Impact of Religion on Halal MeatConsumption Decision Making in Belgium

KARIJN BONNEUniversity College Gent, Belgium

IRIS VERMEIRUniversity College Gent and Gent University, Belgium

WIM VERBEKEGhent University, Belgium

This study investigates the determinants of halal meat consumptionwithin a Belgian Muslim migration population using the theory ofplanned behavior as a conceptual framework, with a focus on therole of self-identity as a Muslim and acculturation in the host coun-try. Cross-sectional data were collected through a survey with 367Muslims mainly originating from North Africa and living inBelgium. Findings reveal that in general, a positive health attitudetoward halal meat predicts the intention to consume halal meatamong Muslims. Perceived lack of safety measures or poor beliefin the safety controls are shown to be potential barriers preventingMuslim consumers from eating halal meat. Low acculturatedMuslims rely strongly on their positive personal attitude toward thehealth status of halal meat, whereas high acculturated Muslims relyon health attitude, animal welfare attitudes, and safety whenintending to consume halal meat. Muslims with a high Muslimself-identity intend to eat halal meat because they believe that it ishealthy whereas Muslims with a low Muslim self-identity are ratherinfluenced by religious peers, together with their personal healthattitude and availability concerns.

KEYWORDS Acculturation, consumer, halal, meat, religion,theory of planned behavior

Submitted: September 2007; First Revision: March 2008; Second Revision: May 2008; Accepted:September 2008

Address correspondence to Karijn Bonne, University College Gent, Department ofBusiness Studies and Public Administration, Belgium. E-mail: Karijn.Bonne@Hogent.be

Journal of International Food & Agribusiness Marketing, 21:526, 2009

Copyright # Taylor & Francis Group, LLCISSN: 0897-4438 print=1528-6983 online

DOI: 10.1080/08974430802480628

5

In many societies, religion plays an influential role in shaping food choice(Dindyal, 2003; Musaiger, 1993). Although religion has been a significantforce in the lives of many individuals, its exact role in consumer food choiceis rather unclear (Delener, 1994). Ample evidence has been provided thatreligion can influence consumer attitude and behavior in general (Delener,1994; Kanekar & Merchant, 2001; Pettinger, Holdsworth, & Gerber, 2004)and food purchasing decisions and eating habits in particular (Asp, 1999;Just, Heiman, & Zilberman, 2007; Mennell, Murcott, & Van Ootterloo, 1992;Mullen, Williams, & Hunt, 2000; Shatenstein & Ghadirian, 1997).

The impact of religion on food consumption depends on the religionitself and on the extent to which individuals interpret and follow the teach-ings of their religion (Heiman, McWilliams, & Zilberman, 2005). Most reli-gions forbid certain foods except for Christianity, which has no foodtaboos (Sack, 2001, p. 218). Especially meat is often strictly regulated in caseswhere religious considerations prevail (Shatenstein and Ghadirian, 1997).

One of the religions with food prohibitions is Islam. It is a religion gov-erned by rules and customs built on five pillars, which every Muslim has toobserve: shahadah or witnessing, salat or prayer, zakah or charity, sawm orfasting, and hajj or pilgrimage. In addition, Muslims have to follow a set ofdietary prescriptions intended to advance their well-being that determinewhich foods are halal (i.e., permitted). These laws are found in the Quranand in the Sunna, the practice of the Prophet Muhammad. They prohibitthe consumption of alcohol, pork, blood, dead meat, and meat that hasnot been slaughtered according to Islamic rulings.

It was estimated that 75% of Muslims in the United States follow theirreligious dietary laws (Hussaini, 1993), meaning that even after having emi-grated most Muslims still eat halal. For example, a recent study revealed that84% of Muslims in France always eat halal meat (Bergeaud-Blackler &Bonne, 2007). Factors explaining differences in adherence to religious diet-ary prescriptions pertain among others to social structures, for example,origin, immigration, and generation differences (Ababou, 2005; Bergeaud-Blackler & Bonne, 2007; Limage, 2000; Saint-Blancat, 2004). However, notonly religious motives determine halal meat consumption but also health;respect for animal welfare and social issues, such as ethnic identity anddegree of acculturation (Bergeaud-Blackler & Bonne, 2007; Bonne &Verbeke, 2006; Bonne, Vermeir, Bergeaud-Blackler, & Verbeke, 2007); andconvenience in cooking and eating (Bonne & Verbeke, 2006).

Investigating consumer decisions toward halal meat is topical, first,given the halal food market size and its evolution, and second, given thepolicy relevance of the issue. The Canadian International Markets Bureau(2006) reported an international halal food trade of $150 billion a year.The total spending power of Muslims in the United States was estimated at$12 billion in 1999 of which $3 billion was for meat and poultry (Riaz,1999). Although there is no estimation yet for halal trade in the European

6 K. Bonne et al.

Union, the potential market size as expressed in terms of the Muslim popula-tion is substantial, that is, estimated between 12 and 18 million individualsbefore EU enlargement (BBC News, 2005; Buijs & Rath, 2006). As wementioned earlier, 75% of Muslims would follow their dietary rules in theUnited States, meaning that even after having emigrated most Muslims stilleat halal. Assuming that this percentage holds also for Muslims in Europe,because they are migrates too, this would yield a potential halal food marketof about 10 million consumers in Europe.

SCOPE AND OBJECTIVES

Major retailers such as Carrefour or Albert Heijn are now testing to includehalal meat in their assortment. Previous attempts in a retail environment wereoften unsuccessful because of lack of insights in food consumption behaviorof Muslims (Ramdani, 2005). Fundamental problems that arise are the differ-ent definitions of halal meat and the different quality certifications. Theseproblems are expected to become important food policy issues in manyEuropean countries (Bergeaud-Blackler, 2004). Nowadays, Muslims aremaking their presence felt socially and politically and are requesting halallabeled food products (Riaz & Chaudry, 2004; Shafie & Othman, 2006).Therefore, marketers, certifying organizations and policymakers need betterinsights in halal meat consumption.

Few studies investigated food decision making within a religious contextthus far, a notable exception being a study with halal meat consumers inFrance (Bonne et al., 2007), which is taken as the point of departure for thecurrent research with a Belgian Muslim sample. The paper by Bonne et al.called for additional research among other European Muslim populations toverify to what extent the findings obtained in France, with its particular halalfood marketing environment, can be generalized to the Muslim migrationpopulations in Europe. Although the institutional environment is rather simi-lar in both Belgium and France, differences pertain mainly to the Muslimpopulation itself and the development of the halal food market and marketingenvironment. The Muslim population in Belgium consists mainly of Moroccanand Turkish immigrants and their descendents. This migration has taken placemore recently than in the case of France, and therefore the migration popula-tion is less rooted in the host (Belgian) society. Furthermore, and partly alsoowing to the more recent migration, the Belgian halal meat market is lessdeveloped than in France, resulting in a more limited offer and availabilityof halal meat in Belgium. Hence, any observed differences in the decision-making process for halal meat between both countries can potentially beattributed to differences in individual and environmental factors.

In a vein similar to Bonne et al. (2007), we aim to investigate the influ-ence of the classical components of the theory of planned behavior (TPB) on

Halal Meat Consumption 7

intention to consume meat within an ethnic and religious minority popula-tion of Muslims originating especially from North Africa and currently livingin Belgium. We argue that meat consumption decisions within a religiouscontext could differ significantly from purchase situations where religiondoes not play a key role. By extending the TPB model with self-identityand acculturation, the influence of the cultural and more specific religiouscontext is captured.

CONCEPTUAL FRAMEWORK AND RESEARCH HYPOTHESES

Conceptual Framework

According to the theory of planned behavior (TPB; Ajzen, 1985, 1991), beha-vioral intention is determined by attitude, subjective norm, and perceivedbehavioral control. Attitude is the psychological tendency that is expressedby evaluating a particular entity with some degree of favor or disfavor (Eagly& Chaiken, 1995). Subjective norm assesses the social pressure on individualsto perform or not to perform certain behavior, that is, the motivation to com-ply with significant others views. Perceived behavioral control is describedas perceptions of the extent to which the behavior is considered controllable.It assesses the degree to which people perceive that they actually have con-trol over enacting the behavior of interest (Liou & Contento, 2001). Controlfactors such as perceived availability or certain barriers may facilitate orinhibit the performance of behavior (Conner & Armitage, 1998; Verbeke &Lopez, 2005). In addition, Conner and Armitage (1998) suggest incorporatinghabit, which measures the degree of automaticity of ones behavior. Habit isdefined as behavior that has become automatic and is beyond an individualsconsciousness. Several studies using TPB to food-related behavior havesuccessfully included habit as an independent predictor of intentions(Honkanen, Olsen, & Verplanken, 2005; Verbeke, Vermeir, & Vackier, 2004).

In the present research, self-identity and acculturation are added to theTPB framework to investigate whether consumers who have different self-iden-tities and levels of acculturation hold different behavioral patterns. Self-identitycan be interpreted as a label that people use to describe themselves. It isassumed to be the result of social interaction and the cause of subsequentbehavior (Biddle, Bank, & Slavings, 1987). Self-identity is believed to influencebehavior insofar as individuals seek to perform behaviors that maintain theirsense of self, for example, eating healthy if one sees oneself as health-conscious(Conner & Armitage, 2006). Several studies (e.g., Armitage & Conner, 1999;Smith, Terry, Manstead, & Louis, 2007; Sparks & Guthrie, 1998; Sparks & Shep-herd, 1992) support the validity andpredictive ability of including self-identity inTPB. In this particular case, the influence of self-identity as a Muslim is inves-tigated through differentiating consumers who highly identify themselves asbeing Muslims from consumers with a rather low identification with being

8 K. Bonne et al.

Muslims. In a similar vein, acculturation is added to the theoretical framework toinvestigate whether consumers who are highly versus lowly acculturated relydifferently on personal attitude, subjective norm, or perceived behavioral con-trol to make their halal meat consumption decision. Acculturation in general isused to denote the process bywhich a racial or ethnic group, usually a minority,adopts the cultural patterns such as beliefs, religion, and language of the hostculture (Jamal, 1996; Laroche, Kim, & Tomiuk, 1999). Hence, it is the mannerinwhich cultural identitymay ormay not changewhen an individual is exposedto a new majority culture. The resulting framework is presented in Figure 1.

Research Hypotheses

Muslims in Europe are mainly immigrants; hence, the question arises whetherthey maintain their food habits or adapt their food choice to their new foodand cultural environment. Stodolska and Livengood (2006) concluded fromtheir study that incentives to preserve traditions such as dress, diet, and leisurestyles among Muslim minorities are strong enough to counteract pressurefrom the outgroup because these customs have strong spiritual underpinningsin addition to cultural roots. In addition, the authors posit that assimilationamong ethnoreligious minorities is slower than among ethnic groups definedby their national origin. Sekhon and Szmigin (2005) showed that even second-generation immigrants are still integrating into the dominant culture ratherthan becoming part of it because they have the desire to stay loyal to the oldergeneration, the home country, and even Eastern values.

With respect to acculturation within the TPB model, Liou and Contento(2001) concluded in their study an increased predictive power of the regres-sion models with individuals who are more acculturated. More specifically,when the degree of acculturation increased, their independent variables

FIGURE 1 Conceptual framework: theory of planned behavior with application to halal meatconsumption (based on Ajzen, 1991; Bonne, Vermeir, Bergeaud-Blackler, & Verbeke, 2007).

Halal Meat Consumption 9

(among others attitude, health concern, self-efficacy) predicted behavioralintention significantly better. Verbeke and Lopez (2005) found the languagefactor in acculturation and time of residence to influence Hispanics foodconsumption in Belgium. Furthermore, previous exploratory researchshowed that time of residence measured as generation associates with driversto consume halal meat (Bonne & Verbeke, 2006). Following this, the firsthypothesis is formulated as follows:

Hypothesis 1: The predictive power of the TPB components forbehavioral intention improves with the degree of accul-turation in the host culture.

In most Muslim societies, one can observe a strong religious socialization,much stronger than within a Western, secularized country (Geels, 1997).The author shows the importance of the Koran for the individual Muslimand concludes that it represents a basis aspect of cultural identity. Moreover,in a migration context, religion can become an important marker of identityand a significant instrument for self-categorization and, therefore, it can notonly help immigrants to find an identity (Burton, 2004) but it can also providethem something to hold on to in an overwhelming flow of changing pro-cesses (Dumont, 2003). As with national groups, religious groups derivetheir distinct identity in part from special food ways (Sarri, Higgins, & Kafa-tos, 2006, p. 7). Halal meat can be considered such a special food shapedby the Islamic dietary prescriptions. Exploratory research using the means-end-chain theory confirmed that Muslim consumers eat halal meat in orderto follow and express their religious teachings (Bonne & Verbeke, 2006).In a related vein, Bergeaud-Blackler and Bonne (2007) showed thatreligion is a very important motivator for eating halal meat. Moreover,Bergeaud-Blackler (2006) described how eating halal is part of a Muslim orIslamic identity among Muslim migrates. However, various levels ofcompliance to the religious beliefs result in different levels of behavioralcommitment (Cosgel & Minkler, 2004). In addition, previous exploratoryresearch suggests that individuals who consider themselves less as Muslimsare less motivated to comply with religious rules and=or peer group pressurewhereas they are more inclined to follow their personal attitudes and convic-tions toward halal meat (Bonne & Verbeke, 2006). Easy availability of halalmeat products could invite them to consume; on the other hand, consump-tion barriers such as price or safety issues may prevent them from eating halalmeat. On the contrary, individuals with a higher Muslim identification couldbe more inclined to follow Islamic rules and customs and therefore be moreopen for peer influence. Therefore, the following hypothesis is set forth:

Hypothesis 2: Individuals with a lower (versus higher) self-identification asMuslims will rely more on individual factors like personal

10 K. Bonne et al.

attitude, perceived behavioral control, perceived availabil-ity, and barriers compared with the motivation to comply.

Furthermore, previous research among Muslims in France revealed abetter regression model for low self-identifying Muslims (Bonne et al.,2007). Therefore, as a part of Hypothesis 2, we hypothesize that in a similarvein, a better TPB model will be found for Muslims who identify themselvesto a lesser degree with Islam.

MATERIALS AND METHOD

Sample and Procedure

Cross-sectional data were collected through a survey in Belgium during sum-mer 2006. All respondents were halal meat consumers. Recruitment wasbased on a snowball sampling technique starting from friends and increasingthrough friends and family of the initial contact persons. Furthermore, own-ers of ethnic food shops and their families, as well as workers of socioculturalorganizations and their members, were asked to participate in the study. Inparallel, a Web survey using the same questionnaire was developed. Thesnowball sampling technique was also used through e-mail and announce-ments were placed on relevant Web sites and Web fora. According to DePelsmacker and Van Kenhove (2002, p. 104), snowball sampling is recom-mended in cases where respondents are difficult to reach, which is the casefor Muslim consumers in Belgium. In total, 367 surveys were completed ofwhich 175 were obtained as a result of the snowball sampling and 192 viaelectronic contact. All questionnaires were self-administered by the respon-dent, hence avoiding potential bias from an interviewer.

Questionnaire and Scaling

The survey used a structured questionnaire in Dutch and French, whichwas pretested and refined prior to fieldwork. The questionnaire includeditems measuring the components of the proposed model of TPB. Behaviorwas measured by asking, How many times a week do you eat halal meat?on a 6-point frequency scale ranging from never to every day. Behavioralintention was measured on an 8-point scale (ranging from 0 to 7) by asking,How many times do you intend to eat halal meat in the next 7 days, todayincluded? Attitude was measured through 14 agreement statements relatedto health, taste, price, safety, animal welfare, and quality on 5-point scalesranging from totally disagree to totally agree. The measure of subjectivenorm consisted of multiple items to assess the motivation to comply. Theitems were phrased as, To what extent do you take the encouragementsto eat halal meat of the following people or institutions into consideration?

Halal Meat Consumption 11

on a 5-point scale from not at all to very much for partner, family in thecountry of origin, family in the host country, friends, religious authorities,children, and the Islamic community in general. Perceived behavioral con-trol was measured with three items. The items read, How much control doyou feel you have over eating halal meat? on a 5-point scale ranging fromno control to complete control; For me, eating halal meat is . . . on a5-point scale from very difficult to very easy; and If I would, I could easilybuy halal meat tomorrow on a 5-point scale from totally improbable tototally probable. Six items were used to assess possible barriers to consum-ing halal meat pertaining to price, availability, lack of time, lack of control,lack of information, and hygiene. All items were measured through a5-point scale ranging from totally no reason to very much a reason fornot consuming halal meat. Furthermore, perceived availability was mea-sured using four items, pertaining both to product and information avail-ability (Halal products are readily available, There is a lot of choice inhalal products, Information on halal labels is clear, There is sufficientinformation available on halal products) on a 5-point Likert scale rangingfrom totally disagree to totally agree. Also habit was measured on a 5-pointLikert scale from totally disagree to totally agree using the statements Eat-ing halal meat is something that I do without reasoning, Eating halal meatis something I do frequently, Eating halal meat is something I do withouthaving to consciously remember, Eating halal meat is something I feelweird if I dont do, and Eating halal meat is something I dont have tothink about doing it.

Finally, the questionnaire measured sociocultural (acculturation andself-identity) and individual characteristics of the respondents. Self-identitywas measured using the statements I consider myself a Muslim, I considermyself a good Muslim, and I consider myself a practicing Muslim on a 5-point Likert scale ranging from totally disagree to totally agree. Individualcharacteristics such as age, gender, origin, generation, place of residence,level of education, occupation, marital status, and number of children werealso included. To assess acculturation, 11 items were used for measuring diet-ary acculturation, ethnic social relations, language use, and media use. A5-point scale ranging from totally Maghrebian (1) to totally Belgian (5)was used.

Data Analyses

The data were analyzed using SPSS 12.0. Mean scores and standard devia-tions of all variables are reported. Internal reliability consistency of themulti-item TPB constructs was assessed using Cronbachs alpha. Factoranalyses were performed. Analyses include independent samples t tests forcomparison of means, linear correlation, and multiple regressions to examinethe determinants of halal meat consumption.

12 K. Bonne et al.

EMPIRICAL FINDINGS

Sample Characteristics

Table 1 presents the sample characteristics. Slightly more women (55.6%)than men (44.4%) participated in the survey. With respect to age, the sampleconsisted of mainly younger respondents (80.1% under 35 years; meanage 29.3 years, SD 8.1). A small but significant difference in mean agebetween the pen-and-paper sample (mean age 30.6 years, SD 9.4) andthe Internet sample (mean age 28.2 years, SD 6.5) was found. Most likely,language and educational barriers among older immigrants have led to arelatively young sample. Most respondents originate from Morocco(86.3%), which is in line with the main Muslim population origin in Belgium.First generation Muslims (i.e., those born abroad) compose 16.8% of the sam-ple and second generation (i.e., those born in Belgium or who came at orbefore the age of 6) account for 83.2% of the sample. The participantswho completed the Web-based questionnaire were significantly highereducated (v2 37.44, p< .001) and belonged significantly more to thesecond generation of Muslim consumers (v2 32.21, p< .001). Although

TABLE 1 Sociodemographic Characteristics of theSample (n 367)GenderMale 44.4Female 55.6

Age25 years 36.22635 years 43.93645 years 13.44655 years 4.9>55 years 1.6

FamilySingle 41.4Married=Living together 55.3Divorced=Widowed 3.3

OriginMoroccan 86.3Algerian 1.6Tunisian 1.4Other 10.7

EducationPrimary school 5.5High school 34.8Bachelor 33.4Master 22.9Master=Doctorate 3.3

GenerationFirst generation 16.8Second or Third generation 83.2

Halal Meat Consumption 13

no sociodemographic data on the Belgian Muslim population are available, itis assumed that there is an overrepresentation in the sample of younger andhigher educated respondents, which places some limits on generalizations tothe broader Muslim population.

Descriptive Statistics

Table 2 presents the mean scores and standard deviation of the componentsincluded in TPB. In general, Muslim consumers feel positive about halal meatconsumption and tend to comply with family and friends as much as withreligious leaders and the Muslim community to eat halal meat. Muslims con-trol their halal meat consumption partly due to a positive perceived availabil-ity of halal meat products. Practical barriers such as price, availability, andlack of time do not strongly restrain them from eating halal meat. Neverthe-less, for some Muslims issues related to the safety of halal meat could formconsumption barriers.

The acculturation score of the respondents varies strongly around arather neutral mean (M 3.18, SD 1.15). However, with respect to dietaryacculturation, the results show that Muslims have a tendency to retain theirdietary behavior (dietary acculturation score: M 2.25, SD .69), despitethe fact that 82% of the sample is born in Belgium and those born in NorthAfrica already reside in Belgium for 16.7 years on average (SD 10.6). With

TABLE 2 Cronbachs Alpha, Mean Scores (M), and Standard Deviations(SD) on 5-Point Scales for the Measured Constructs

Numberof Items

CronbachsAlpha M SD

1. Behavior 1 3.27 1.172. Behavioral intention 1 4.52 1.953. Attitude 14 .87 4.10 .58

Price 3 .64 3.77 .81Health 4 .71 4.18 .71Safety 3 .84 4.34 1.18Quality 3 .71 4.30 1.16Animal welfare 1 4.53 .91Taste 1 4.46 .86

4. Motivation to comply 4Internal 2 .89 3.90 1.27External 2 .92 3.91 1.11

5. Perceived Behavioral Control 2 .71 4.56 .726. Barriers 5

Practical 3 .77 1.48 .88Safety 2 .89 2.26 1.42

7. Availability 4 .81 3.55 1.018. Habit 5 .76 4.07 1.009. Self-Identity 3 .75 4.29 .7010. Acculturation 11 .74 3.18 1.16

14 K. Bonne et al.

TABLE3

PearsonCorrelationCoefficients

fortheMeasuredConstructs(n367)

12

34.1

4.2

56

7.1

7.2

89

10

1.Behav

ior

2.Behav

ioralintention

.637

3.Attitude

.165

.167

4.Motivationto

comply

Internal

0.069

0.068

.173

External

.114

.110

.125

5.Perceivedco

ntrol

.241

.208

.324

.151

.127

6.Perceivedav

ailability

.152

.219

.397

.173

.123

.482

7.Perceivedbarriers

Practical

barriers

.067

.084

.284

.011

.004

.243

.220

Safety

barriers

.181

.234

.233

.047

.036

.258

.228

8.Hab

it.155

.096

.191

.142

.093

.171

.126

.157

.118

9.Se

lf-Identity

.200

.143

.292

.191

.147

.242

.284

.128

.208

.216

10.Acculturation

.183

.094

.101

.111

.113

.077

.061

.015

.015

.054

.062

p