Trust me? Consumer trust in expert information on food product labels

Journal Pre-proof Trust me? Consumer trust in expert information on food product labels Christoph D.D. Rupprecht, Lei Fujiyoshi, Steven R. McGreevy, Ichiro Tayasu PII:

S0278-6915(20)30058-2

DOI:

https://doi.org/10.1016/j.fct.2020.111170

Reference:

FCT 111170

To appear in:

Food and Chemical Toxicology

Received Date: 6 September 2019 Revised Date:

14 January 2020

Accepted Date: 28 January 2020

Please cite this article as: Rupprecht, C.D.D., Fujiyoshi, L., McGreevy, S.R., Tayasu, I., Trust me? Consumer trust in expert information on food product labels, Food and Chemical Toxicology (2020), doi: https://doi.org/10.1016/j.fct.2020.111170. This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. © 2020 Published by Elsevier Ltd.

CRediT author statement: Christoph Rupprecht: Conceptualization, Methodology, Formal Analysis, Investigation, Writing - Original Draft, Writing - Review & Editing, Visualization. Lei Fujiyoshi: Conceptualization, Methodology, Investigation, Writing - Original Draft, Writing- Review & Editing, Visualization. Steven McGreevy: Conceptualization, Methodology, Investigation, Writing - Original Draft, Writing- Review & Editing, Visualization, Supervision, Project administration. Ichiro Tayasu: Conceptualization, Methodology, Investigation, Writing- Review & Editing, Visualization, Supervision, Project administration, Funding acquisition.

Title: Trust me? Consumer trust in expert information on food product labels Christoph D. D. Rupprecht , Lei Fujiyoshi , Steven R. McGreevy , Ichiro Tayasu a,*

a

a

(name order: given name – middle initials – family name) Research Institute for Humanity and Nature, Kyoto, Japan (457-4 Motoyama, Kamigamo, Kita-ku, Kyoto, 603-8047 JAPAN)

a

*Corresponding author: Christoph D. D. Rupprecht

457-4 Motoyama, Kamigamo, Kita-ku, Kyoto, 603-8047 JAPAN E-mail addresses: [email protected] (C. Rupprecht), [email protected] (L. Fujiyoshi), [email protected] (S. McGreevy), [email protected] (I. Tayasu)

a

1

Trust me? Consumer trust in expert

2

information on food product labels

3 4

Keywords

5

product labels; food information; consumer survey; expert knowledge; traceability; food

6

metrology;

7 8

Highlights

9

* Usefulness of food product label information depends on consumers’ trust.

10

* Online survey used to clarify consumer trust in sources of label information across four

11

food products.

12

* Expert labels based on scientific analysis were highly trusted regardless of food type or

13

country.

14

* Expert labels might play an important role as trusted sources of information in global food

15

system.

16

17

Abstract

18

Food product labels can provide consumers with rich, specific, expert-certified product

19

information. However, sources of label information differ. How do consumers then evaluate

20

label trustworthiness of expert labels in comparison to other commonly used label types? We

21

present results from a representative online survey (N=10,000) of consumers in Japan, the

22

USA, Germany, China and Thailand using professionally designed labels for four food types

23

(milk, honey, oil, wine) and five different sources of food information (farmers,

24

government/administration, producer associations, experts, and consumers). We tested

25

label legibility through identification of the label information source and asked respondents to

26

evaluate the trustworthiness of labels using a six-scale instrument ranging from overall label

27

trust to purchase intent. Results show that label legibility varied between countries, with

28

expert labels scoring lowest. Nevertheless, respondents correctly identifying all label

29

information sources chose expert labels as the most or second-most trustworthy across all

30

countries and food types, while consumer labels scored low. Demographic factors exhibited

31

weak influence. Results suggest expert labels might play an important role as trusted

32

sources of information in an increasingly complex global food system. Finally, we consider

33

the implications of the study for a potential institutionalization of expert labels based on the

34

Japanese context.

35

36

1. Introduction

37

Food consumers want assurance that the food they eat is safe and that the information

38

accompanying food products is accurate. In a complex, interconnected, and global food

39

system, trusting food labels and sources of food information is difficult. Food tampering,

40

deliberate mislabeling, and scandals only further decrease public trust in those that provide

41

food (Charlebois et al., 2016; Mazzocchi et al., 2008). To address these concerns, efforts for

42

greater transparency in the food chain, whether it be through traceability measures (i.e.

43

Badia-Melis et al., 2015; Regattieri et al., 2007) or food metrology (i.e. METROFOOD, 2019),

44

have been increasingly deployed. For instance, stable isotope analysis has been used to

45

authenticate the origin of food in cases where fraud was suspected. Scientific testing to

46

verify food quality and safety (e.g. Blaauboer et al., 2016) has the potential to provide the

47

assurance consumers seek.

48 49

As reliable as these scientific tests may be, more fundamental hurdles remain at the level of

50

consumer evaluation of information provided in food labeling and how it affects the decision-

51

making process. Consumers must successfully interpret the source of the information,

52

understand the information, and trust it enough for it to factor into the decision to buy. Many

53

scholars in the fields of marketing, consumer studies, and food policy have researched the

54

role of “quality labels” that indicate quality assurance or certification, and how scientific

55

information related to nutrition, origin, and production conditions (i.e. organic) on food

56

labeling affect consumer decision making processes (Aprile et al., 2012; Cowburn and

57

Stockley, 2005; Grunert and Wills, 2007; Newman et al., 2014; Tonkin et al., 2015). A

58

common refrain in research of this kind points to the high volume and diversity of labels on

59

the market today and how this makes it difficult for consumers to know who provides the

60

information and whether it is trustworthy. Navigating a sea of food labels can cause “label

61

fatigue” and undermine the original intention behind providing information in the first place

62

(Atkinson and Rosenthal, 2014; Ilbery et al., 2005). Other studies focus on the design or

63

level of detailed information on labels as it relates to credibility (D’Souza et al., 2007).

64 65

Quality labels for food products are issued by an array of public and private sources-- third-

66

party certifiers, business associations, government bodies, and farmers are some examples.

67

Research analyzing these sources of information for food labeling and how consumers

68

evaluate their credibility or trustworthiness remains limited. Moussa and Touzani (2008)

69

used signal theory to examine how consumers perceived the credibility of quality labels from

70

three different sources of information. However, they were interested in how to measure

71

perceived credibility signaled by the label in general and not if the specific sources of

72

information were seen as trustworthy (Moussa and Touzani, 2008). Following up on this

73

work, Atkinson and Rosenthal (2014) tested the degree to which source, argument

74

specificity, and product involvement influenced consumer trust of ecolabels on milk in the US.

75

They found that eco-labels for milk were more trusted if they had detailed information and

76

were issued from the government and businesses (Atkinson and Rosenthal, 2014). Eden et

77

al. (2008) used focus group interviews to examine consumer confidence in food and non-

78

food related assurance schemes by government, commercial, and third-party sources and

79

found that consumers tend to doubt their legitimacy. In particular, assurance schemes by

80

NGOs were confusing and consumers were suspicious of the organizations’ intentions,

81

biases, and capacities for monitoring and testing. The authors advised that “knowledge

82

intermediaries”

83

simplistically...and ready to respond positively to information about the supply chain” (Eden

84

et al., 2008a, pp. 2, 13). These studies recognize a need to know more about how

85

consumers evaluate existing sources of food information and how scientific testing for

86

greater food transparency might allow independent scientific “expert” bodies, with no interest

87

in the sale of products, to emerge as a new source of trustworthy information.

(assurance

schemes

and

labels)

not

“imagine

consumers

too

88 89

Independent neutral researchers and their scientific techniques to verify food authenticity are

90

an ideal source of food-related information in the global food system. Stable isotope analysis

91

of various food items has been conducted by many researchers and reported as a useful

92

tool to verify geographic origin (Camin et al., 2010; Crittenden et al., 2007; Guo et al., 2010;

93

Kukusamude and Kongsri, 2018) and purity (Jamin et al., 2005; Meier Augenstein, 2010).

94

Specific examples that use stable isotope analysis include identifying the geographic origin

95

at a regional scale for olive oil in Italy (Camin et al., 2010), milk in Australia (Crittenden et al.,

96

2007), beef in China (Guo et al., 2010), and rice in Thailand (Kukusamude and Kongsri,

97

2018). As for food purity, carbon isotope ratios of corn oil sold in the United Kingdom were

98

analyzed on behalf of the Food Standards Agency and found that 35% of oils on the market

99

contained oil from undeclared sources not indicated on the labeling (Meier Augenstein,

100

2010). Another example that measured the hydrogen isotope ratios of citric acid in orange

101

juice confirmed that products had been spiked with commercially produced citric acid (Jamin

102

et al., 2005). All of these studies suggest the potential role of stable isotope analysis in

103

clarifying origin and purity, leading to greater food authenticity. In Europe, research

104

infrastructure is being developed to mainstream laboratory testing of food for quality

105

assurance (i.e. METROFOOD, 2019). This points to a need for integrating, and

106

institutionalizing science-based methods to assure consumer trust in food quality

107

internationally, an effort our work aims to contribute to through providing data to inform policy

108

and practices. We take this topic up again in the discussion section. These examples also

109

suggest the potential for the creation of new, science-based sources of food information.

110 111

The assurance that scientific experts might provide would be perceived differently across

112

different country contexts. Country-level differences of trust for food information sources is

113

not well studied. McGreevy and Akitsu (2016) surveyed (n = 5825) parents of elementary

114

school students, asked how much they trusted different sources of food information in five

115

countries (Japan, Korea, France, Germany, and the USA), and found significant differences

116

between them. For example, in Japan and Korea, the degree of trust was higher for products

117

with a photo of a farmer’s face, while consumers in the USA, France, and Germany were

118

generally distrustful of government and farmers. Around the world, there is concern for public

119

trust in science and scientific expertise (Haerlin and Parr, 1999; Priest et al., 2003) and the

120

degree to which this trust varies in different countries and cultural contexts is still largely

121

unknown.

122 123

In this study, a large-scale, multi-country online survey of consumers was conducted to

124

clarify consumer evaluation of trustworthiness of food labels from different sources, with

125

special consideration for labels from “expert” sources. Throughout the paper, the term

126

“expert” is defined as “independent, neutral researchers who use scientific methods to test

127

and analyze foods” (See Table 2). The following research questions were examined: 1)

128

What food characteristics to consumers value?; 2) What sources of food label information to

129

consumers trust most?; 3) How do consumers evaluate labels from expert sources in

130

comparison to other commonly used label sources?; 4) Do socio-demographic factors (age,

131

income, sex, children in household) influence the trust of expert labels?

132

133

2. Material and methods

134

2.1. Survey design

135

The survey tested four food types in five countries with five different sources of label

136

information for each food type. There are compelling reasons as to why milk, cooking oil,

137

honey, and wine were selected as food types for testing. Milk and cooking oil are everyday

138

products used on a daily basis, while honey and wine are more expensive, luxury products.

139

Every one of the four food types either has a history of food tampering, deliberate

140

mislabeling, or is given special value for its geographic origin (Table 1). There are cases of

141

each of the food types being diluted, enriched, or blended with alternative substances,

142

compromising its purity. Wine in particular, and to a lesser extent honey, milk, and oil, is

143

regarded as having higher quality if it originates from certain regions (e.g. wine from

144

Bordeaux, olive oil from Liguria). In each of these circumstances, scientific testing in the form

145

of stable isotope analysis could verify to a high degree of certainty the purity or geographic

146

origin of the product and its ingredients.

147 148

Table 1

149

Examples of food fraud and mislabelling. Food

Country

Year

Type

Issue

Honey

Japan

2018

origin

Imported honey was Sankei newspaper labelled “Japanese (Sankei 2018)

Olive

Italy

2016

origin

Source

as https://www.sankei.com/regi honey” on/news/180119/rgn180119 0057-n1.html

Extra virgin olive oils Reuters

made in Syria and https://jp.reuters.com/article/

oil

Turkey were labelled olives-idJPKCN0VD0DK as “made in Italy” (Reuters 2016) Wine

USA

2016

origin

Wine from Napa

outside CBS San Francisco was https://sanfrancisco.cbslocal.

Valley

labelled as “made in

com/2016/11/04/formernapa-valley-winemaker-

Napa Valley”

arrested-for-fraud-

(CBS SF 2016) mislabeling-wines/

Milk

Japan

1996

ingredient

Skim milk was mixed Ministry of Economy, Trade with raw milk and and Industry labelled

as

“whole

https://www.meti.go.jp/policy/ economy/chizai/chiteki/pdf/1

milk”

9hyoji/houkokusho.pdf

(METI 1996) Honey

Thailand

2017

ingredient

Honey diluted with Bangkok Post corn

syrup,

sugar, https://www.bangkokpost.co

and honey fragrance (Bangkok Post 2017)

m/thailand/general/1255515/ adulterated-honey-seized-inchiang-mai

Milk

China

2008

ingredient

Milk formula

and

infant The Guardian

adulterated https://www.theguardian.com

with melamine (The /world/2008/dec/02/china Guardian 2008) Honey

Germany,

2010

origin

Honey imported from Deutsche

Welle

USA,

China re-labeled and https://www.dw.com/en/us-

China

exported

to

(Deutsche

USA stings-german-chineseWelle

5966640

2010) Wine

Germany, Austria

1985

ingredient,

Austrian

origin

mixed

honey-smuggling-scam/a-

was Wikipedia

wine with

toxic https://en.wikipedia.org/wiki/

substances

and

labeled as “made in

1985_diethylene_glycol_win e_scandal

Germany” (Wikipedia 2019) 150 151 152

Japan, Germany, USA, China, and Thailand were selected as survey countries for a number

153

of reasons including trade relations, history of food safety scares, development of middle

154

class consumer culture, and potential for developing science expert food assurance bodies.

155

Significant food trade relationships exist between many of these countries: China and

156

Thailand are major exporting countries to Japan, Germany and the USA. Food scandals and

157

tampering have occurred in every country, some more so than others. Japan, USA, and

158

Germany have highly developed consumer cultures, while China and Thailand are quickly

159

shifting toward widespread consumerism. The majority of the countries included have the

160

scientific capacity for conducting food-related testing (Guo et al., 2010; Kukusamude and

161

Kongsri, 2018).

162 163

For this consumer survey, farmers, government or administrative bodies, producer

164

associations, other consumers, and scientific experts were selected as the five types of

165

sources of food label information for the following reasons (See Table 2). Many consumers

166

feel disconnected from the origins of their food and seek direct relationships with farmers as

167

a way to assure themselves of food safety (Kneafsey et al., 2008). Building on the work of

168

McGreevy and Akitsu (2016), the survey sought to test consumer trust in farmers labels, in

169

particular labels with a farmer’s face, against other label types from mainstream information

170

sources. Government/administration and producer association labeling on food products is

171

common and consumers may have varying degrees of trust for these two sources depending

172

on the history of food scandal in their country and sentiment toward government and industry

173

in general. With the rise of online shopping platforms like Amazon and Alibaba, consumers

174

tend to encounter product reviews performed by other consumers on a regular basis (Chen,

175

2017). Research points to consumer reviews as having strong influence over perceived trust

176

of businesses (Banerjee et al., 2017) and purchasing intent (Ketelaar et al., 2015).

177 178

Table 2

179

Five types of sources of label information and their definitions. Label information

Definition/description of label information source.

source Producers

People who grow or produce the food.

Government/admin

Departments of the government responsible for food regulations and

istration

laws.

Producer

Body representing businesses that process, distribute and sell food.

association Experts

Independent, neutral researchers who use scientific methods to test and analyze foods.

Consumers

People who evaluate food based on their personal experience or preference.

180

181

2.2. Data collection and analysis

182

Data was collected across five countries (Japan, Germany, USA, China, Thailand) through

183

an online survey coordinated by Macromill, a major Japanese market research company.

184

Respondents were recruited from the service’s consumer panels, consented to and received

185

compensation for their participation according to the service’s policies (exact compensation

186

per participant is not disclosed but likely less than US$10 or equivalent). Responses were

187

collected during 2018 and sampled to be as representative as possible for the population

188

demographic of the respective countries. The study was approved by the home institution’s

189

research ethics committee (RIHN 2017-9).

190 191

The survey instrument (see Appendix A) consisted of 30 questions in five parts and was

192

designed in line with the research questions of this study: (1) respondents’ socio-

193

demographic data; (2) respondent’s trust in different sources of food information; (3)

194

importance of food characteristics (safety, region of origin, purity, taste, freshness, price) for

195

respondents for four food types (milk, honey, oil, wine); (4) matching labels used in the

196

questionnaire to food information providers; (5) a modified set of questions to measure

197

respondents’ trust in different labels (producer, government/administration, producer

198

association, expert, consumer) across for food types (milk, honey, oil, wine). Question types

199

included predominantly Likert-scale questions with some single-choice questions. The

200

survey instrument was conceived in Japanese and translated into English, German, Chinese

201

and Thai. For all languages, native speakers helped to ensure the survey instrument was

202

linguistically correct and easy to read. Questions in part five drew on previous research by

203

Moussa and Touzani (2008) and their refined ‘perceived credibility of quality labels’

204

measurement scale. This scale was further modified by replacing a question about

205

expertness with a question measuring purchase intent (Table 3). Labels used in part four

206

and five were designed in collaboration with a professional designer to avoid introducing bias

207

towards existing logo marks and product brands. In part five, labels were displayed enlarged

208

from a mock product on a background designed to evoke a shopping context (Fig. 1). Stock

209

photos of individuals pictured in producer labels were adapted to geographic context.

210

Data was analyzed using descriptive and inferential statistics following procedures described

211

by Field et al. (2012) using jamovi (jamovi Team, 2019), R (R-Project, 2018) and its package

212

likert (Bryer et al., 2016). Mann-Whitney and Spearman tests were used to analyze the

213

relationship between variables.

214

215 216

Fig. 1. Images of labels on products in front of shopping background as used in part 5 of the

217

survey instrument: a) USA milk expert label, b) China oil producer association label, c)

218

Thailand honey producer label, d) Germany honey producer label, e) Japan wine consumer

219

label, f) Japan wine government/administration label.

220 221

Table 3

222

Modified perceived label trust measurement scale (based on Moussa and Touzani, 2008). I can trust what this label says. This label is honest. The creator of this label has good intentions. The creator of this label has passed strict tests before issuing it. This label inspires confidence. I would purchase this item.

223

224

3. Results

225

3.1. Sample characteristics

226

Samples for the population demographics of the respective countries with 2,000 valid

227

responses per country (total n = 10,000) were collected as representative as possible (older

228

respondents in Thailand are slightly underrepresented, Table 4). Countries varied in age

229

structure, income structure, and proportion of households with children.

230 231

Table 4

232

Sample characteristics across five countries (Japan, Germany, USA, China, Thailand)

233

through an online survey.

Japan

USA

Germany

China

Thailand

Female

50.7%

50.8%

50.3%

48.9%

51.1%

Male

49.3%

49.2%

49.7%

51.1%

48.9%

20s

14.0%

20.1%

15.8%

23.2%

23.8%

30s

18.7%

18.9%

15.8%

21.8%

26.7%

40s

17.4%

20.3%

22.3%

23.0%

24.3%

50s

17.1%

19.5%

18.6%

16.4%

17.5%

60s

19.2%

13.6%

14.7%

9.9%

7.0%

70s

13.6%

7.7%

12.8%

5.8%

0.8%

1 (lowest)

9.4%

15.3%

8.3%

10.0%

4.3%

2

27.2%

21.1%

15.6%

32.9%

12.0%

3

26.4%

20.3%

34.0%

36.7%

26.4%

4

16.2%

16.0%

20.3%

11.0%

27.2%

5

10.6%

11.0%

12.3%

4.6%

12.5%

6

5.0%

5.9%

4.3%

3.2%

8.6%

7 (highest)

5.1%

10.4%

5.2%

1.6%

9.1%

Children in

Yes

15.5%

30.6%

22.6%

68.3%

51.9%

household

No

84.5%

69.3%

77.4%

31.8%

48.0%

Sex

Age

Income

234

235

3.2. Trust in food information sources

236

Respondent’s trust in food source information was mostly positive (Fig. 2). Experts were

237

highly trusted as sources of food information. Government/administration and producer

238

associations were least trusted in Japan, USA and Germany, while in China and Thailand

239

government/administration was perceived as more trustworthy. Producers were most highly

240

trusted in the USA, but trusted least in China. Consumers were comparatively less trusted in

241

Japan than in other countries.

242

243 244

Fig. 2. Respondents’ trust in sources of food information.

245

3.3. Importance of food characteristics for consumers?

246

Importance of food characteristics (safety, origin, purity, taste, freshness, price) varied widely

247

between food types and countries (Table 5). Price features prominently in Japan, but

248

respondents generally considered it less important in other countries. Freshness for milk and

249

taste for wine were seen as important across all countries. Origin was considered to be of

250

low importance in all countries across all food types. In China and Thailand, safety and purity

251

were ranked comparatively high across all foods.

252

253

Table 5

254

Importance of different food characteristics by country/food*.

Milk

Honey

Oil

Wine

Country

Japan

USA

Germany

China

Thailand

Safety

4.06

4.11

4.35

4.59

4.69

Origin

3.67

3.67

4.09

3.99

4.38

Purity

3.70

4.16

4.49

4.23

4.65

Taste

3.98

4.28

4.50

4.24

4.44

Freshness

4.14

4.36

4.56

4.54

4.64

Price

4.09

3.94

3.88

3.89

4.07

Safety

3.92

3.89

4.26

4.54

4.59

Origin

3.73

3.54

4.05

3.86

4.34

Purity

3.75

4.00

4.43

4.34

4.58

Taste

3.86

4.11

4.56

4.21

4.39

Freshness

3.62

3.96

4.25

4.33

4.47

Price

3.96

3.81

3.83

3.91

4.08

Safety

3.85

3.88

4.22

4.59

4.62

Origin

3.36

3.39

3.76

3.86

4.20

Purity

3.48

3.89

4.35

4.23

4.53

Taste

3.49

3.77

4.36

4.15

3.94

Freshness

3.60

3.83

4.22

4.32

4.37

Price

4.00

3.94

3.91

3.96

4.08

Safety

3.37

3.53

3.99

4.39

4.37

Origin

3.30

3.26

3.70

4.01

4.16

Purity

3.08

3.55

4.17

4.11

4.29

Taste

3.59

3.95

4.43

4.27

4.40

Freshness

3.10

3.58

3.73

4.03

3.97

Price

3.59

3.65

3.75

3.92

4.03

Note: colour coding is applied on a per-food per-country basis (e.g. Milk in Japan). Yellow represents higher importance, blue lower importance. Values are not adjusted for absolute differences between countries.

255

3.4. Consumers evaluation of expert labels

256

3.4.1. Instrument validity & label legibility

257

Reliability of the modified trust measurement set (Table 6) was high (Cronbach’s alpha >

258

0.9). The number of respondents who correctly identified the information source of all

259

respective labels varied highly (Japan 57%, USA 33%, Thailand 32%, Germany 30%, China

260

16%). Label legibility was lowest for expert labels (Table 6). This has implications for the

261

design and implementation of expert labels revisited in the discussion section.

262 263

Table 6

264

Legibility of labels by country. Japan

USA

Germany

China

Thailand

Producer

83.8%

72.8%

71.3%

62.2%

66.7%

Government/admin.

83.9%

74.5%

72.0%

70.2%

70.6%

Producer association

77.0%

60.5%

62.4%

55.4%

56.6%

Expert

68.2%

46.6%

43.5%

28.0%

49.6%

Consumer

82.9%

72.2%

72.1%

71.3%

72.0%

Note: colour coding is applied on a per-country basis.

265

3.4.2. Label trust

266

Respondents across all countries showed high trust levels for expert-sourced labels across

267

all food types (Table 7). In Japan, USA, and Germany, expert label trustworthiness was most

268

highly ranked, while in China and Thailand expert labels were ranked second place after

269

government/admin labels. Consumer label trustworthiness was perceived as very low across

270

all countries and food types. In China and Thailand, producer label trustworthiness was also

271

consistently low, while in other countries they generally placed in the middle. Respondents

272

also showed variation in absolute trust levels between countries. Average label trust across

273

all food types ranged between 3.23 (Japan) and 3.81 (China). Trust below 3.0 (indicating

274

lack of trust) only occurred for consumer labels, and only in Japan (for all food types) and

275

Germany (for oil). This indicates respondents in general perceived trustworthiness of the

276

labels in this study as neutral (3.0) to trustworthy (above 3).

277 278

Table 7

279

Respondents’ trust in labels by country. Food

Label type

Japan

USA

Germany

China

Thailand

All

Milk

Producer

3.31

3.51

3.50

3.63

3.47

3.484

Government/admin.

3.39

3.68

3.49

4.08

4.18

3.764

Producer assoc.

3.32

3.31

3.38

3.77

3.67

3.490

Expert

3.49

3.84

3.81

4.04

4.11

3.858

Consumer

2.76

3.24

3.04

3.62

3.24

3.18

3.36

3.56

3.57

3.63

3.47

3.518

Government/admin.

3.38

3.70

3.52

4.06

4.19

3.770

Producer assoc.

3.26

3.47

3.55

3.73

3.66

3.534

Expert

3.49

3.85

3.86

4.03

4.11

3.868

Consumer

2.76

3.28

3.04

3.59

3.25

3.184

Honey Producer

Oil

Wine

Producer

3.29

3.45

3.44

3.57

3.38

3.426

Government/admin.

3.39

3.69

3.49

4.05

4.20

3.764

Prod Assoc

3.26

3.35

3.29

3.77

3.60

3.454

Expert

3.49

3.83

3.79

4.04

4.09

3.848

Consumer

2.73

3.25

2.99

3.58

3.22

3.154

Producer

3.35

3.50

3.50

3.58

3.38

3.462

Government/admin.

3.30

3.62

3.44

4.03

4.07

3.692

Prod Assoc

3.20

3.35

3.42

3.73

3.60

3.460

Expert

3.41

3.72

3.76

4.03

4.02

3.788

Consumer

2.74

3.27

3.01

3.59

3.22

3.166

Note: colour coding is applied on a per-country basis.

280

3.4.3. Statistical analysis: factors associated with label trust

281

In general, socio-demographic factors had only limited effects on label trust (Table 8).

282

Among socio-demographic factors, the presence of children in the household most strongly

283

affected label trust, and respondents from these households generally perceived labels as

284

more trustworthy than respondents from households without children. Income only had a

285

noteworthy effect in China, where higher income was correlated with higher label trust. Both

286

trust in the source of food information and stronger preferences for specific food

287

characteristics were correlated with higher label trust. The specific trust question in the trust

288

measurement scale was strongly correlated with willingness to buy.

289 290

Table 8

291

Factors associated with label trust. Factor

Effect observed

Effect range*

Age

All respondents**: higher age was weakly correlated

rs=-0.02 to -0.12

with lower trust across all labels, except for the milk producer label. Correctly identified**: higher age was very weakly

rs=-0.04 to -0.07

correlated with lower label trust for experts (all foods). Income

All respondents: higher income was moderately

rs=0.21 to 0.25

correlated with higher label trust in China, but only

(China only)

weakly, not correlated, or weakly negatively correlated in all other countries. Correctly identified: higher income was moderately

rs=0.20 to 0.31

correlated with higher label trust in China, but only

(China only)

weakly, not correlated, or weakly negatively correlated in all other countries. Sex

All respondents: women had generally slightly higher

d=0.03 to 0.13

trust in labels, except for consumer labels, of which oil and wine consumer labels were slightly more trusted by men. Correctly identified: women had slightly higher trust in

d=0.05 to 0.15

producer association labels for all foods (and government labels for honey only), while men had slightly higher trust in consumer labels for honey, oil and wine. Children in

All respondents & correctly identified: respondents in

d=0.24 to 0.44

household

households with children generally showed

(all respondents)

moderately higher label trust across all food types

d=0.20 to 0.42

and labels.

(correctly identified)

Food

All respondents & correctly identified: trust in food

rs=0.30 to 0.47

information

information source was moderately correlated with

(all respondents)

source trust

trust in the associated label for all food types.

rs=0.30 to 0.47 (correctly identified)

Food

All respondents & correctly identified: stronger

rs=0.12 to 0.35

preferences

preferences for food characteristics were weakly to

(all respondents)

moderately correlated with label trust, with stronger

rs=0.06 to 0.36

correlations for government and expert labels with

(correctly identified)

safety and purity preferences. Correlation was weakest for consumer labels as well as for price preferences. Willingness

All respondents & correctly identified: Willingness to

to buy***

buy was strongly correlated with label trust for all

rs=0.70 to 0.80

labels and food types. Notes: Only effect sizes/ranges for tests with p<.05 are reported. *Effect range based on Spearman correlation (rs), Mann-Whitney U test (Cohen’s d). ** Due to the low label legibility for some labels and countries, tests were conducted for all respondents (n=10,000), and again for respondents who correctly identified all labels (n=3,343). *** For this test, only the correlations between question 5 (label trust) and 6 (willingness to buy) of the comprehensive trust scale were tested. 292

293

294

4. Discussion

295

4.1. Major findings and further research

296 297

Trust in the five sources of information for each country was relatively the same with only a

298

few noticeable differences (see Fig. 2). Food information from experts was the most

299

trustworthy for all countries, which strengthens the argument for food metrology and public

300

trust in science. One notable difference was the gap in perceived trustworthiness for

301

government-derived food information between China and Thailand (high trust) and the US,

302

Japan, and Germany (very low trust). In both China and Thailand, food safety is a concern

303

for consumers and the government, which might explain the high level of trust. For example,

304

in China, Liu et al. (2014) also found high levels of trust for the government, which was

305

viewed as having concern for citizens’ health.

306 307

A major finding of this research is that the expert labels were deemed highly trustworthy in

308

each of the five countries surveyed, ranking first in Japan, USA, and Germany and second in

309

China and Thailand after government/admin labels (See Table 7). Public trust in scientific

310

experts to provide assurance for food safety and quality is strong, indicating a demand for

311

this kind of expert-sourced information in the food marketplace. However, of all the labels

312

surveyed, expert labels were consistently ranked as the least legible (see Table 6), meaning

313

that survey participants had the hardest time understanding the source of the label. This is

314

most likely due to the fact that expert labels of this design are currently not encountered in

315

anyway by consumers. Special attention should be paid as to how expert-sourced labels are

316

designed if such a label would be created by scientists. Further research is needed to

317

identify what “signals” are appropriate in signifying expert-sourced information and how they

318

might be distinct from existing government, administrative, or producer association signifiers.

319 320

Another difference amongst the countries was in producer trustworthiness, with four of the

321

five countries seeing farmers as trustworthy. Again, issues related to food safety could

322

explain why Chinese consumers ranked producers as being the least trustworthy of the five

323

sources of information surveyed, a sentiment confirmed by Zhang et al. (2016). High

324

trustworthiness for farmers in the US was not seen in previous large sample surveys (i.e.

325

McGreevy and Akitsu, 2016; Sapp et al., 2010), making these findings unique and deserving

326

of more research. The high level of trust for farmers was not reflected when considering the

327

producer labels. Producer labels featuring a farmers’ face were generally seen as having

328

middle to low levels of trustworthiness in each country when compared to the other sources,

329

which is consistent with previous findings (e.g. McGreevy and Akitsu, 2016), and confirms

330

some of the same arguments that Eden et al. (2008a, 2008b) have written about regarding

331

“knowledge fixes” to bridge the perceived “disconnect” between consumers and producers.

332

Eden et al. (2008a, 2008b) point out that it may not be the information on labels that is most

333

convincing, but the spaces (the retail grocery stores or farmers markets) in which consumers

334

encounter food and, if present, the producers themselves that provide assurance and

335

feelings of reconnection. In this study, the labels used for the survey were contextualized by

336

images that signified shopping in a supermarket, which may have contributed to survey

337

participants viewing farmer’s faces on labels in another way than they would in, for example,

338

a farmers market where other forms of assurance might be provided beyond labeling.

339

Further research is needed to elaborate the difference in high generalized trust of producers

340

and low level of trust for labels with farmer’s faces.

341 342

A few patterns were observable when looking at what food characteristics consumers value,

343

but in general, cross-country results were diverse (see Table 5). Among the options of safety,

344

origin, purity, taste, freshness, and price for milk, honey, oil, and wine, only Japanese

345

consumers ranked price as an important factor. China and Thailand ranked safety and purity

346

as a priority for every food item. Unsurprisingly, freshness for milk and taste for wine ranked

347

highly in every country, while safety, purity, and taste were important for honey. Origin was

348

consistently ranked as the lowest prioritized characteristic, even for products such as oil and

349

wine, which sometimes valorize their geographic origins as a way to add value or for

350

marketing purposes. These findings have implications for the receptiveness of expert-

351

sourced labels that use stable isotope analysis to verify product origin, since consumers

352

generally don’t consider origin as an important characteristic when shopping.

353 354

Opposite to expert-sourced labels, consumer-sourced labels were seen as the least

355

trustworthy of those surveyed. Research on how internet users evaluate electronic word-of-

356

mouth (eWOM) and online products reviews show that perceived trustworthiness is

357

influential when making online purchases (Reichelt et al., 2014). However, this study shows

358

that when compared to other more established or authoritative sources, consumer-sourced

359

information on food products is less trusted. Particularly intriguing is how Japanese

360

consumers distrusted other consumers as sources of food information, while the other four

361

countries were quite similar. More research is needed to understand how word of mouth

362

information is interpreted and trusted in different socio-cultural settings, as comparative work

363

between Amazon USA and Japan by Lin and Kalwani (2018) suggests. With the rise in

364

online shopping for food products, it will be interesting to monitor levels in trust of consumer

365

reviews going forward.

366 367

Socio-demographic factors were fairly insignificant on how they affected label trust (see

368

Table 8). Across all studied countries, respondents with children in the household perceived

369

labels as more trustworthy. This can be explained in part by how parents need to reach

370

quick decisions for purchasing food based on labeling and packaging and are more likely to

371

trust labels at face-value (Abrams et al., 2015). Higher incomes in China correlated to high

372

levels of trust, but wasn’t an important factor elsewhere.

373 374

4.2. Institutionalization of expert assurance for quality labels?

375 376

In this study, expert labels were highly trusted for all food types and across all countries.

377

Following up on our short introduction review of the role of scientific testing in food quality

378

assurance, this suggests that independent, neutral researchers who use scientific methods

379

to test and analyze food and the information obtained from stable isotope analysis (i.e. for

380

origin and purity) have high potential to become a trustworthy source of information in the

381

complex food marketplace. The emergence of a widely used expert label is in line with the

382

trend toward greater supply chain transparency and the development of broad research

383

platforms worldwide (i.e. METROFOOD, 2019).

384

The demand for expert assurance of food safety and integrity could be linked to the

385

institutionalization of an expert assurance label in close partnership with food metrology

386

platforms and scholars. Here we speculate as to how such an institutional structure might be

387

constructed based on a preliminary survey of food labeling institutions in Japan. At present,

388

certification authorities are mainly governmental (Ministries or prefectures) or producer

389

associations, and expert bodies (i.e. research platforms or scientists) that certify food quality

390

do not exist. Table 9 shows five typical quality food labels, the certification authority, and the

391

role of experts taking part in commissions or scientific analysis, but not acting as certification

392

authorities themselves.

393 394

Table 9

395

Five typical food quality labels in Japan and related certification authority and role of experts. Label

Name of label

Certification authority

Label of food for

Ministry of

specified health uses

Health,Labor and Welfare

Role of experts ・scientific analysis ・member of food safety commission ・member of consumer commission

Label of JAS (Japanese

Ministry of

Agricultural Standard)-

Agriculture, Forestry

・scientific analysis

certified organic food

and Fisheries

Label of geographical

Ministry of

indication

Agriculture, Forestry

・member of academic expert commission

and Fisheries

Label of E (Excellent

Prefecture

・member of certification

Quality, Exact commission Expression, and harmony with Ecology) Label of certification of

Japan Salt Industry

safety and health

Association

standard of salt

・scientific analysis ・member of judging commission

396 397

398 399 400

Fig. 3. An example institutional structure for expert assurance quality labeling.

401

We propose an institutional model for expert assurance in Fig. 3. Food producers who want

402

scientifically certified labels on their food products (e.g. origin or purity) submit an application

403

to the expert organization. The expert organization tests the products using various scientific

404

methods (e.g. stable isotope analysis) and the results of the analysis are sent to an

405

independent, neutral expert committee to be judged and verified. After certification, the

406

expert organization provides scientifically certified labels for the food products to the

407

producers.

408 409

Like many certification systems, the question of who bears the costs for testing and

410

organizational capacities will be an issue. But if the demand for greater transparency in the

411

food system is strong enough and if, for example, regulatory bodies in charge of food

412

standards for international trade see the value in supporting expert certification, such a

413

platform is conceivable at the national level. Countries that suffer from food scandals and

414

low consumer trust for food safety, such as China or Thailand, would be ideal candidates to

415

assemble experts in an institutional structure to provide consumers the assurance they

416

desire.

417 418

4.3. Study limitations

419 420

This study had a few limitations that need to be recognized. The survey was conducted

421

online, which allowed for rapid and large number sampling, but can limit the population

422

surveyed. The same online survey company was used in each country to eliminate

423

differences in process. Certain demographic groups were underrepresented in the survey--

424

for example, older participants were less represented in Thailand. The labels used in the

425

survey were not existing product labels, so results could differ if existing labels were used.

426

That said, this choice was made to avoid bias (for example based on product brand) and

427

focus on the authoritative information source in each case. Label designs and text used in

428

them were localized with the help of native language experts to make the labels seem as

429

legitimate as possible.

430

431

Comparisons between countries were limited as this was out of the scope of the paper as it

432

deserves a thorough analysis of social, cultural, and historical elements to help explain the

433

differences observed. More research on the cultural differences of food label perception and

434

trust are needed in the future to better explain how country-specific contextual factors and

435

follow-up studies on this topic are planned by the authors.

436

5. Conclusion

437

When consumers evaluate what information on food product labels they trust, experts

438

emerge as a group that can provide highly trusted information when compared with other

439

sources of information. This holds true across all countries and food types surveyed in this

440

study. Government labels were trusted in countries with a history of food safety scandals

441

(China and Thailand) and labels signifying consumer-sourced information were distrusted in

442

every country. While trust for producers was high, labels with producers’ faces were seen as

443

less trustworthy. Demographic factors only had limited influence on these results. The broad

444

appeal of expert information on food product labels thus suggests that expert labels might

445

play an important role in providing consumers with trusted information in an increasingly

446

complex global food system. However, expert labels were difficult to identify accurately as

447

originating from expert sources, suggesting that labels of this kind would require thoughtful

448

design. In light of recent developments around food chain transparency which heavily rely on

449

experts, expert labels and their institutionalization merit the attention of researchers as well

450

as that of various stakeholders. In this process, questions around label legibility identified in

451

this study will need to be addressed, as will questions around what institutional models for

452

expert assurance and certification systems are appropriate and effective.

453 454 455 456

457

Acknowledgements

458

This research was supported by the Research Institute for Humanity and Nature (RIHN: a

459

constituent member of NIHU), the Environmental Traceability Core Project No. 14200076,

460

and the FEAST Project No. 14200116. Special thanks to Sittidaj Pongkijvorasin

461

(Chulalongkorn University) and Ma Jia (Shanghai Academy of Agricultural Sciences) for their

462

assistance with translation and to Duncan Brotherton for graphic design. We also thank the

463

judges at the 1st ISO-FOOD International Symposium on Isotopic and Other Techniques in

464

Food Safety and Quality for awarding this research the Best Poster Award.

465 466 467

Appendix A. Supplementary data

468

Supplementary data to this article can be found online at ***********.

469

References

470

Abrams, K.M., Evans, C., Duff, B.R.L., 2015. Ignorance is bliss. How parents of preschool

471

children make sense of front-of-package visuals and claims on food. Appetite 87, 20–

472

29. https://doi.org/10.1016/j.appet.2014.12.100

473

Aprile, M.C., Caputo, V., Nayga Jr, R.M., 2012. Consumers’ valuation of food quality labels:

474

the case of the European geographic indication and organic farming labels:

475

Consumers’ valuation of food quality labels. Int. J. Consum. Stud. 36, 158–165.

476

https://doi.org/10.1111/j.1470-6431.2011.01092.x

477

Atkinson, L., Rosenthal, S., 2014. Signaling the Green Sell: The Influence of Eco-Label

478

Source, Argument Specificity, and Product Involvement on Consumer Trust. J.

479

Advert. 43, 33–45. https://doi.org/10.1080/00913367.2013.834803

480

Badia-Melis, R., Mishra, P., Ruiz-García, L., 2015. Food traceability: New trends and recent

481

advances. A review. Food Control 57, 393–401.

482

https://doi.org/10.1016/j.foodcont.2015.05.005

483

Banerjee, S., Bhattacharyya, S., Bose, I., 2017. Whose online reviews to trust?

484

Understanding reviewer trustworthiness and its impact on business. Decis. Support

485

Syst. 96, 17–26. https://doi.org/10.1016/j.dss.2017.01.006

486

Bangkok Post. 2017. “Adulterated honey seized in Chiang Mai.” Available online at:

487

https://www.bangkokpost.com/thailand/general/1255515/adulterated-honey-seized-

488

in-chiang-mai (accessed January 8, 2020)

489

Blaauboer, B.J., Boobis, A.R., Bradford, B., Cockburn, A., Constable, A., Daneshian, M.,

490

Edwards, G., Garthoff, J.A., Jeffery, B., Krul, C., Schuermans, J., 2016. Considering

491

new methodologies in strategies for safety assessment of foods and food ingredients.

492

Food Chem. Toxicol. 91, 19–35. https://doi.org/10.1016/j.fct.2016.02.019

493

Bryer, J., Speerschneider, K., Bryer, M.J., 2016. Package ‘likert.’

494

Camin, F., Larcher, R., Perini, M., Bontempo, L., Bertoldi, D., Gagliano, G., Nicolini, G.,

495

Versini, G., 2010. Characterisation of authentic Italian extra-virgin olive oils by stable

496

isotope ratios of C, O and H and mineral composition. Food Chem. 118, 901–909.

497

https://doi.org/10.1016/j.foodchem.2008.04.059

498

CBS SF Bay Area 2016. Former Napa Valley Winemaker Arrested For Fraud, Mislabeling

499

Wines. https://sanfrancisco.cbslocal.com/2016/11/04/former-napa-valley-winemaker-

500

arrested-for-fraud-mislabeling-wines/ (accessed 4 September 2019).

501

Charlebois, S., Schwab, A., Henn, R., Huck, C.W., 2016. Food fraud: An exploratory study

502

for measuring consumer perception towards mislabeled food products and influence

503

on self-authentication intentions. Trends Food Sci. Technol. 50, 211–218.

504

https://doi.org/10.1016/j.tifs.2016.02.003

505

Chen, C.-W., 2017. Five-star or thumbs up? The influence of rating system types on users’

506

perceptions of information quality, cognitive effort, enjoyment and continuance

507

intention. Internet Res. 27, 478–494. https://doi.org/10.1108/IntR-08-2016-0243

508

Cowburn, G., Stockley, L., 2005. Consumer understanding and use of nutrition labelling: a

509

systematic review. Public Health Nutr. 8, 21–28. https://doi.org/10.1079/PHN2004666

510

Crittenden, R.G., Andrew, A.S., LeFournour, M., Young, M.D., Middleton, H., Stockmann, R.,

511

2007. Determining the geographic origin of milk in Australasia using multi-element

512

stable isotope ratio analysis. Int. Dairy J. 17, 421–428.

513

https://doi.org/10.1016/j.idairyj.2006.05.012

514

D’Souza, C., Taghian, M., Lamb, P., Peretiatko, R., 2007. Green decisions: demographics

515

and consumer understanding of environmental labels. Int. J. Consum. Stud. 31, 371–

516

376. https://doi.org/10.1111/j.1470-6431.2006.00567.x

517

Deutsche Welle, 2010. US stings German-Chinese honey-smuggling scam.

518

https://www.dw.com/en/us-stings-german-chinese-honey-smuggling-scam/a-

519

5966640 (accessed January 8, 2020).

520

Eden, S., Bear, C., Walker, G., 2008a. Understanding and (dis)trusting food assurance

521

schemes: Consumer confidence and the ‘knowledge fix.’ J. Rural Stud. 24, 1–14.

522

https://doi.org/10.1016/j.jrurstud.2007.06.001

523

Eden, S., Bear, C., Walker, G., 2008b. Mucky carrots and other proxies: Problematising the

524

knowledge-fix for sustainable and ethical consumption. Geoforum 39, 1044–1057.

525

https://doi.org/10.1016/j.geoforum.2007.11.001

526

Field, A., Miles, J., Field, Z., 2012. Discovering Statistics Using R. SAGE.

527

Grunert, K.G., Wills, J.M., 2007. A review of European research on consumer response to

528

nutrition information on food labels. J. Public Health 15, 385–399.

529

https://doi.org/10.1007/s10389-007-0101-9

530

Guo, B.L., Wei, Y.M., Pan, J.R., Li, Y., 2010. Stable C and N isotope ratio analysis for

531

regional geographical traceability of cattle in China. Food Chem. 118, 915–920.

532

https://doi.org/10.1016/j.foodchem.2008.09.062

533 534 535

Haerlin, B., Parr, D., 1999. How to restore public trust in science. Nature 400, 499. https://doi.org/10.1038/22867 Howard, P.H., Allen, P., 2010. Beyond Organic and Fair Trade? An Analysis of Ecolabel

536

Preferences in the United States: Beyond Organic and Fair Trade? Rural Sociol. 75,

537

244–269. https://doi.org/10.1111/j.1549-0831.2009.00009.x

538

Ilbery, B., Morris, C., Buller, H., Maye, D., Kneafsey, M., 2005. Product, Process and Place:

539

An Examination of Food Marketing and Labelling Schemes in Europe and North

540

America. Eur. Urban Reg. Stud. 12, 116–132.

541

https://doi.org/10.1177/0969776405048499

542

Jamin, E., Martin, F., Santamaria-Fernandez, R., Lees, M., 2005. Detection of exogenous

543

citric acid in fruit juices by stable isotope ratio analysis. J. Agric. Food Chem. 53,

544

5130–5133. https://doi.org/10.1021/jf050400b

545

jamovi Team, T., 2019. jamovi (Version 0.9.5.3-15).

546

Ketelaar, P.E., Willemsen, L.M., Sleven, L., Kerkhof, P., 2015. The Good, the Bad, and the

547

Expert: How Consumer Expertise Affects Review Valence Effects on Purchase

548

Intentions in Online Product Reviews. J. Comput.-Mediat. Commun. 20, 649–666.

549

https://doi.org/10.1111/jcc4.12139

550 551 552

Kneafsey, M., Cox, R., Holloway, L., Dowler, E., Venn, L., Tuomainen, H., 2008. Reconnecting Consumers, Producers and Food: Exploring Alternatives. Berg. Kukusamude, C., Kongsri, S., 2018. Elemental and isotopic profiling of Thai jasmine rice

553

(Khao Dawk Mali 105) for discrimination of geographical origins in Thung Kula Rong

554

Hai area, Thailand. Food Control 91, 357–364.

555

https://doi.org/10.1016/j.foodcont.2018.04.018

556

Lin, H.-C., Kalwani, M.U., 2018. Culturally Contingent Electronic Word-of-Mouth Signaling

557

and Screening: A Comparative Study of Product Reviews in the United States and

558

Japan. J. Int. Mark. 26, 80–102. https://doi.org/10.1509/jim.17.0016

559 560

Liu, R., Pieniak, Z., Verbeke, W., 2014. Food-related hazards in China: Consumers’ perceptions of risk and trust in information sources. Food Control 46, 291–298.

561 562

https://doi.org/10.1016/j.foodcont.2014.05.033 Mazzocchi, M., Lobb, A., Bruce Traill, W., Cavicchi, A., 2008. Food Scares and Trust: A

563

European Study: Food Scares and Trust: A European Study. J. Agric. Econ. 59, 2–

564

24. https://doi.org/10.1111/j.1477-9552.2007.00142.x

565

McGreevy, S.R., Akitsu, M., 2016. Steering Sustainable Food Consumption in Japan: Trust,

566

Relationships, and the Ties that Bind, in: Sustainable Consumption, The

567

Anthropocene: Politik—Economics—Society—Science. Springer, Cham, pp. 101–

568

117. https://doi.org/10.1007/978-3-319-29665-4_7

569

Meier Augenstein, W., 2010. Stable Isotope Forensics: An introduction to the forensic

570

application of stable isotope analysis. John Wiley & Sons, Ltd.

571

METROFOOD, 2019. Infrastructure for promoting Metrology in Food and Nutrition.

572

https://www.metrofood.eu/ (accessed 22 July 2019).

573

Ministry for Economy, Trade, and Industry 2008. 表示に係る不正競争行為に関する 調査研究

574

報告書 (in Japanese, Research report on acts of unfair competition regarding

575

labeling).

576

https://www.meti.go.jp/policy/economy/chizai/chiteki/pdf/19hyoji/houkokusho.pdf

577

(accessed 4 September 2019).

578

Moussa, S., Touzani, M., 2008. The perceived credibility of quality labels: a scale validation

579

with refinement. Int. J. Consum. Stud. 32, 526–533. https://doi.org/10.1111/j.1470-

580

6431.2008.00713.x

581

Newman, C.L., Turri, A.M., Howlett, E., Stokes, A., 2014. Twenty Years of Country-of-Origin

582

Food Labeling Research: A Review of the Literature and Implications for Food

583

Marketing Systems. J. Macromarketing 34, 505–519.

584

https://doi.org/10.1177/0276146714529306

585

Priest, S.H., Bonfadelli, H., Rusanen, M., 2003. The “Trust Gap” Hypothesis: Predicting

586

Support for Biotechnology Across National Cultures as a Function of Trust in Actors.

587

Risk Anal. 23, 751–766. https://doi.org/10.1111/1539-6924.00353

588 589 590

Regattieri, A., Gamberi, M., Manzini, R., 2007. Traceability of food products: General framework and experimental evidence. J. Food Eng. 10. Reichelt, J., Sievert, J., Jacob, F., 2014. How credibility affects eWOM reading: The

591

influences of expertise, trustworthiness, and similarity on utilitarian and social

592

functions. J. Mark. Commun. 20, 65–81.

593

https://doi.org/10.1080/13527266.2013.797758

594

Reuters 2016. 硫酸銅まぶしたオリーブや偽オリーブ油を押収、イタリア警察( in Japanese,

595

Italian Police seizes fake olive oil and olive oil containing copper sulfate).

596

https://jp.reuters.com/article/olives-idJPKCN0VD0DK (accessed 4 September 2019).

597

R-Project, 2018. R version 3.5.2.

598

Sapp, S.G., Arnot, C., Fallon, J., Fleck, T., Soorholtz, D., Sutton-Vermeulen, M., Wilson,

599

J.J.H., 2010. Consumer Trust in the U.S. Food System: An Examination of the

600

Recreancy Theorem*. Rural Sociol. 74, 525–545. https://doi.org/10.1111/j.1549-

601

0831.2009.tb00703.x

602

Sankei Shimbun 2018. はちみつの産地偽装 上野物産に是正指示

ハンガリー産を「国

603

産」に(in Japanese, Hungary-produced honey mislabeled as produced in Japan;

604

Ueno Bussan instructed to issue correction).

605

https://www.sankei.com/region/news/180119/rgn1801190057-n1.html (accessed 4

606

September 2019).

607

The Guardian. 2008. “Chinese figures show fivefold rise in babies sick from contaminated

608

milk.” Available online at: https://www.theguardian.com/world/2008/dec/02/china

609

(accessed January 8, 2020)

610

Tonkin, E., Wilson, A.M., Coveney, J., Webb, T., Meyer, S.B., 2015. Trust in and through

611

labelling – a systematic review and critique. Br. Food J. 117, 318–338.

612

https://doi.org/10.1108/BFJ-07-2014-0244

613 614

Wikipedia 2020, 1985 diethylene glycol wine scandal. https://en.wikipedia.org/w/index.php?title=1985_diethylene_glycol_wine_scandal&old

615 616

id=934040608 (accessed January 8, 2020) Zhang, L., Xu, Y., Oosterveer, P., Mol, A.P.J., 2016. Consumer trust in different food

617

provisioning schemes: evidence from Beijing, China. J. Clean. Prod. 134, 269–279.

618

https://doi.org/10.1016/j.jclepro.2015.09.078

619 620 621 622 623

Declaration of interests ☒ The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. ☐The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: