless, the researchers found that single traffic labels were the most influential (p < 0.03) compared to other labels especially in situations where consumers compared the nutrient content of similar products. With regards to calorie serving estimations, both single and multiple labels were the most accurate (p < 0.001) in making such estimations. Whereas front-of-package labels enable the participants to accurately estimate the product’s nutritional content compared to no label, none of the conditions changed the consumers’ purchase intentions. Hence, Gorski et al (2018) concluded that while no type of label was superior to the other in influencing consumer purchase choice, three-star labelling emerged to be the best in providing consumer education. This study, therefore, acts as both a background source and exhibit source to show evidence of an influence of food labelling on food choices. There are few studies that have used qualitative methods to evaluate the relationship between food labelling and consumer behaviour. To fill this research gap, Mhurchu et al (2017) conducted a randomised controlled trial study to explore three different types of food labelling (health rating labels, traffic light rating labels, and nutrition information labels) affected consumer purchase behaviour. The study relied on smartphone technology used by participants to receive product labels on their screens while recording their food purchases. Therefore, the barcode technology enabled the participants to actively participate in collecting objective data on food purchase and selection behaviour. The researchers then conducted a post-study analysis to evaluate the frequency of label viewing, the use of labelling by the producing company, and the relationship between the food label and the healthiness of the purchased product. A total of 1255 participants viewed the food labels, and a total of 66,915 barcoded items were purchased. The participants viewed 23% of all the purchased products, although this frequency decreased overtime. The participants were less likely to view labels of honey, eggs, sugar, vegetables, meat and fruits. Nonetheless, all the products whose labels were viewed and subsequently purchased by the purchased were had significant health qualities that those whose labels were viewed but not purchased. The nutrient profile of these products had a mean difference of −0.90 (95% CI -1.54 to −0.26). Ultimately, the authors concluded that there is a significant association between food labeling and consumer purchase behavior – in terms of the healthiness of food products purchase. Therefore, food labeling plays an important role in influencing consumer purchase behavior.

In another recent study by Graham et al (2017) the authors intended to explore the influence on explained or unexplained nutrition labels on the food choices of parents and children. Hence, the randomized control trial intended to investigate whether parents and children could select healthier foods when labeled with front-of-package labels and multiple traffic lights coded labels. The results indicated no relationship between multiple traffic labels and consumer’s purchase choice of healthy foods. There is an increasing trend of front-of-pack labels as an effective mechanism for enhancing consumer choice of healthy diets, and subsequently improving the health and well-being of the population. A study by Pettigrew et al (2017) explored the consumer preferences for various types and characteristics of front-of pack labels to provide nutritional information on food packets – in Australia. Whereas a majority of existing studies focus on two main types of front-of-pack labels in Australia namely those expressing the quantity of specific nutrients taken on a daily basis, as well as those that indicate nutrient content using the color system, the study by Pettigrew et al’s (2017) study focused on the health star rating system, consumers’ nominee of their most preferred labeling criterion, sampling consumers to lower socio-economic status, and children with a right of purchasing and consuming foods on their own or those who influence their parents to make food purchases. The researchers then conducted an online survey among the Australian participants to evaluate their most preferred labeling criterion between the front-of-pack label, traffic label and healthy star rating label. The study results revealed health-star rating label as the most preferred (40%) across all the participant demographics (i.e. children, upper or lower socio-economic status, males or female). Conversely, the daily intake label was the least preferred i.e. 20%. Ease of interpretation, salience and ease of use were among the reasons provided by the participants to justify their preference. Ideally, this study provides some evidence that a population-based nutrition labeling is an important tool used by consumers in making their food purchase choices. In another study by Cecchini & Warin et al (2016), the researchers intended to evaluate the influence of food labeling on consumer food choices and feeding behavior. The study’s main aim was to investigate the influence of food labeling on consumers’ selection of healthier foods, especially with regards to calorie content. Moreover, the study’s interest was inclined towards investigating the effectiveness of various food labeling criteria such as the traffic light criterion and daily amount guideline. The review pooled together randomized control trials that homogenously reported their outcomes. Ultimately, the study results found a positive correlation between food labeling and consumers’ choice of food, with food labeling found to increase consumer choice of healthier food by 17.95%. Moreover, the study results indicated that food labeling would increase the consumers’ calorie choice by 3.59% although these results had no statistical significance. Other findings by Cecchini & Warin et al (2016) are that traffic ligh labeling criterion was more effective in influencing consumer’s choice of healthier foods followed by other labeling criteria. However, the researchers could not conduct a further analysis of various labeling criterion and their single-handed effects on calorie choice. Nonetheless, the results from this study indicate that food labeling is an effective was of influencing consumer choice – with regards to healthier foods. Moreover, more interpretive labels have emerged to have a greater influence on the consumer choice. In a study by Olstad et al (2015) the researchers intended to evaluate the effectiveness of traffic light food labeling scheme on improving consumer purchase in sport facility and recreational environments. Adding to the already existing evidence that traffic light labeling has a positive effect on consumer’s purchasing behavior (i.e. selection of food) within such settings, Olstad et al (2015) examined the influence of menu traffic light food labeling on the consumers’ ability to select healthier foods within sports and recreational environments. The study implemented traffic light labeling scheme on food menus for a period of one week, with a one-week pre and post- assessment. A total of 322 participants completed the one week baseline assessment, while 313 participants participated in the intervention. To evaluate the effect, the researchers measured the change in the volumes of participants’ purchases of products labeled with yellow, green and red lights by comparing these volumes at baseline and after the intervention. Moreover, the change assessment also examined the association between the change and the demographic characteristics of the participants. The results indicated a 3.3% increase in purchase of foods with green label, with a statistical significance of p < 0.05, while the purchase of foods with red label decreased by 3.2% with a statistical significance of p < 0.05 too. However, the study did not find any difference in the effectiveness of traffic light labeling triggered by any demographic characteristic. Moreover, the demographic characteristics did not have any influence on the average daily revenues. Ultimately, the authors concluded that the traffic light labels increased the consumer’s purchase of healthy foods, reduced the purchase of unhealthy foods in recreational and sports environment.

More like a perfect exhibit source, the study by Maubach et al (2014) intended to explore the impact of various food labeling schemes such as multiple traffic labeling, star rating and daily intake guide labeling on consumer purchase behavior as well as their product perceptions. The study design also covered the health claims and nutrient contents of the investigated food product. Furthermore, the authors were interested in evaluating whether the participants used more or less information while making the food choices. The study results indicated that whereas the food choices were generally similar among with regards to the star labels and multiple traffic light labels, the multiple had a significant influence on participants’ preference because the foods selected with regards to this label were generally less healthy. On the other hand, the participants were more inclined to choosing less nutritional foods that had health claims, although the effect of health claims on consumer choice as less pronounced in the case of multiple traffic light labels. Nonetheless, the multiple traffic labels had more influence on consumer choice compared to the stars system. Generally, the results of this study imply the existence of evidence that different schemes of food labeling have different influence on consumer purchase behavior. Existing research evidence has shown that understandable and easily accessible food nutrient information displayed on food packs has an influence on shoppers’ ability to make food choices, which are mostly healthier choices. The study by Watson et al (2014) intended to investigate the shopper’s dependability on various front-pack labels to identify and make buying choices on healthier foods. The online questionnaire study aimed to examine a range of seven food labeling schemes including a variety of traffic light labeling, daily intake labeling, star labeling applied on nine different food products mostly purchased locally. The results indicated that in over 80% of the shopping time, participants could choose to buy healthier foods using any of the seven food labeling schemes. Moreover, no labeling scheme was found to be more superior to the other in influencing participants’ purchase choice. Nonetheless, schemes that only provided energy content information, as well as those with less numerical labels were relatively less influential to consumer choice.

Chapter 3

3.0 Research Methodology

3.1 Research paradigm

From an epistemological point of view, the research paradigm is concerned with the type of knowledge acquired within a specific subject area (Bryman et al., 2015). Moreover, according to Collis & Hussey (2013), it also entails the nature of assumptions made by the researcher in the process of the study, and can either be a positivism paradigm, realism paradigm or interpretivism paradigm. Coolican (2017) defines the positivist paradigm as the consideration of scientific models as the best approaches to evaluate a social reality. It is associated with deductive reasoning where theories and hypotheses are developed and tested, respectively (Craig & Douglas, 2005). Moreover, according to Forrester (2010), positivism is also associated with inductive reasoning, which entails the achievement of knowledge through data collection. Secondly, a researcher can take realism paradigm, which also considers scientific inquiries but uses what people experience through their senses to understand a phenomenon (Gill et al., 2010). Lastly, according to Krathwol et al. (2005), the interpretivism research paradigm refutes the use of scientific models to understand a phenomenon, but instead propose that researchers should use the uniqueness of human behaviour to understand the social world. The current study intended to evaluate the influence of food calorie labelling on consumer purchase behaviour, and therefore its primary interest was centralized on human behaviour. So, the researcher selected the interpretivism research philosophy as the most appropriate one.

3.2 Research Approach

Roberts (2010) states that two main research approaches can be taken by a researcher, namely: inductive approach and deductive approach. In deductive approaches, the researcher begins by developing theories and hypotheses based on the collected data, then going ahead to confirm or reject these hypotheses or revise the theory (Saunders et al., 2009). On the other hand, Krathwol et al. (2005) write that inductive research approach entails the use of already available data to develop a particular theory – and therefore researchers who use the deductive approach will make an observation, find out any form of behaviour from that observation, and derive conclusions from them. In the current study, the researcher partly followed the inductive approach because data was collected through survey questionnaires, while statistical methods were used to confirm or refute the hypotheses. Hence, the binary research approach enabled the researcher to have a comprehensive data analysis that could allow the derivation of valid conclusions (Flick, 2008).

3.3 Research method

According to Coolican (2017), there are two primary research methods that can be adopted by a researcher, namely: qualitative and quantitative research method. Quantitative research method involves answering research questions by evaluating the relationship between two or more variables (Miles et al., 1994). Moreover, the relationship between these variables can be measured using a variety of statistical methods, although it may be appropriate for studies that have clearly defined research hypotheses sand theories to be tested. On the flipside, qualitative research methodology enables the research to evaluate subjective human behaviour using non-statistical methods of data analysis (Moisander & Valtonen, 2006). On that note, the current study aims to explore the influence of calorie food labelling on subjective consumer choice. According to Collis & Hussey (2013), this can be associated with a naturalistic inquiry, which entails the examination of complex human experience. Therefore, the current study adopted a mix of qualitative and quantitative research methods. Qualitative research methods were exemplified in the process of data collection, wherein survey questionnaires were used to gauge the researcher’s opinions and behaviours. On the other hand, quantitative research methods were used in the data analysis section, where statistical tools were used to analyse and compare consumer purchase choices. Several theoretical underpinnings justify the application of mixed methods in the current study. For instance, according to Forrester (2010), qualitative methods enable a holistic exploration of subjective pathways that would allow for the development of theories. On the other hand, Collis & Hussey (2013) observe that quantitative methods are underpinned on scientific methods and are therefore suitable behavioural and social sciences by enabling an understanding of unique human nature. Therefore to draw on the benefits of both qualitative and quantitative methods, the researcher applied a mixed methodology.

3.4 Data Collection

The study targets 5 participants per restaurant, each restaurant being visited ones. Three person’s data collection teams will be assigned to each restaurant to maintain a count of restaurant entrants for purposes of calculating the response rate. The data collection teams will approach five customers (18-year-old) from each restaurant and ask them to provide their purchase receipts and later participate in a brief online survey administered through Qualtrics (an online survey platform); a 100 Saudi Riyals food purchase voucher will be awarded to the participants as an incentive after completing the survey. The data collection teams will also hand a card with a link to the survey website for participants to log in and participate at their convenience. Attached in Appendix 1 is a sample of the brief survey questionnaire. The survey was conducted in the English language, and no personal identifier was collected from the participants.

3.5 Data Analysis

The researcher will then enter all the items listed in the receipts in a database. Moreover, the number of calories will be assigned to each item in the receipts using the restaurant’s published calorie information. However, the number of calories will be adjusted to the participant’s reported extras (e.g., salads and vinaigrette) – if the extras’ calorie information is available. Alternatively, all the extras whose calorie information is not available will be assigned the lowest calorie content in that category. The researcher will then calculate the total number of calories per participant by adding the calories across all items indicated in their respective receipts. A mean calorie per purchase will then be calculated for each restaurant. A benchmark of 1000 calories diet will be used because it represents half of the standard reference (i.e., 2000), and all the purchases will be categorized 250 increments of calories to determine the total distribution. Ultimately, the researcher will use SPSS version 15.0 to conduct a statistical analysis of the collected data. Furthermore, the researcher will use t-test (α < 0.05) to identify the mean differences between the calories, while the χ2 test will be used to determine the p values.

3.6 Sampling Method

The Kingdom’s Ministry of Health maintains a register of all licensed food restaurants in RSA. The researcher included licensed food service restaurants that had implemented the mandatory calorie labelling on menus by January 2019 – information that was either available on the internet or in the restaurant’s websites. Conversely, the researcher excluded all ice-cream parlours in the register because the main focus of the study was calorie patterns on food and beverage purchases. After excluding ice-cream parlours, three restaurants remained and were eligible for inclusion of into the study. The three restaurants were conveniently sampled from a list of over 100 suitable samples within Riyadh’s Central Business District. This sample included three of the most popular fast-food restaurants in Riyadh, namely: Pastalita-Sahafa, Pinoy Fastfood, and Hardee's Fast food. Because of the easier accessibility of these food chains within Riyadh’s CBD, they were selected as the most convenient locations for data collection.

3.7 Ethical consideration

Because the study involved human subjects, ethical considerations will be deemed vital for maintaining the integrity of the survey. First, the participant’s integrity will be ensured by respecting the anonymity and confidentiality of the participants. Besides, each participant will be allowed to willingly provide their response to the questionnaires to ensure that the study is independent and impartial. The study will also ensure that no personal (e.g., email address, phone number, or full names) information of the participants is sought from them. Furthermore, as recommended by Bryman et al. (2015), the participants were required to sign informed consent (Appendix 2) – and they could withdraw their participation at will. Last but not least, the researcher will seek ethical approval from the ethics committee.

Chapter 4

4.0 Results

No sampled site was excluded from the study; all the three sites were located in Riyadh’s CDB. Besides, none was closed on the day of the study, all shared their names but concealed their affiliations. Each restaurant had a cooperative manager and all of them produced valid receipts. From the three sites, a total of 15 handed their receipts while 14 participated in the online survey. Therefore, one receipt was excluded from the study while the other 14 were considered valid for analysis as all of them indicated valid caloric values. We attribute the equal rate of participation in all the restaurants to the fact that they were all located within the same area (CBD). The average amount of calories purchased by the participants was 827 calories, while an average of 1000 calories or above was purchased by 34% of the participants. Besides, as illustrated in table 1, 15% of the participants purchased an average of 1250 calories. Participants who purchased chicken meals had the highest amount of calorie intake, while those who purchased sandwich meals had the lowest amount of calorie purchase.

4.1 Observed calorie information

93% (14) of the respondents participated in the survey questions evaluating the amount of calories purchased in the food. Apart from participants who purchased from Pastalita-Sahafa, 4% of the participants agreed to have seen the calorie information as currently provided. Interestingly, participants from Pastalita-Sahafa were significantly more likely to report seeing the menu calorie information than participants from the other two restaurants i.e. 34% compared to 4%, with a P value of < .001. Among the participants from Pastalita-Sahafa, those who agreed to have seen calorie information on the menu purchased food with 52 fewer caloric intakes than those who reported not seeing calorie information on the food menus. The latter group purchased an average of 714 calories compared to the former group who purchased an average of 766 calories, translating to a significant difference of P < .01. Furthermore, only 17% of those who saw calorie content purchased higher calorie content (calories ≥ 1000) food compared to 23% of those who never saw the calorie information (P < .01). Likewise, among the participants from Pastalita-Sahafa who saw the calorie information, 37% agreed that the information influenced their purchase choices. Nonetheless, participants who reported seeing calorie information on the menus and using such information to make purchase choices purchased food with an average of 647 calories compared to those who did not use that information to make food purchase choices who bought food with an average of 746 calories. Ideally, this means that the latter group bought 99 less calories compared to the former group, with a difference of (P < .001). The study also found that those who saw the calorie information and used such information to make purchase decisions were less likely to purchase food with more than 1250 calories. Interestingly though, there was no statistically significant calorie intake between participants who saw calorie information and did not use it to make purchase decision and those who did not see the calorie information. The latter group purchased an average of 746 calories compared to the former group who purchased an average of 766 calories, translating to a P value difference of .29.

Chapter 5

5.0 Discussion

Despite the availability of food calorie information on the menus, a relatively low number of participants reported to have seen the calorie information in their respective menus. This result is consistent with the findings of two other studies (i.e. Burton et al 2006 & Wootan & Osbom 2006). A closer observation revealed that the two of the three restaurants placed limited food calorie information on the deli foods near the registers and this was associated with a higher number of participants seeing the information, although the association was not statistically significant. More than one third of the participants reported that the calorie information affected their purchase choices. When we objectively analysed and measured the calorie content indicated on the receipts, it emerged that the participants who saw and used the calorie information to make purchase decisions purchased food with fewer calories compared to those who did not see the calorie information or saw the calorie information but did not use such information to make their purchase decision. However, the healthy role of providing calorie information is exemplified in the findings that participants bought food with high energy content, above 30% bought food with an average of 1000 calories in one meal. This finding generally confirms the earlier hypothesis that displayed caloric content affects consumer choice behaviour. Restaurants in the city of Riyadh and those in several other cities in RSA are required to indicate calorie information on their menus (Burton et al, 2006). In the current study, which involved restaurants with calorie content labels on their menus, some participants (5%) did not see the calorie information on the menus, especially when the information were not displayed on prominent formats such as counter charts, posters, distant walls, websites or counter mats. But a higher percentage of those who saw the calorie information were experienced in one of the restaurants which displayed the calorie information next to the point of purchase. A possible implication of this finding is that a prominent display of calorie information may contribute to more number of consumers seeing such information and using that information to make purchase decisions. However the study findings regarding the association between consumer purchase behaviour in the three restaurants and calorie information displayed on the food menus is subject to two main limitations, first the selected participants might not be a representative of all restaurant customers in RSA. there was no uniformity in the amount of calories purchased in all the three restaurants (the analysis only worked with averages), and some participants purchased fewer calories in one restaurant compared to the other; this could imply that food in that restaurant had lower calories compared to the other, or that consumers in that restaurant are more likely to purchase fewer calories compare to consumers in the other restaurants. Nonetheless, even if the analysis was done on one restaurant of the three restaurants, the overall results indicate that participants who saw calorie information would still purchase food with fewer calories. Similarly, the popularity of these restaurants in Riyadh is significant; therefore if they were to make the calorie information visible at the point of purchase, participants in all the three restaurants would identify and use the information to make their purchase decisions. Secondly, it could be possible that the participants who reported to have seen the calorie information could have been interested in having such information due to weight concerns, and were more interested in looking for such information compared to those who did not see the calorie information. However, it is important to note that the participants who saw but did not use such information to make purchase decision, as well as those who did not see such information purchased relatively similar amounts of calories – implying that they had comparable purchase patterns.

Based on the prevalence of fast food consumption in Riyadh, even a small reduction in food calorie content could contribute to a significant reduction in the general population level calorie consumption (Hill et al 2003, Veerman et al 2007). However, a large population of fast food consumers in Riyadh lack the information needed to make healthy food choices at point of purchase. Nonetheless, it is surprising that the mandatory food calorie information on restaurant menus did not receive much opposition compared to the case of New York where a similar directive attracted a series of litigations (Basset et al, 2008). Finally, the calorie content in fast foods is high are unhealthy. Whereas most fast food restaurants in RSA claim to be publishing the nutritional information of all their food and beverage products, some of the methods used in publishing such information may not be as effective as expected. With this regard, this study has established that placing the calorie information at the point of purchase is more effective and may contribute to lower consumption of caloric food by consumers who reported to have seen the information. Consequently, this study recommends that public health authorities in RSA should develop more policies aimed at making calorie information displayed to consumers in more prominent ways (e.g. at the point of purchase) to allow for more consumer information and reduce obesity-related illnesses by reducing consumer caloric intake.

Chapter 6

6.0 Conclusion and Recommendations

The findings of this study indicate that all participants who saw calorie labels, regardless of using the information or not, made generally healthier food choices with regards to the food brand. A majority of the participants who saw the calorie labels used the information to make choices on the brand of food to purchase. Despite these findings, the current study recommends future research to further explore how consumers use calorie labelling not only in restaurant settings but also on supermarket s and other types of food stores. Moreover, future research should also evaluate factors that enhance the consumer’s usage of the calorie information, as well as its effects on the consumer’s subconscious mind. Nonetheless, we observe that legislations on mandatory food labelling should be conscious of necessary environmental changes, the need for educational campaigns and social marketing as well as the need for various behavioural strategies that would make them more effective. However, nutritionists have the role to help consumers make good food choices by recommending the usage of healthy caloric intake. Nutritionists and public health practitioners can also play a role in promoting healthier caloric intake by encouraging hoteliers to avail food with wider range of food nutrients that consumers can choose from.

References

Bryman, Alan and Emma Bell (2015) Business Research Methods, Oxford University Press.

Burton, S., Creyer, E., Kees, J., Huggins, K. (2006) Attacking the obesity epidemic: the potential health benefits of providing nutrition information in restaurants. Am J Pub Health;96(9):1669–1675.

Collis, J., & Hussey, R. (2013) Business Research: A Practical Guide for Undergraduate

Coolican, H. (2017) Research Methods and Statistics in Psychology, Psychology Press. Craig, C. S. and Douglas, S. P. (2005) International Marketing Research, John Wiley & Sons.

Cecchini, M., Warin., L. (2016) Impact of Food Labelling Systems on Food Choices and Eating Behaviors: A Systematic Review and Meta-Analysis of Randomized Studies: Impact of Food Labelling, Obesity Reviews, 2016. DOI: 10.1111/obr.12364.

Ducrot, M., Pauline et al. (2016) Impact of Different Front-of-Pack Nutrition Labels on Consumer Purchasing Intentions , American Journal of Preventive Medicine, 2016. DOI: 10.1016/j.amepre.2015.10.020.

Gorski, F., Mary T. et al. (2018) Comparing Five Front-of-Pack Nutrition Labels’ Influence on Consumers’ Perceptions and Purchase Intentions, Preventive Medicine, 2018. DOI: 10.1016/j.ypmed.2017.10.022.

Graham, P., Dan J., et al. (2017) Impact of Explained v. Unexplained Front-of-Package Nutrition Labels on Parent and Child Food Choices: A Randomized Trial, Public Health Nutrition, 2017. DOI: 10.1017/S1368980016002676.

Thorndike, A., Riis, J., Sonnenberg, L., Levy, D. (2014) Traffic-Light Labels and Choice Architecture, American Journal of Preventive Medicine, 2014. DOI: 10.1016/j.amepre.2013.10.002.

Gorski, T., Werth, M., Musicus, A., Bragg, A., Graham, J., Elbel, B., & Roberto, A. (2018). Comparing five front-of-pack nutrition labels' influence on consumers' perceptions and purchase intentions. Preventive medicine, 106, 114–121. doi:10.1016/j.ypmed.2017.10.022

Miles, B. & Michael H. (1994) Qualitative Data Analysis: An Expanded Sourcebook, 2nd Edition. Sage Publications, Thousand Oaks, California. Whilst this book is now quite ‘old’ by referencing standards, it is a classic text for qualitative methods.

Olstad, P. Dana, L. et al. (2015) Using Traffic Light Labels to Improve Food Selection in Recreation and Sport Facility Eating Environments, Appetite, 2015. DOI: 10.1016/j.appet.2015.04.057.

Roberts, M. (2010) The Dissertation Journey: A Practical and Comprehensive Guide to Planning, Writing, and Defending Your Dissertation, Corwin Press.

Roberto A., Bragg, A., Seamans J., Mechulan, L., Novak, N., Brownell, D. (2012) Evaluation of Consumer Understanding of Different Front-of-Package Nutrition Labels, 2010-2011. Prev Chronic Dis;9:120015.

Veerman, J., Barendregt, J., Beeck, V., Seidell, J., Mackenbach, J. (2007) Stemming the obesity epidemic: a tantalizing prospect. Obesity. 2007;15 2365–2370.

Wootan, M., Osborn, M. (2006) Availability of nutrition information from chain restaurants in the United States. Am J Prev Med. 2006;30:266–268.